Proceedings of the Third International Conference on Invasive ...

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong><strong>Conference</strong> on Invasive SpartinaSAN FRANCISCO, CALIFORNIANovember 8-10, 2004Editors:Debra R. Ayres, Science EditorDrew W. Kerr, Management EditorStephanie D. Ericson, Copy and Layout EditorPeggy R. Ol<strong>of</strong>son, Executive Editor2010

CONFERENCE AND PROCEEDINGS FUNDINGCalifornia State Coastal ConservancySan Francisco Bay-Delta Science Consortium (Agreement #46000001642)University <strong>of</strong> California, DavisNational Science FoundationCalifornia Sea GrantSan Francisco Bay Joint VentureSan Francisco Estuary ProjectSan Francisco Estuary InstituteU.S. Environmental Protection Agency, Region 9Suggested Citation:Ayres, D.R., D.W. Kerr, S.D. Ericson and P.R. Ol<strong>of</strong>son, eds. 2010. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong><strong>Conference</strong> on Invasive Spartina, 2004 Nov 8-10, San Francisco, CA, USA. San Francisco EstuaryInvasive Spartina Project <strong>of</strong> <strong>the</strong> California State Coastal Conservancy: Oakland, CA.Printed on recycled paper.

FORWARD & ACKNOWLEDGEMENTSThe <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartina convened to provide a forum for <strong>the</strong> bestand latest Spartina research, and an opportunity to discuss new developments in Spartina science with marshland managers and technical experts who had extensive experience with this invasive genus. The <strong>the</strong>me <strong>of</strong><strong>the</strong> conference, “Linking science and management,” reflected <strong>the</strong> desire <strong>of</strong> <strong>the</strong> organizers to improve bothfields through increased exposure and interaction with each o<strong>the</strong>r, continuing and extending a focus onuniting managers and researchers that was established at earlier international conferences on this subject. Thefirst international Spartina conference was held in 1990 in Seattle, Washington, and <strong>the</strong> second in 1997 inOlympia, Washington. We are pleased to have hosted <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on InvasiveSpartina here in San Francisco, California in 2004.The success <strong>of</strong> this event was <strong>the</strong> result <strong>of</strong> hard work and support <strong>of</strong> a number <strong>of</strong> individuals andorganizations. The <strong>Conference</strong> Organizing Committee helped to guide <strong>the</strong> development <strong>of</strong> <strong>the</strong> conferenceand addressed organizational issues as <strong>the</strong>y arose. The <strong>Conference</strong> Organizing Committee included:Ms. Marcia Brockbank, Program Manager, San Francisco Estuary ProjectDr. Mike Conner, Executive Director, San Francisco Estuary InstitutePr<strong>of</strong>. Edwin Grosholz, Department <strong>of</strong> Environmental Science and Policy, University <strong>of</strong> California, DavisMr. Paul Hedge, Project Manager, National Oceans Office, AustraliaMr. Doug Johnson, Executive Director, California Invasive Plants CouncilMs. Peggy Ol<strong>of</strong>son, Director, San Francisco Estuary Invasive Spartina ProjectDr. Kim Patten, Director, Long Beach Research and Extension Unit, Washington State UniversityDr. Bobbye Smith, Regional Science Liaison to Office <strong>of</strong> Research and Development, USEPA Region 9Ms. Erin Williams, Non-native Invasive Species Program Coordinator, U.S. Fish and Wildlife ServiceThe Science Program Committee developed <strong>the</strong> conference program, selected presentations andspecial guest presenters, and provided support throughout <strong>the</strong> conference. The Science Program Committeeincluded:Dr. Lars Anderson, Exotic and Invasive Weed Research Laboratory, U.S. Department <strong>of</strong> AgricultureDr. Debra Ayres, Department <strong>of</strong> Evolution and Ecology, University <strong>of</strong> California, DavisDr. Peter Baye, San Francisco Estuary Invasive Spartina ProjectMs. Janie Civille, Department <strong>of</strong> Evolution and Ecology, University <strong>of</strong> California, DavisDr. Joshua Collins, Wetlands Program, San Francisco Estuary InstitutePr<strong>of</strong>. Edwin Grosholz, Department <strong>of</strong> Environmental Science and Policy, University <strong>of</strong> California, DavisDr. Kim Patten, Long Beach Research and Extension Unit, University <strong>of</strong> WashingtonDr. Drew Talley, Romberg Tiburon Center, San Francisco Bay National Estuarine Research ReserveMajor financial support for <strong>the</strong> conference was provided by <strong>the</strong> California State CoastalConservancy, San Francisco Bay-Delta Science Consortium (Agreement #46000001642), and University <strong>of</strong>California, Davis. Additional support for <strong>the</strong> conference was provided by <strong>the</strong> National Science Foundation,California Sea Grant, San Francisco Bay Joint Venture, San Francisco Estuary Project, San FranciscoEstuary Institute, and U.S. Environmental Protection Agency, Region 9.A luncheon and exhibit on <strong>the</strong> second day <strong>of</strong> <strong>the</strong> conference provided an opportunity for participantsto network while learning about emerging treatment technologies. We are grateful to <strong>the</strong> followingcompanies for <strong>the</strong>ir sponsorship <strong>of</strong> this luncheon: Aquatic Environments, Inc., BASF Corporation, Cygneti

Enterprises West, Inc., Helena Chemical Company, Monsanto Company, Nufarm Turf & Specialty, Wilbur-Ellis Company, and Wilco Industries.Production <strong>of</strong> this proceedings volume has been a long and fragmented process, because <strong>of</strong> <strong>the</strong>sporadic availability <strong>of</strong> time and funding. We thank <strong>the</strong> authors for both <strong>the</strong>ir time and effort in preparing<strong>the</strong>se excellent papers, and <strong>the</strong>ir patience as we worked through several years <strong>of</strong> staff changes, competingpriorities, and funding loss. We are especially grateful to Debra Ayres, who ultimately coordinated <strong>the</strong> peerreview <strong>of</strong> <strong>the</strong> 36 papers in <strong>the</strong> first three sections <strong>of</strong> this volume; to Drew Kerr, who reviewed and edited <strong>the</strong>16 papers in <strong>the</strong> last section; and to Stephanie Ericson, who worked countless hours for over four years toedit, organize, and layout <strong>the</strong> entire volume.Peggy R. Ol<strong>of</strong>sonOrganizing Committee Chair andDirector, San Francisco Estuary Invasive Spartina Projectii

TABLE OF CONTENTSForward & Acknowledgments ................................................................................................................. iIntroduction .......................................................................................................................................... viiChapter 1: Spartina Biology ................................................................................................................ 1California Cordgrass (Spartina foliosa), an Endemic <strong>of</strong> Salt Marsh Habitats along <strong>the</strong> Pacific Coast <strong>of</strong>Western North AmericaM.C. Vasey ................................................................................................................................................ 3Local and Geographic Variation in Spartina-herbivore InteractionsS.C. Pennings ............................................................................................................................................ 9Speciation, Genetic, and Genomic Evolution in SpartinaM.L. Ainouche, A. Baumel, R. Bayer, K. Fukunaga, T. Cariou and M.T. Misset .................................... 15Evolution <strong>of</strong> Invasive Spartina Hybrids in San Francisco BayD.R. Ayres and D.R. Strong .................................................................................................................... 23Evolving Invasibility <strong>of</strong> Exotic Spartina Hybrids in Upper Salt Marsh Zones <strong>of</strong> San Francisco BayM.R. Pakenham-Walsh, D.R. Ayres, and D.R. Strong ............................................................................. 29Varying Success <strong>of</strong> Spartina spp. Invasions in China: Genetic Diversity or Differentiation?S. An, Y. Xiao, H. Qing, Z. Wang, C. Zhou, B. Li, S. Shi, D. Yu, Z. Deng, and L. Chen .......................... 33Spartina densiflora x foliosa Hybrids Found in San Francisco BayD.R. Ayres and A.K.F. Lee ...................................................................................................................... 37Fungal Symbiosis: A Potential Mechanism Of Plant InvasivenessR.J. Rodriguez, R.S. Redman, M. Hoy, and N. Elder ............................................................................... 39Is Ergot a Natural Component <strong>of</strong> Spartina Marshes? Distribution and Ecological Host Range <strong>of</strong>Salt Marsh Claviceps purpureaA.J. Fisher ............................................................................................................................................... 43Mechanisms <strong>of</strong> Sulfide and Anoxia Tolerance in Salt Marsh Grasses in Relation to ElevationalZonationB.R. Maricle and R.W. Lee ...................................................................................................................... 47Effects <strong>of</strong> Salinity on Photosyn<strong>the</strong>sis in C 4 Estuarine GrassesB.R. Maricle, O. Kiirats, R.W. Lee and G.E. Edwards ............................................................................ 55Chapter 2: Spartina Distribution and Spread ................................................................................... 59A Tale <strong>of</strong> Two Invaded Estuaries: Spartina in San Francisco Bay, California and Willapa Bay,WashingtonD.R. Strong ............................................................................................................................................. 61Spartina In China: Introduction, History, Current Status and Recent ResearchS. An, H. Qing, Y. Xiao, C. Zhou, Z. Wang, Z. Deng, Y. Zhi and L. Chen ............................................... 65Spread <strong>of</strong> Invasive Spartina in <strong>the</strong> San Francisco EstuaryK. Zaremba, M. McGowan, and D. R. Ayres ........................................................................................... 73iii

Remote Sensing, LiDAR and GIS Inform Landscape and Population Ecology, Willapa Bay,WashingtonJ.C. Civille, S.D. Smith and D.R. Strong ................................................................................................ 83Implications <strong>of</strong> Variable Recruitment for <strong>the</strong> Management <strong>of</strong> Spartina alterniflora in Willapa Bay,WashingtonJ.G. Lambrinos, D.R. Strong, J.C. Civille and J. Bando ........................................................................ 87Pollen Limitation in a Wind-Pollinated Invasive Grass, Spartina alternifloraH.G. Davis, C.M. Taylor, J.G. Lambrinos, J.C. Civille and D.R. Strong ............................................... 91Invasive Hybrid Cordgrass (Spartina alterniflora x S. foliosa) Recruitment Dynamics in OpenMudflats <strong>of</strong> San Francisco BayC.M. Sloop, D.R. Ayres and D.R. Strong ................................................................................................ 95The Influence <strong>of</strong> Intertidal Zone and Native Vegetation on <strong>the</strong> Survival and Growth <strong>of</strong>Spartina anglica in Nor<strong>the</strong>rn Puget Sound, WA, USAC.E. Hellquist and R.A. Black ................................................................................................................ 99Will Spartina anglica Invade Northwards with Changing Climate?A.J. Gray and R.J. Mogg ...................................................................................................................... 103Competition among Marsh Macrophytes by Means <strong>of</strong> Vertical Geomorphological DisplacementJ.T. Morris ............................................................................................................................................ 109Modeling <strong>the</strong> Spread <strong>of</strong> Invasive Spartina Hybrids in San Francisco BayR.J. Hall, A.M. Hastings and D.R. Ayres .............................................................................................. 117Modeling <strong>the</strong> Spread and Control <strong>of</strong> Spartina alterniflora in a Pacific EstuaryC.M. Taylor, A. Hastings, H.G. Davis, J.C. Civille and F.S. Grevstad ................................................ 121Hybrid Cordgrass (Spartina) and Tidal Marsh Restoration in San Francisco Bay: If You Build ItThey Will ComeD.R. Ayres and D.R. Strong .................................................................................................................. 125Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina ..................................................................... 127Assessment <strong>of</strong> <strong>the</strong> Potential Consequences <strong>of</strong> Large-scale Eradication <strong>of</strong> Spartina anglicafrom <strong>the</strong> Tamar Estuary, TasmaniaM. Sheehan and J.C. Ellison................................................................................................................. 129Spartina Invasion Changes Intertidal Ecosystem Metabolism in San Francisco BayA.C. Tyler and E.D. Grosholz ............................................................................................................... 135Mechanistic Processes Driving Shifts in Benthic Infaunal Communities Following SpartinaHybrid Tidal Flat InvasionC. Neira, E.D. Grosholz and L.A. Levin ............................................................................................... 141Spartina alterniflora Invasions in <strong>the</strong> Yangtze River Estuary, China: A SynopsisB. Li, C-Z. Liao, X-D. Zhang, H-L. Chen, Q. Wang, Z-Y. Chen, X-J. Gan, J-H. Wu, B. Zhao,Z-J. Ma, X-L. Cheng, L-F. Jiang, Y-Q. Luo, and J-K. Chen ................................................................. 147The Role <strong>of</strong> Spartina anglica Production in Bivalve Diets in Nor<strong>the</strong>rn Puget Sound, WA, USAC.E. Hellquist and R.A. Black .............................................................................................................. 153Contrasting Effects <strong>of</strong> Spartina foliosa and Hybrid Spartina on Benthic InvertebratesE.D. Brusati and E.D. Grosholz ........................................................................................................... 161Impacts <strong>of</strong> Benthic Invertebrates on Sediment Porewater Ammonium and Sulfide: Consequencesfor Spartina Seedling Growth.U.H. Mahl, A.C. Tyler and E.D. Grosholz ............................................................................................ 165Quantifying <strong>the</strong> Potential Impact <strong>of</strong> <strong>the</strong> Spartina Invasion on Invertebrate Food Resources foriv

Community Spartina Education and Stewardship ProjectK. O'Connell ......................................................................................................................................... 265Biological Control <strong>of</strong> SpartinaF.S. Grevstad, M.S. Wecker and D.R. Strong ....................................................................................... 267Ecological Investigations <strong>of</strong> Natural Enemies for an Interstate Biological Control Programagainst Spartina GrassesD. Viola, L. Tewksbury and R. Casagrande ......................................................................................... 273Potential for Sediment-Applied Acetic Acid for Control <strong>of</strong> Spartina alternifloraL.W.J. Anderson ................................................................................................................................... 277vi

INTRODUCTION“Maritime Spartina species define and maintain <strong>the</strong> shoreline along broad expanses <strong>of</strong> temperatecoasts where <strong>the</strong>y are native. The large Spartina species grow lower on <strong>the</strong> tidal plane than o<strong>the</strong>rvascular plants; tall, stiff stems reduce waves and currents to precipitate sediments from turbidestuarine waters. With <strong>the</strong> right conditions, roots grow upward through harvested sediments toelevate <strong>the</strong> marsh. This engineering can alter <strong>the</strong> physical, hydrological, and ecologicalenvironments <strong>of</strong> salt marshes and estuaries. Where native, Spartinas are uniformly valued, mostlyfor defining and solidifying <strong>the</strong> shore. The potential to terrestrialize <strong>the</strong> shore was <strong>the</strong> rationale <strong>of</strong>many <strong>of</strong> <strong>the</strong> scores <strong>of</strong> Spartina introductions. In a time <strong>of</strong> rising sea levels, <strong>the</strong>se plants are valued asa barrier to <strong>the</strong> sea in native areas and in China and Europe where <strong>the</strong>y have been cultivated. Incontrast, in North America, Australia, Tasmania, and New Zealand, and in some parts <strong>of</strong> China,nonnative Spartinas are seen as a bane both to ecology and to human uses <strong>of</strong> salt marshes andestuaries” (Strong and Ayres 2009).Three conferences on Spartina were held spread should be developed to understandduring <strong>the</strong> last two decades to detail current population dynamics that could inform controlunderstanding <strong>of</strong> <strong>the</strong> biological, ecological, and strategies. Attendees recommended <strong>the</strong>se basicpolitical repercussions <strong>of</strong> invasive Spartina and its steps to address invasive Spartina control: identifycontrol. The first international conference on a lead agency, appoint a single person to be ainvasive Spartina was held in Seattle, Washington, “Spartina Czar,” inventory <strong>the</strong> extent <strong>of</strong> invasion,USA in 1990 (Mumford et al. 1991). In addition to identify routes <strong>of</strong> spread, enlist public support,organizers, <strong>the</strong>re were 31 attendees, 17 evaluate various control methods, and continuepresentations, and discussion on outcomes <strong>of</strong> four workgroup discussions.regional strategy workgroups (all on <strong>the</strong> Pacific The second conference was held in Olympia,coast <strong>of</strong> <strong>the</strong> United States). While most <strong>of</strong> <strong>the</strong> Washington in 1997 (Patten 1997). There werepresenters were local to Washington State, 130 attendees from five countries (United States,Spartina researchers also came from <strong>the</strong> U.S. East Canada, Australia, New Zealand, and UnitedCoast, <strong>the</strong> United Kingdom and New Zealand. Kingdom) at <strong>the</strong> 29 presentations. Like <strong>the</strong> firstMany questions were posed at <strong>the</strong> conference. conference, it included presentations on invasiveAttendees deliberated on potential changes to <strong>the</strong> Spartina biology, distribution, impacts, andhabitat (sedimentation rates and detrital control. However, <strong>the</strong> 1997 conference also hadbreakdown) and fauna (invertebrates, fish, and presentations addressing <strong>the</strong> “political ecology <strong>of</strong>birds) as a result <strong>of</strong> invasion, and during and Spartina control” (Perkins 1997). Topics <strong>of</strong> <strong>the</strong>sefollowing control. As well, questions arose on <strong>the</strong> papers included public activities, and riskefficacies <strong>of</strong> biological, chemical and mechanical assessment associated with control measures; “<strong>the</strong>control methods, and potential constraints on hysteria over <strong>the</strong> cordgrass” (Cohen 1997) andmethod — environmental, political and/or “<strong>the</strong> crisis <strong>of</strong> civil dialogue” (Markham 1997)economic. It was agreed that models <strong>of</strong> Spartina were noted. New topics at <strong>the</strong> 1997 conferencevii

included <strong>the</strong> documentation <strong>of</strong> hybrids betweenS. alterniflora and S. foliosa in San Francisco Bay(Daehler and Strong 1997), <strong>the</strong> use <strong>of</strong> GISmapping and associated databases to inventoryand model future spread <strong>of</strong> Spartina (Harringtonet al. 1997), <strong>the</strong> first empirical data on drift <strong>of</strong>Spartina seed (Sayce et al. 1997), increases inshorebird populations after Spartina die-back in<strong>the</strong> United Kingdom (Gray et al. 1997), andcomparisons <strong>of</strong> benthic invertebrates betweenSpartina stands and mudflat (Luiting et al. 1997).Attendees suggested that additional research beconducted to develop models <strong>of</strong> populationdynamics and genetic shifts, to monitor spread andenvironmental impacts <strong>of</strong> invasive Spartina, toimprove mapping and modeling in order to helpprioritize control and evaluate costs. They alsoproposed carrying out research to quantifysediment accretion rates in invaded areascompared with non-invaded habitats, to evaluatephysiological tolerance and vulnerabilities <strong>of</strong>Spartina, assess environmental impacts associatedwith Spartina removal and specific controltechniques.The third international conference on invasiveSpartina was held in San Francisco, California in2004. Hosted by <strong>the</strong> California State CoastalConservancy and <strong>the</strong> U.S. EnvironmentalProtection Agency, <strong>the</strong> meeting drew 147presenters and attendees from seven countries,including China, France, <strong>the</strong> United States,Canada, Australia, New Zealand, and <strong>the</strong> UnitedKingdom. Fifty-seven oral and posterpresentations were shared; 40 on science and 17on control or management. Of <strong>the</strong> 11 papers onSpartina biology in this volume, half were focusedon genetics, with Spartina hybridizations being<strong>the</strong> primary context for genetics studies. Thirteenpapers on distribution and spread <strong>of</strong> Spartinaincluded spatio-temporal analysis using GIStechniques, predictions <strong>of</strong> spread due tocompetition, climate change, evolution, andrestoration; dynamics <strong>of</strong> seed set and seedlingrecruitment; and two simulation models describingspread. Studies on impacts to <strong>the</strong> ecosystemranged from increased sedimentation underSpartina canopies, to changes in invertebrateviiicommunities and food webs, to shifts in birdcommunities, populations, and behaviors due to<strong>the</strong> Spartina invasion.Of <strong>the</strong> 17 papers on control and management,seven papers were focused on control per se,while five detailed <strong>the</strong> <strong>of</strong>tentimes complex andbureaucracy-laden routes <strong>of</strong> Spartina control inSan Francisco Bay, Washington State, Tasmaniaand New Zealand — mostly benefiting from <strong>the</strong>perspective <strong>of</strong> hindsight. Two papers considered<strong>the</strong> long and hard task that lies ahead to restoreestuaries that have undergone major increases inmarsh elevation due to Spartina invasion — whatHacker and Dethier (this volume) called <strong>the</strong>“legacy <strong>of</strong> invasion.”A primary reason for hosting <strong>the</strong> 2004Spartina conference in San Francisco was to giveparticipants a first-hand look at <strong>the</strong> invasion,through helicopter and “mud-level” field trips —with an eye toward assessing <strong>the</strong> extent <strong>of</strong> <strong>the</strong>invasion, and determining whe<strong>the</strong>r control was, infact, feasible or even necessary. These questionswere discussed by an expert panel at <strong>the</strong> end <strong>of</strong> <strong>the</strong>conference. Overwhelmingly, <strong>the</strong> panel agreedthat control was both possible and urgentlyneeded. With this unequivocal recommendation,<strong>the</strong> California State Coastal Conservancyproceeded to fund and coordinate an aggressiveregional program to eradicate introduced Spartina(including hybrids) from <strong>the</strong> San FranciscoEstuary. By fall 2010, <strong>the</strong> program hadsuccessfully reduced <strong>the</strong> area <strong>of</strong> invasive Spartinafrom more than 800 net acres in 2005, to less than100 net acres, a reduction <strong>of</strong> 90 percent. Theprogram is ongoing, with full effective eradicationexpected before 2020.The papers that follow in this conferenceproceedings volume detail <strong>the</strong> advances that havebeen made in our understanding <strong>of</strong> Spartinabiology and demography, <strong>the</strong> pr<strong>of</strong>oundenvironmental effects that result from <strong>the</strong>seinvasions worldwide, and suggest <strong>the</strong> scope <strong>of</strong>unresolved issues.Debra AyresUniversity <strong>of</strong> California, Davis

REFERENCESCohen E.S. 1997. Local Spartina management planning forcommunity oriented research and development. In: Patton, K.,ed. Second <strong>International</strong> Spartina <strong>Conference</strong> <strong>Proceedings</strong>.Washington State University Cooperative Extension, LongBeach, WA. pp. 58-59.Daehler C.C. and D.R. Strong. 1997. Invasions and hybridization <strong>of</strong>cordgrasses, Spartina spp. In: Patton, K., ed. Second<strong>International</strong> Spartina <strong>Conference</strong> <strong>Proceedings</strong>. WashingtonState University Cooperative Extension, Long Beach, WA. p. 17.Gray A.J., A.F. Raybould and S.L. Brown. 1997. Theenvironmental impact <strong>of</strong> Spartina anglica: past, present andpredicted. In: Patton, K., ed. Second <strong>International</strong> Spartina<strong>Conference</strong> <strong>Proceedings</strong>. Washington State UniversityCooperative Extension, Long Beach, WA. pp. 39-45.Hacker S.D. and M.N. Dethier. Where do we go from here?Alternative control and restoration trajectories for a marine grass(Spartina anglica) invader in different habitat types. In: Ayres,D.R., D.W. Kerr, S.D. Ericson and P.R. Ol<strong>of</strong>son, eds.<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on InvasiveSpartina, 2004 Nov 8-10, San Francisco, CA, USA. SanFrancisco Estuary Invasive Spartina Project <strong>of</strong> <strong>the</strong> CaliforniaState Coastal Conservancy: Oakland, CA. (this volume).Harrington J.A., M.B. Harrington and C.J. Berlin. 1997. ModelingSpartina in Willapa Bay. . In: Patton, K., ed. Second<strong>International</strong> Spartina <strong>Conference</strong> <strong>Proceedings</strong>. WashingtonState University Cooperative Extension, Long Beach, WA. pp.23-26.Luiting V.T., J.R. Cordell, A.M. Olso and C.A. Simenstad. 1997.Does exotic Spartina alterniflora change benthic invertebrateassemblage? In: Patton, K., ed. Second <strong>International</strong> Spartina<strong>Conference</strong> <strong>Proceedings</strong>. Washington State UniversityCooperative Extension, Long Beach, WA. pp. 48-50.Markham D.C. 1997. Sustainable development and Spartinacontrol – a case study on <strong>the</strong> crisis <strong>of</strong> civil dialog. In: Patton, K.,ed. Second <strong>International</strong> Spartina <strong>Conference</strong> <strong>Proceedings</strong>.Washington State University Cooperative Extension, LongBeach, WA. pp. 64-67.Mumford T.F, P. Peyton., J.R. Sayce. and S. Harbell, eds. 1991.Spartina Workshop Record. Washington Sea Grant Program,University <strong>of</strong> Washington, Seattle.Perkins J.H. 1997. The political ecology <strong>of</strong> Spartina andglyphosate. In: Patton, K., ed. Second <strong>International</strong> Spartina<strong>Conference</strong> <strong>Proceedings</strong>. Washington State UniversityCooperative Extension, Long Beach, WA. p. 82.Patten, K., ed. 1997. Second <strong>International</strong> Spartina <strong>Conference</strong><strong>Proceedings</strong>. Washington State University CooperativeExtension, Long Beach, WA.Sayce K., B. Dumbauld and J. Hidy. 1997. Seed dispersal in drift <strong>of</strong>Spartina alterniflora. In: Patton, K., ed. Second <strong>International</strong>Spartina <strong>Conference</strong> <strong>Proceedings</strong>. Washington State UniversityCooperative Extension, Long Beach, WA. pp. 27-31.Strong, D.R. and D.R. Ayres. 2009. Spartina Introductions andConsequences in Salt Marshes: Arrive, Survive, Thrive, andsometimes Hybridize. In: Silliman B.R., E. Grosholz and M.Bertness, eds. Human Impacts on Salt Marshes: A GlobalPerspective. University <strong>of</strong> California Press, Berkeley, CAix

CHAPTER ONESpartina Biology

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyCALIFORNIA CORDGRASS (SPARTINA FOLIOSA), AN ENDEMIC OF SALT MARSH HABITATSALONG THE PACIFIC COAST OF WESTERN NORTH AMERICAM.C. VASEYDepartment <strong>of</strong> Biology, San Francisco State University, San Francisco, CA 94132; andDepartment <strong>of</strong> Environmental Studies, University <strong>of</strong> California, Santa Cruz, CA 95064; mvasey@sfsu.eduCalifornia cordgrass (Spartina foliosa) is an endemic <strong>of</strong> <strong>the</strong> California Floristic Province in westernNorth America. This paper reviews its historic and contemporary distribution, pr<strong>of</strong>iles its habitatcharacteristics including key adaptive traits, and makes <strong>the</strong> case that S. foliosa should be recognizedas a foundation species within salt marshes characteristic <strong>of</strong> <strong>the</strong> California Floristic Province. Thisinformation is <strong>the</strong>n considered in <strong>the</strong> context <strong>of</strong> introgressive hybridization between S. foliosa and<strong>the</strong> closely related Spartina alterniflora. Clearly, <strong>the</strong> hybrid between S. foliosa and S. alterniflora isprogressively spreading genes <strong>of</strong> S. alterniflora into pure stands <strong>of</strong> S. foliosa throughout <strong>the</strong> SanFrancisco Bay estuary. While this raises a concern about <strong>the</strong> potential “extinction” <strong>of</strong> S. foliosa as adistinct genotype, it is suggested that a greater problem may be <strong>the</strong> ecological implications <strong>of</strong> thisintrogression for salt marshes in <strong>the</strong> San Francisco Bay estuary. Ra<strong>the</strong>r than eradication per se, it issuggested that focused containment may be <strong>the</strong> best strategy for minimizing <strong>the</strong> impact <strong>of</strong> thishybrid on salt marshes in <strong>the</strong> California Floristic Province <strong>of</strong> this region. Fur<strong>the</strong>r, continued adaptivemanagement studies that evaluate <strong>the</strong> impacts <strong>of</strong> this species on salt marsh restoration arerecommended.INTRODUCTIONIn this review, I will examine what we know aboutCalifornia cordgrass and attempt to frame this discussion in<strong>the</strong> context <strong>of</strong> conservation management challenges thatinvasive Spartina alterniflora pose to <strong>the</strong> San Francisco Bayestuary. The approach will be to discuss what is known <strong>of</strong><strong>the</strong> historic and contemporary distribution <strong>of</strong> Spartinafoliosa, to focus on <strong>the</strong> suite <strong>of</strong> adaptive traits that makes S.foliosa such an important component <strong>of</strong> low marsh habitatsin west coast estuaries, and to pr<strong>of</strong>ile S. foliosa as a“foundation species” in this environment. I will <strong>the</strong>nsummarize this information in <strong>the</strong> context <strong>of</strong> threats posedby hybridization <strong>of</strong> S. foloiosa with <strong>the</strong> invasive Spartinaalterniflora.HISTORIC AND CONTEMPORARY DISTRIBUTION OFSPARTINA FOLIOSAMacDonald and Barbour (1974) conducted a survey <strong>of</strong>salt marsh vegetation along <strong>the</strong> North American Pacificcoast ranging from Point Barrow, Alaska to Cabo San Lucas,Baja California. In general, <strong>the</strong>y found three dominant saltmarsh vegetation communities over this extensive range: anarctic, boreal, and north temperate assemblage is dominatedby nor<strong>the</strong>rn grasses such as Deschampsia caespitosa andsedge species such as Carex lygnbyei, which ranges from <strong>the</strong>Seward Peninsula down to Drake’s Estero in Marin County;a south temperate and subtropical assemblage ranges fromSan Francisco Bay to Laguna San Ignacio in Baja California,which is <strong>the</strong> assemblage occupied by S. foliosa; and south <strong>of</strong>this region, tropical mangrove forests and scrub assemblagesdominate estuarine tidal wetlands.The historic distribution <strong>of</strong> S. foliosa occurs in thisintermediate zone, also known as <strong>the</strong> California FloristicProvince (Raven and Axelrod 1978). The California FloristicProvince occurs along a cismontane region from sou<strong>the</strong>rnOregon down through nor<strong>the</strong>rn Baja California. It ischaracterized by a Mediterranean-type climate consisting <strong>of</strong>long, hot and dry summers punctuated by short, wet and coldwinters. The California Floristic Province has recently beenrecognized as one <strong>of</strong> twenty-five global “biodiversity hotspots,” with over 48% <strong>of</strong> <strong>the</strong> plant species being endemics(Myers et al. 2000; Calsbeek et al. 2003). Spartina foliosa isamong <strong>the</strong>se endemic species.The particular distribution <strong>of</strong> S. foliosa occurs in twomajor disjunct regions: (1) nor<strong>the</strong>rn California (SanFrancisco Bay region) and, approximately 565 kilometersaway, (2) a series <strong>of</strong> geographically proximal populationsfrom Orange County in sou<strong>the</strong>rn California through southcentralBaja California. The historic nor<strong>the</strong>rn Californiadistribution has been confused by one case <strong>of</strong>misidentification and two o<strong>the</strong>r cases <strong>of</strong> relatively recentnor<strong>the</strong>rn range extensions. The case <strong>of</strong> misidentification wasfirst recognized by P. Faber (personal communication) whoperceived that <strong>the</strong> extensive stands <strong>of</strong> Spartina longrecognized in Humboldt Bay wetlands were actuallySpartina densiflora (a caespitose species from SouthAmerica) ra<strong>the</strong>r than S. foliosa. Later, Spicher and Josselyn(1985) published a confirmation <strong>of</strong> this observation.Barnhart et al. (1992) speculate that S. densiflora may havebeen introduced to Humboldt Bay as long ago as <strong>the</strong> 1860sas part <strong>of</strong> shingle or dry ballast deposited during <strong>the</strong> height<strong>of</strong> shipping activities in this formerly bustling port. Spicher-3-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaand Josselyn (1985) also reported a local population <strong>of</strong> S.foliosa at Bodega Bay; however, this population was almostcertainly not present when Barbour (1970) and MacDonaldand Barbour (1974) reported on salt marsh vegetation inBodega Bay wetlands. In <strong>the</strong> 1990s, a large population <strong>of</strong> S.foliosa was observed colonizing <strong>the</strong> accreting delta <strong>of</strong>Lagunitas Creek in sou<strong>the</strong>rn Tomales Bay (Baye 2004,personal communication). Howell (1949, 1970) in his flora<strong>of</strong> Marin, however, does not mention this locality for S.foliosa nor do any historic herbarium collections suggest thatei<strong>the</strong>r <strong>the</strong> Bodega Bay or Tomales Bay populations occurredprior to <strong>the</strong> latter part <strong>of</strong> <strong>the</strong> twentieth century. MacDonaldand Barbour (1974) specifically remark upon <strong>the</strong> absence <strong>of</strong>S. foliosa in Tomales Bay.On <strong>the</strong> o<strong>the</strong>r hand, Howell (1949, 1970) and historicherbarium collections do confirm that S. foliosa hashistorically been found in Drake’s Estero in sou<strong>the</strong>rn PointReyes as well as <strong>the</strong> relatively nearby Bolinas Lagoon inMarin County. O<strong>the</strong>r than <strong>the</strong>se two outer coastal estuaries,all o<strong>the</strong>r historic nor<strong>the</strong>rn California populations <strong>of</strong> S. foliosaare found within <strong>the</strong> lower portion <strong>of</strong> <strong>the</strong> San Francisco Bayestuary, ranging from tidal wetlands <strong>of</strong> South San FranciscoBay up through San Pablo Bay and <strong>the</strong> lower tidal wetlands<strong>of</strong> <strong>the</strong> Petaluma River, Sonoma Creek, and <strong>the</strong> Napa River.Spartina foliosa continues up into <strong>the</strong> Carquinez Straightsbut is largely unknown from Suisun Bay. In this easternregion, it is apparently constrained by low brackish marshvegetation dominated by Schoenoplectus acutus andSchoenoplectus californicus and its distribution is dynamicdepending on decadal shifts in <strong>the</strong> salinity gradient in thisportion <strong>of</strong> <strong>the</strong> estuary. It is somewhat remarkable that <strong>the</strong>reare no historic or current occurrences <strong>of</strong> S. foliosa from <strong>the</strong>Golden Gate south along <strong>the</strong> central California coast to PointConception, including likely tidal wetlands such as ElkhornSlough and Morro Bay. Peter Baye (personalcommunication 2004) suggests that before <strong>the</strong> construction<strong>of</strong> jetties, <strong>the</strong>se central Californian lagoons were not tidalduring summers (due to low flows) so it is possible that <strong>the</strong>ydid not provide suitable habitat.In summary, S. foliosa in nor<strong>the</strong>rn California washistorically concentrated within <strong>the</strong> lower, saline reaches <strong>of</strong><strong>the</strong> San Francisco Bay estuary where it was an importantconstituent <strong>of</strong> <strong>the</strong> extensive salt marshes that historicallydominated tidal wetlands in this region. It appears to beslowly moving north along <strong>the</strong> coast and is now present inBodega Bay. Its eastern distribution in <strong>the</strong> San FranciscoBay estuary is dynamic and dependant on long-term shifts <strong>of</strong><strong>the</strong> salinity gradient in <strong>the</strong> Carquinez Straight and lowerSuisun Bay subregion.The second historic center <strong>of</strong> distribution for S. foliosais relatively well documented by MacDonald and Barbour(1974), Trnka and Zedler (2000), and Wiggins (1980). Thenor<strong>the</strong>rnmost historic locality is Mugu Lagoon (OrangeCounty), considerably south <strong>of</strong> Point Conception. There is<strong>the</strong>n a fairly continuous occurrence <strong>of</strong> S. foliosa whereverconditions are suitable down to <strong>the</strong> Tijuana Estuary in SanDiego County (e.g. Anaheim Bay, Bolsa Chica Bay,Newport Bay, Santa Margarita River, Los PeñasquitosLagoon, San Elijo Lagoon, San Diegito Lagoon, MissionBay, and San Diego Bay). This distribution continues in arelatively consistent fashion along <strong>the</strong> coast <strong>of</strong> BajaCalifornia in Estero de Punta Banda, Bahia de San Quintin,Laguna Guerrero Negro, Ojo de Liebre, Laguna San Ignacio,and Bahia de la Magdalena. It is <strong>of</strong> interest that Bahia de laMagdalena is at a boundary between tropical and subtropicalmarine ecosystems. It may be that S. foliosa is able to extendfur<strong>the</strong>r south than most California Floristic Province plantendemics because it is responding as much to <strong>the</strong>se marineinfluences as to Mediterranean climate influences. South <strong>of</strong>Laguna Guerrero Negro, S. foliosa begins to co-occur withmangroves and mangrove-like shrubs such as Rhizophoramangle, Laguncularia racemosa, Conocarpus erecta, andAvicennia germinans. MacDonald and Barbour (1974) pointout that <strong>the</strong>se same mangrove associates co-occur withsmooth cordgrass (S. alterniflora) in marshes near Tabascoon <strong>the</strong> Gulf <strong>of</strong> Mexico. Thus, <strong>the</strong> closest geographic distancebetween California cordgrass and smooth cordgrass is at S.foliosa’s sou<strong>the</strong>rn distributional limit. There are only about1,200 kilometers separating <strong>the</strong> Gulf coast <strong>of</strong> Mexico from<strong>the</strong> Pacific coast <strong>of</strong> Baja California.HABITAT CHARACTERISTICS AND ADAPTIVE TRAITSThroughout its entire range, S. foliosa occupies adistinctive zone associated with coastal salt marshes. In mostcases, it is <strong>the</strong> only vascular plant species present in thiszone, which roughly corresponds to bay or channel marginsat or near <strong>the</strong> mean high tide line (Schoenherr 1995). Everyday, S. foliosa habitat is inundated by salt water for one toseveral hours. Fur<strong>the</strong>r, it is exposed to <strong>the</strong> forces <strong>of</strong> waveaction and high velocity channel flows. This combination <strong>of</strong>high salinity, prolonged inundation, and daily hydrologicaldisturbance is obviously beyond <strong>the</strong> tolerance limits foro<strong>the</strong>r salt marsh vascular plant species. Spartina foliosa isable to tolerate <strong>the</strong>se extreme conditions because <strong>of</strong> a uniquesuite <strong>of</strong> adaptive traits that include: (1) C4 photosyn<strong>the</strong>sis, aphysiological drought-adaptive mechanism which meansthat about half as much water is needed for <strong>the</strong> same amount<strong>of</strong> carbohydrate produced; (2) epidermal salt glands whichalso help to conserve water by reducing salt concentrationsin cell sap; (3) possession <strong>of</strong> deep rhizomes (about 15 cm)which regenerate new segments each season and createextensive, branching mats that penetrate anaerobic sedimentsand anchor shoreline habitat; (4) stems and leaves that arecomposed <strong>of</strong> aerenchymous tissue that allows oxygen totravel from aerial portions <strong>of</strong> <strong>the</strong> plant down to <strong>the</strong> rhizomesthat are embedded in anaerobic sediments; and (5) stems andleaves that are about one meter tall, enabling vegetativestructures to remain above extreme high tide levels so thatoxygen can reliably be transported to <strong>the</strong> roots.-4-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyThese vegetative adaptive traits are also complementedby a number <strong>of</strong> important reproductive adaptations thatinclude: (1) wind pollination; (2) asynchronous flowering;i.e. protogyny (i.e. females flower first, <strong>the</strong>n males comeinto bloom) which promotes out-crossing yet allows someoverlap in flowering time so that individuals can partiallyself-fertilize (reproductive insurance); (3) production <strong>of</strong>large numbers <strong>of</strong> small seeds that can be dispersed byflotation (hydrochory); (4) seeds with awns and o<strong>the</strong>rstructures that make <strong>the</strong>m available for attachment to birdfea<strong>the</strong>rs and potential avian dispersal (Vivian-Smith andStiles 1994); (5) vegetative rhizome fragments that can betransported by water to fur<strong>the</strong>r spread individual genotypes;(6) phenology cued into optimal conditions for flowering(late summer), seed set (fall) and dispersal during winterstorm events; and (7) timing <strong>of</strong> dispersal coinciding withfresh conditions in <strong>the</strong> estuary during winter flooding, suchthat fresh water conditions promote germination andestablishment.The unique suite <strong>of</strong> vegetative and reproductiveadaptations <strong>of</strong> S. foliosa have resulted in its successfuloccupation <strong>of</strong> an important low marsh niche within westcoast tidal wetlands found in <strong>the</strong> California FloristicProvince (and beyond through south-central BajaCalifornia). In fact, as I will argue below, S. foliosa could beviewed as a “foundation species” i.e. a species that has apr<strong>of</strong>ound effect on tidal wetland functions such assuccession, productivity, and habitat structure. Thesefunctions ultimately facilitate <strong>the</strong> occupation <strong>of</strong> such tidalwetlands by a myriad <strong>of</strong> microbial, algal, plant, invertebrate,fish, and bird species.SPARTINA FOLIOSA AS A FOUNDATION SPECIESThe concept <strong>of</strong> a foundation species has been articulatedby Dayton (1972), Bruno and Bertness (2001), and Bruno etal. (2003) in <strong>the</strong> context <strong>of</strong> recognizing facilitation as anessential element <strong>of</strong> contemporary ecological <strong>the</strong>ory. Unlikekeystone species, which are proportionately rare incommunities and yet exert a disproportionate influence overcommunity structure through processes such as predation orecological engineering, foundation species are habitatformingdominant species that provide <strong>the</strong> framework for <strong>the</strong>assembly <strong>of</strong> an entire community. California cordgrassappears to be an excellent example <strong>of</strong> a foundation species.The importance <strong>of</strong> S. foliosa as an initiator <strong>of</strong> tidalwetland succession in San Francisco Bay tidal wetlands haslong been appreciated. For example, Howell (1949) makes<strong>the</strong> following comment: “Pacific (i.e. California) cord grassis generally <strong>the</strong> first plant to appear on tidal flats where itfrequently establishes broad pure stands. Later it issucceeded by Salicornia and a more diversified salt marshassociation as higher ground is built up around it. In thislater association Spartina still occurs as a narrow fringealong tidal sloughs and also occasionally as a localizedcolony in low areas”. In recently restored tidal wetlands in<strong>the</strong> bay today, such as Carl’s Marsh in Sonoma County orPond 2A in Napa County, this exact pattern has beenobserved. Spartina foliosa is usually <strong>the</strong> first species tocolonize barren mud flats by floating seeds and rhizomefragments. As clones grow and establish, rhizomatous matstrap sediment and facilitate <strong>the</strong> incremental rise <strong>of</strong> a marshplain. As elevations become suitable, o<strong>the</strong>r vascular plantspecies colonize <strong>the</strong> emerging marsh plain and eventuallydisplace S. foliosa to <strong>the</strong> accreting edges <strong>of</strong> <strong>the</strong> marsh,margins <strong>of</strong> drainage channels, or it persists in lowdepressions in <strong>the</strong> marsh where duration <strong>of</strong> inundationapparently detracts o<strong>the</strong>r marsh plain species from becomingestablished (personal observation).Once established, dense low marsh stands <strong>of</strong> S. foliosaprovide a key role in marsh productivity. Each year, S.foliosa rhizomes put out fresh shoots that rapidly grow intotall, mature stems and leaves. A considerable phase <strong>of</strong>carbon fixation <strong>the</strong>n takes place before flowering begins inmid summer (June). After fruiting in <strong>the</strong> fall, stems die backand much <strong>of</strong> <strong>the</strong> vegetative matter <strong>of</strong> <strong>the</strong> stand contributes to<strong>the</strong> detritus base <strong>of</strong> <strong>the</strong> wetland food web. Fur<strong>the</strong>r, stemwrack is washed up to <strong>the</strong> marsh/upland transition zone,providing habitat and nutrients for a variety <strong>of</strong> organismsthat inhabit this interface. Along with this direct contributionto <strong>the</strong> wetland food web, S. foliosa also provides animportant indirect contribution because its rhizomatous massprovides key habitat for a diverse assemblage <strong>of</strong> algae,including nitrogen-fixing cyanobacteria. Dawson and Foster(1982) describe this phenomenon as follows: “The mudbeneath and between California cord grass is covered withvarious algae, including films <strong>of</strong> golden-colored, unicellulardiatoms, <strong>the</strong> brownish-green Enteromorpha, redPolysiphonia, <strong>the</strong> brownish-green, siphonous Vaucheria, andblue-green algae. All <strong>of</strong> <strong>the</strong>se kinds <strong>of</strong> algae can beimportant contributors to marsh productivity and, inaddition, some <strong>of</strong> <strong>the</strong> blue-greens can convert nitrogen gasinto o<strong>the</strong>r nitro-nutrients for <strong>the</strong> algae as well as <strong>the</strong>flowering plants.”Stands <strong>of</strong> S. foliosa also provide habitat structure that isimportant for a number <strong>of</strong> species that occupy tidalwetlands. The relationship between California cordgrass andCalifornia and light-footed clapper rails (both subspecies <strong>of</strong>Rallus longirostris) is an excellent case in point. Both forcover and nesting habitat, California cordgrass is essentialfor <strong>the</strong> success <strong>of</strong> this species. Boyer and Zedler (1996) alsopoint out that insects occupy stands <strong>of</strong> S. foliosa and <strong>the</strong>seundoubtedly provide food resources for passerine birds (e.g.marsh wrens and song sparrows) that live in salt marshhabitats.Finally, because <strong>of</strong> its importance as a foundationspecies, particularly in terms <strong>of</strong> succession, S. foliosa is alsoan essential element for tidal wetland restoration projects.Seeds, rhizome fragments, and plugs <strong>of</strong> S. foliosa can beused to revegetate formerly diked wetlands that are restored-5-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinato tidal action. An interesting example is <strong>the</strong> restoration <strong>of</strong>Muzzi Marsh in Marin County which, because <strong>of</strong> an alteredhydrology, is still dominated by large meadows <strong>of</strong> Californiacordgrass. One <strong>of</strong> <strong>the</strong> goals <strong>of</strong> this project was to recoverhabitat for California clapper rails and, indeed, this specieshas colonized Muzzi Marsh and now hosts a thriving,nesting population (Page and Evens 1987). In San FranciscoBay, it is likely that <strong>the</strong> use <strong>of</strong> S. foliosa material in marshrestoration projects will have to be much more carefullycontrolled because <strong>of</strong> <strong>the</strong> possibility that hybrid sourcematerial may contaminate seed sources. In fact, one <strong>of</strong> <strong>the</strong>greatest challenges <strong>of</strong> <strong>the</strong> invasion <strong>of</strong> S. alterniflora and itsintrogression with S. foliosa revolves around <strong>the</strong> desperateneed and opportunity to restore historic tidal wetlands inlower San Francisco Bay in <strong>the</strong> face <strong>of</strong> <strong>the</strong> potential for <strong>the</strong>hybrids to occupy restoration sites and send <strong>the</strong> successionaltrajectory into a completely unknown realm.SUMMARY AND IMPLICATIONS OF HYBRIDIZATIONSpartina foliosa has great significance as a foundationspecies, shaping <strong>the</strong> structure and function <strong>of</strong> salt marshassemblages within <strong>the</strong> California Floristic Province. Theecological dilemma is that its genetically-compatible sisterspecies, S. alterniflora (Baumel et al. 2002), possessessimilar adaptive traits combined with greater robustness,fertility, and ecological amplitude than S. foliosa.Fortunately, a majority <strong>of</strong> <strong>the</strong> range and individualpopulations <strong>of</strong> S. foliosa lie in <strong>the</strong> sou<strong>the</strong>rn center <strong>of</strong> itsdistribution from Point Mugu in Orange County throughBahia Magdalena in Baja California. Unfortunately, since70% <strong>of</strong> <strong>the</strong> acreage <strong>of</strong> salt marsh habitats in Californiaoccurs in <strong>the</strong> San Francisco Bay estuary, it is likely that <strong>the</strong>largest extant populations <strong>of</strong> S. foliosa are at risk due tocontamination by hybridization. As pointed out by Daehlerand Strong (1997), Ayres et al. (2003), and Baye (2004),given <strong>the</strong> rapid spread S. foliosa x S. alterniflora hybrids, <strong>the</strong>array <strong>of</strong> short form and tall form recombinants, <strong>the</strong> malesuperiority <strong>of</strong> <strong>the</strong>se hybrids (Antilla et al. 1998, 2000), and<strong>the</strong> greater ecological amplitude <strong>of</strong> <strong>the</strong>se hybrid genotypes, itis probable that within <strong>the</strong> San Francisco Bay estuary, S.foliosa will ultimately be transformed into a new entity thatcombines traits <strong>of</strong> both S. foliosa and S. alterniflora.Just as S. foliosa is a foundation species in SanFrancisco Bay tidal wetlands, and a myriad <strong>of</strong> species havelife history traits that are adapted to this species, it is likelythat <strong>the</strong> hybrid entity will take its place as a foundationspecies as well, triggering a cascade <strong>of</strong> responses by specieswhose ecological roles are entwined with S. foliosa. It is alsolikely that some species will benefit from this emergenthybrid entity while o<strong>the</strong>rs will not. These implications arecurrently under investigation, as revealed by several papersin this volume. In any case, given <strong>the</strong> fact that <strong>the</strong> hybrid canspread by wind-born pollen, it is hard to imagine that <strong>the</strong>future demography <strong>of</strong> S. foliosa in San Francisco Bay willnot be influenced by S. alterniflora to some degree. The factthat hybrids appear to be more successful than pure S.alterniflora suggests that S. foliosa is making an importantcontribution to <strong>the</strong> adaptive success <strong>of</strong> this entity. In thatsense, S. foliosa is not so much going “extinct” as it is beingtransformed into a more fertile and ecologically successfulnew organism that combines traits <strong>of</strong> both species, analternative condition that Arnold (1997) describes in a morepositive light as “genetic enrichment”.In a review <strong>of</strong> <strong>the</strong> role <strong>of</strong> natural hybridization andevolution, Arnold (1997) points out that <strong>the</strong>re are both socioculturaland scientific underpinnings to <strong>the</strong> view that allhybridization is “completely maladaptive”. Yet, we arediscovering that hybridization is an important evolutionarymechanism. Stebbins (1950, 1957) argued that a high degree<strong>of</strong> genetic variability is required for rapid rates <strong>of</strong> adaptationand speciation. His idea was that genetic recombination fromhybridization between differently adapted species, ra<strong>the</strong>rthan mutation, is <strong>the</strong> most likely source <strong>of</strong> such variation.The challenge for hybrid speciation under natural conditionsis to (1) produce a fit recombinant and (2) keep thisgenotype intact in <strong>the</strong> face <strong>of</strong> potential genetic swamping by<strong>the</strong> parents (Arnold 1996, Rieseberg 1997, Turelli et al.2001). The hybridization event between S. foliosa and S.alterniflora has been mediated by human causes ra<strong>the</strong>r thanshifting climate or some o<strong>the</strong>r “natural” event. As a result,<strong>the</strong> more robust but perhaps less locally adapted S.alterniflora does not exist in high enough numbers to swampout <strong>the</strong> hybrid. On <strong>the</strong> o<strong>the</strong>r hand, <strong>the</strong> far more abundant andlocally adaptive S. foliosa appears to lack <strong>the</strong> pollen fertilityto swamp out <strong>the</strong> invading hybrids (Antilla et al. 1998).What seems to be happening is that S. alterniflora genes areintrogressing into S. foliosa populations faster than pure S.alterniflora individuals are colonizing o<strong>the</strong>r parts <strong>of</strong> <strong>the</strong> bay(Antilla et al. 2000). In that sense, over many generations, ifpure S. alterniflora can be constrained or even eradicated, itis more likely that S. foliosa will assimilate S. alternifloragenes into its populations than <strong>the</strong> contrary (i.e. S.alterniflora genes are introgressing into pure S. foliosapopulations). Ultimately, fertile recombinant genotypesappear to be poised to generate a new evolutionary unitwithin San Francisco Bay whose ultimate fate and impact on<strong>the</strong> ecology <strong>of</strong> <strong>the</strong> wetlands <strong>of</strong> this estuary, and elsewhere,are both unknown and potentially beyond our control.While it is still possible, one question should beaddressed that could have legal implications for <strong>the</strong> potentialmanagement <strong>of</strong> this situation. That is, is <strong>the</strong>re any evidencethat <strong>the</strong> nor<strong>the</strong>rn S. foliosa populations are geneticallydistinct from sou<strong>the</strong>rn populations <strong>of</strong> this species? If it couldbe established that nor<strong>the</strong>rn populations represent a distinctpopulation segment, <strong>the</strong>n one could make <strong>the</strong> case that SanFrancisco Bay area S. foliosa should be protected under <strong>the</strong>Endangered Species Act. How that might be done is wellbeyond <strong>the</strong> scope <strong>of</strong> this paper, but it might help focus astrategy <strong>of</strong> containment and conservation <strong>of</strong> S. foliosa that-6-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina Biologywould provide additional leverage to help protect <strong>the</strong> areas<strong>of</strong> San Francisco Bay salt marshes that are so far notimpacted by <strong>the</strong> Spartina hybrid.In <strong>the</strong> long term, it certainly is possible that this hybridevolutionary unit will spread beyond San Francisco Bay. Infact, in 2001, a hybrid individual was found in BolinasLagoon (35 kilometers away), and in 2002, five hybridswere found in Drake’s Estero, some 45 kilometers from <strong>the</strong>nearest San Francisco Bay population <strong>of</strong> hybrid entities(Zaremba and McGowan 2004). Hybrids within <strong>the</strong> bay haveincreased some 317%, particularly in <strong>the</strong> Central Bay(Zaremba and McGowan 2004). Given that <strong>the</strong> Bay is amagnet for migratory shorebirds and waterfowl, and thatmany species are flying south during fall migration, it is hardto imagine that <strong>the</strong> hybrid won’t eventually colonize <strong>the</strong>sou<strong>the</strong>rn center <strong>of</strong> distribution for S. foliosa. On <strong>the</strong> o<strong>the</strong>rhand, even given dispersal to sou<strong>the</strong>rn California, it ispossible that <strong>the</strong> hybrid will not fare as well as S. foliosa in<strong>the</strong> hotter and drier sou<strong>the</strong>rn habitats. None<strong>the</strong>less, it isadvisable for sou<strong>the</strong>rn California and Baja Californiawetland managers to remain vigilant to this possibility and toact quickly to eliminate hybrids, as was <strong>the</strong> case in BolinasLagoon and Drake’s Estero.As known by wetland ecologists for decades, S. foliosais a vital force in shaping salt marsh communities in SanFrancisco Bay, sou<strong>the</strong>rn California, and Baja California. Theintroduction <strong>of</strong> S. alterniflora into San Francisco Bay hasmost likely begun a new chapter in <strong>the</strong> evolutionary fate <strong>of</strong>S. foliosa, and in so doing, is reorganizing its relationship toa host <strong>of</strong> o<strong>the</strong>r tidal wetland species in this region. We needto forge a realistic strategy to deal with this unexpecteddevelopment. Massive eradication efforts could have <strong>the</strong>irown unexpected consequences, especially if directedtowards <strong>the</strong> hybrid. From <strong>the</strong> perspective <strong>of</strong> S. foliosa, <strong>the</strong>fact that its genome is being transformed into a new hybridentity is not, in my view, extinction per se, although thisperspective can certainly be argued and it is not aninsignificant concern. The more compelling issue, however,is how <strong>the</strong> hybrid will affect future salt marsh and mud flathabitats. Given that large-scale wetland restoration projectsare planned for <strong>the</strong> South Bay and elsewhere, it might beworth taking a pluralistic approach to managing this crisis.Design restoration activities to test different approaches,from aggressive prevention to passive observation. Give<strong>the</strong>se trials enough time to evaluate <strong>the</strong> ecological impacts <strong>of</strong>wetland restoration projects that are influenced by hybrids.To <strong>the</strong> degree possible, as in <strong>the</strong> eradication effort at Bolinasand Drake’s Estero, aggressively contain hybrid colonistswherever <strong>the</strong>y can be caught in time. Like it or not, humansare part <strong>of</strong> nature, and nature is full <strong>of</strong> surprises. In <strong>the</strong> longterm, perhaps <strong>the</strong> best we can do is manage this complexitywith respect, active curiosity (including variousexperimental approaches), and perhaps, if not fatalism, atleast a degree <strong>of</strong> tolerance and humble appreciation for <strong>the</strong>evolutionary creativity inherent in this remarkable situation.ACKNOWLEDGMENTSI thank Peggy Ol<strong>of</strong>son for inviting me to contribute thispaper to <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on InvasiveSpartina. Constructive comments on this manuscript havebeen contributed by Drew Talley and Tom Parker. MarkBertness provided helpful feedback on <strong>the</strong> topic <strong>of</strong>foundation species.REFERENCESAntilla, C.K., C.C. Daehler, N.E. Rank, D.R. Strong. 1998. Greatermale fitness <strong>of</strong> a rare invader (Spartina alterniflora, Poaceae)threatens a common native (Spartina foliosa) with hybridization.Amer. J. Bot. 85: 1597-1601.Antilla, C.K., R.A. King, C. Ferris, D.R. Ayres, D.R. Strong. 2000.Reciprocal hybrid formation <strong>of</strong> Spartina in San Francisco Bay.Mol. Ecol. 9:765-770.Arnold, M.L. 1996. Natural Hybridization and Introgression.Princeton Univ. Press, Princeton, NJ.Arnold, M.L. 1997. Natural Hybridization and Evolution. OxfordUniversity Press, NY.Ayres, D.R., D.R. Strong, P. Baye. 2003b. Spartina foliosa(Poaceae) – a common species on <strong>the</strong> road to rarity? Madrono50: 209-213.Barbour, M.G. 1970. The flora and plant communities <strong>of</strong> BodegaHead, California. Madrono 20:289-313.Baumel, A., M.L. Ainouche, R.J. Bayer, A.K. Ainouche, M.T.Misset.. 2002. Molecular phylogeny <strong>of</strong> hybridizing species from<strong>the</strong> genus Spartina Schreb. (Poaceae). Molecular Phylogeneticsand Evolution 22: 303-314.Baye, P. 2004. DRAFT. A review and assessment <strong>of</strong> potential longtermecological consequences <strong>of</strong> <strong>the</strong> introduced cordgrassSpartina alterniflora in <strong>the</strong> San Francisco Estuary. A technicalreport prepared for <strong>the</strong> San Francisco Estuary Invasive SpartinaProject, Berkeley, CA.Boyer, K. and J.B. Zedler. 1996. Damage to cordgrass by scaleinsects in a constructed salt marsh: effects. Estuaries 19:1-12.Bruno, J.F. and M.D. Bertness. 2001. Habitat modification andfacilitation in benthic marine communities. In: M.D. Bertness etal., eds. Marine Community Ecology. Sinauer Press, pp. 201-218.Bruno, J.F., J.J. Stachowicz, M.D. Bertness. 2003. Inclusion <strong>of</strong>facilitation into ecological <strong>the</strong>ory. Trends in Evolution andEcology 3: 119-125.Calsbeek R.; J.N. Thompson; J.E. Richardson. 2003. Patterns <strong>of</strong>molecular evolution and diversification in a biodiversity hotspot:<strong>the</strong> California Floristic Province. Mol. Ecol. 12: 1021-1029.Daehler, C.C. and D.R. Strong. 1996. Status, prediction, andprevention <strong>of</strong> introduced cordgrasses (Spartina spp.) invasions inPacific estuaries, USA. Biol. Cons. 78:51-58.Dayton, P.K. 1972. Toward an understanding <strong>of</strong> communityresilience and <strong>the</strong> potential effects <strong>of</strong> enrichments to <strong>the</strong> benthos<strong>of</strong> McMurdo Soiund Antarctica. In Parker, B.C. eds,<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> Colloquium on Conservation Problems inAntarctica. Allen Press, pp. 81-96.Dawson, E.Y. and M.S. Foster 1982. Seashore Plants <strong>of</strong> California.U.C. Press, Berkeley.Howell, J.T. 1949. Marin Flora; manual <strong>of</strong> <strong>the</strong> flowering plantsand ferns <strong>of</strong> Marin County, California. U.C. Press, Berkeley.Howell, J.T. 1970. Marin Flora; manual <strong>of</strong> <strong>the</strong> flowering plantsand ferns <strong>of</strong> Marin County, California (Including a Supplement).U.C. Press, Berkeley.-7-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaMacDonald, K.B. and M.G. Barbour. 1974. Beach and salt marshvegetation <strong>of</strong> <strong>the</strong> North Americcan Pacific Coast. In Reimold,R.J. and W.H. Queen (Eds) Ecology <strong>of</strong> Halophytes. AcademicPress, N.Y., pp.175-224.Myers, N., R.A. Mittermeier, C.G. Mittermeier, G.A.B. da Fonseca,J. Kents. 2000. Biodiversity hotpots for conservation priorities.Nature 403: 853-858.Page, G.W., and J.G. Evens. 1987. The sizes <strong>of</strong> clapper railpopulations at Corte Madera Ecological Reserve, Muzzi Marsh,San Clemente Creek, and Triangle Marsh. Report to MarinAudubon Society from Point Reyes Bird Observatory, May 1987.PRBO Contribution Number 367.Raven, P.H and D.I. Axelrod. 1978. Origin and Relationships <strong>of</strong><strong>the</strong> California Flora. U.C. Press, Berkeley.Reiseberg, L.H. 1997. Hybrid origins <strong>of</strong> plant species. Ann. Rev.Ecol. Syst. 28: 359-389.Schoenherr, A. 1995. A Natural History <strong>of</strong> California. U.C. Press,Berkeley, CA.Spicher, D. and M. Josselyn. 1985. Spartina (Gramineae) inNor<strong>the</strong>rn California: distribution and taxonomic notes. Madrono32: 158-167.Stebbins, G.L. 1950. Variation and evolution in plants. ColumbiaUniv Press, NY.Stebbins, G.L. 1957. The role <strong>of</strong> hybridization in evolution. Am.Philos. Soc. 103: 231-251.Trnka, S. and J.B. Zedler. 2000. Site conditions, not parentalphenotype, determine <strong>the</strong> height <strong>of</strong> Spartina foliosa. Estuaries23: 572-582.Turelli, M., N.H. Barton, J.A. Coyne. 2001. Theory and speciation.Trends Ecol. Evol. 16: 330-343.Vivian-Smith, G. and E.W. Stiles. 1994. Dispersal <strong>of</strong> salt marshseeds on <strong>the</strong> feet and featyers <strong>of</strong> waterfowl. Wetlands 14: 316-319.Wiggins, I. 1980. Flora <strong>of</strong> Baja California. Stanford Univ. Press,Palo Alto, CA.Zaremba, K. and M.F. McGowan. 2004. San Francisco EstuaryInvasive Spartina Project Monitoring Report for 2003. SanFrancisco Estuary Invasive Spartina Project, Berkeley, CA.-8-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyLOCAL AND GEOGRAPHIC VARIATION IN SPARTINA-HERBIVORE INTERACTIONSS.C. PENNINGSDepartment <strong>of</strong> Biology and Biochemistry, University <strong>of</strong> Houston, Houston TX 77204; spennings@uh.eduSpartina alterniflora is consumed by a variety <strong>of</strong> herbivores, and <strong>the</strong> nature <strong>of</strong> <strong>the</strong>se plant-herbivoreinteractions varies on both local and geographic scales. The palatability <strong>of</strong> S. alterniflora toherbivores varies within single marshes as a function <strong>of</strong> elevation. Tall-form plants, which occur atlow elevations close to creek banks, are more palatable to herbivores than are short-form plants,which occur at middle elevations on <strong>the</strong> marsh platform. The proximate causes <strong>of</strong> this localvariation in palatability include variation in leaf nitrogen content and chemical defenses.Differences in <strong>the</strong>se leaf traits are driven by variation in sediment biogeochemistry across <strong>the</strong>elevational gradient. The palatability <strong>of</strong> S. alterniflora to herbivores also varies geographically.High-latitude (New England) plants along <strong>the</strong> Atlantic Coast <strong>of</strong> <strong>the</strong> United States are more palatableto herbivores than are low-latitude (South Atlantic Bight) plants. The proximate causes <strong>of</strong> thislatitudinal variation in palatability include variation in leaf nitrogen content, toughness, andchemical defenses. Differences in palatability and leaf traits persisted over 5 clonal generations in acommon-garden greenhouse environment, and thus are probably genetically determined. A number<strong>of</strong> processes may contribute to driving this variation in leaf traits. The most likely ultimate causesare latitudinal variation in herbivore pressure and latitudinal variation in plant phenology.Understanding spatial variation in interactions between herbivores and S. alterniflora within itsnative range may shed insights into how <strong>the</strong>se interactions develop when S. alterniflora isintroduced to new regions, and may inform biocontrol efforts.Keywords: common-garden experiment, elevation, herbivory, latitude, leaf traits, salt marsh,Spartina alternifloraINTRODUCTIONEarly salt-marsh studies discounted <strong>the</strong> importance <strong>of</strong>herbivory to salt marsh ecology because herbivores did notappear to play a major role in energy flow through <strong>the</strong> marshfood web (Smalley 1960; Teal 1962). More recent work,however, has shown that herbivores can affect <strong>the</strong> biomass,distributions and reproduction <strong>of</strong> a variety <strong>of</strong> salt marshplants (Srivastava and Jefferies 1996; Bortolus and Iribarne1999; Pennings and Bertness 2001; Silliman and Zieman2001; Rand 2003; Silliman and Bortolus 2003; Ho andPennings 2008). Thus, a general understanding <strong>of</strong> <strong>the</strong>ecology <strong>of</strong> salt-marsh plants requires consideration <strong>of</strong> plan<strong>the</strong>rbivoreinteractions.In order to obtain a general understanding <strong>of</strong> plan<strong>the</strong>rbivoreinteractions, it is necessary to consider spatialvariation. Even within <strong>the</strong> limited confines <strong>of</strong> salt marshhabitats, interactions between plants and herbivores are notexactly <strong>the</strong> same everywhere. They vary with marshelevation (Silliman and Bertness 2002; Goranson et al.2004), local plant community composition (Rand 1999,2003, 2004), and latitude (Pennings et al. 2001; Penningsand Silliman 2005; Pennings et al. 2007). Thus, a generalunderstanding <strong>of</strong> plant-herbivore interactions in salt marshesmust incorporate spatial variation on a variety <strong>of</strong> scales.Here, I address variation in interactions betweenherbivores and Spartina alterniflora at local and geographicscales. At <strong>the</strong> local scale I focus on <strong>the</strong> intertidal gradientacross <strong>the</strong> marsh. Along this gradient, soil biogeochemistryvaries markedly, producing strong variation in plantmorphology and palatability to herbivores. At <strong>the</strong>geographic scale I focus on latitudinal variation along <strong>the</strong>Atlantic Coast <strong>of</strong> <strong>the</strong> United States. Latitudinal differencesin palatability <strong>of</strong> salt marsh plants to herbivores are strikingand general across <strong>the</strong> plant and herbivore community.Understanding variation in plant-herbivore interactions atboth <strong>the</strong>se scales provides a framework for syn<strong>the</strong>sizingresults <strong>of</strong> different studies done in different locations, allowstests <strong>of</strong> general ecological <strong>the</strong>ory, and may shed insights intoecological processes obtained when Spartina is introducedinto new geographic regions.LOCAL VARIATIONSpartina alterniflora varies in palatability to herbivoreswithin individual marshes because <strong>of</strong> variation in soilbiogeochemistry that mediates plant traits. Salt marshhabitats are physically stressful for plants. Periodic flooding<strong>of</strong> marsh soils leads to high levels <strong>of</strong> sulfides, low redoxlevels, low levels <strong>of</strong> oxygen, and low bioavailability <strong>of</strong>nitrogen, a suite <strong>of</strong> factors that are inimicable to vigorous-9-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaplant growth (Ponnamperuma 1972; Mendelssohn andMorris 2000). The high salinity <strong>of</strong> salt marsh soils fur<strong>the</strong>rlimits plant growth by reducing soil water potential,damaging cellular processes, and interfering with nitrogenuptake (Drake 1989; Mendelssohn and Morris 2000).Flooding and salinity levels vary across <strong>the</strong> marshlandscape (Pennings and Bertness 2001). Flooding is leastat <strong>the</strong> terrestrial border <strong>of</strong> <strong>the</strong> marsh, which is rarely flooded,and increases at lower marsh elevations; however, stresscaused by flooding may be reduced at low elevationsimmediately adjacent to creekbanks, where <strong>the</strong>re is a highlevel <strong>of</strong> exchange between pore water and <strong>the</strong> water column(Howes et al. 1981; Howes and Goehringer 1994). Salinitylevels are low at <strong>the</strong> terrestrial border, close to those <strong>of</strong> <strong>the</strong>water column at <strong>the</strong> creekbank, and may peak in <strong>the</strong> highmarsh if conditions favor evaporation <strong>of</strong> water andconcentration <strong>of</strong> salts (Pennings and Bertness 1999).Because <strong>of</strong> this variation in soil biogeochemistry acrosselevation, almost all salt marsh plants exhibit strong spatialvariation in height and o<strong>the</strong>r morphological traits (Richardset al. 2005). For S. alterniflora, which occurs at middle tolow marsh elevations, plants at middle elevations are short(to < 50 cm) and plants at low elevations, especially close tocreekbanks, are tall (to > 200 cm). This spatial variation inmorphology is strongly driven by variation in <strong>the</strong> abioticenvironment, although genetic variation among plants alsoaffects morphology (Valiela et al. 1978; Anderson andTreshow 1980; Gallagher et al. 1988; Pr<strong>of</strong>fitt et al. 2003).Given strong variation in plant height and morphologyacross <strong>the</strong> intertidal zone, it is not surprising that palatabilityto herbivores also varies across intertidal gradients. Lowmarshplants are more palatable than mid-marsh plants to avariety <strong>of</strong> herbivores, including geese (Buchsbaum et al.1984), hemiptera (Denno et al. 1980; Denno et al. 1996),grasshoppers (Goranson et al. 2004) and snails (Silliman andBertness 2002).The plant traits mediating this variation in palatabilityhave not been explicitly examined; however, low- and midmarshplants differ in two important ways that are likely toaffect palatability to herbivores. First, low-marsh plantshave a higher nitrogen content than do plants occurring in<strong>the</strong> mid-marsh (Gallagher et al. 1980; Bowdish and Stiling1998). Second, <strong>the</strong> concentration <strong>of</strong> phenolics (a class <strong>of</strong>chemical compounds that <strong>of</strong>ten deter feeding by herbivores)is lower in low-marsh plants than in mid-marsh plants(Buchsbaum et al. 1984). Variation in <strong>the</strong>se two traits likelyexplains much <strong>of</strong> <strong>the</strong> preference <strong>of</strong> herbivores for low-marshversus mid-marsh S. alterniflora.The palatability <strong>of</strong> S. alterniflora to herbivores is alsoaffected by past herbivory (Denno et al. 2000). Feeding byherbivores may reduce plant nutritional content and/orinduce defenses against herbivores. In low-latitude marshes,gastropods that damage S. alterniflora are much moreabundant in <strong>the</strong> mid marsh than <strong>the</strong> low marsh (Silliman andBertness 2002). To <strong>the</strong> extent that this pattern is generalamong herbivores, herbivore feeding damage in <strong>the</strong> middlemarsh zone could reinforce patterns <strong>of</strong> plant palatabilitycreated by biogeochemical differences across elevation.Variation in palatability to herbivores across marshelevation is not unique to S. alterniflora. A number <strong>of</strong> saltmarsh plants vary in palatability to herbivores as a function<strong>of</strong> marsh elevation and/or salinity (Hemminga and vanSoelen 1988; Levine et al. 1998; Moon and Stiling 2000;Goranson et al. 2004). Given that <strong>the</strong> physical stressgradients in salt marshes affect both nitrogen availability andplant size, it is likely that most salt marsh plants vary innitrogen content and toughness across elevation, and itwould be surprising if variation in <strong>the</strong>se factors did notaffect palatability to herbivores. The details <strong>of</strong> how physicalstress mediates plant palatability in salt marshes, however,appear to be species-specific, depending on both <strong>the</strong> plantand <strong>the</strong> herbivore involved, and <strong>the</strong>refore cannot be reducedto a simple generalization that applies across all species(Goranson et al. 2004).LATITUDINAL VARIATIONSpartina alterniflora also varies in palatability toherbivores across latitude. Paired feeding preference assayscomparing S. alterniflora from New England versus <strong>the</strong>South Atlantic Bight found that four species <strong>of</strong> herbivoresstrongly preferred to eat <strong>the</strong> high-latitude plants versus <strong>the</strong>low-latitude plants (Pennings et al. 2001). Results weresignificant in 19 <strong>of</strong> 21 assays, and did not depend on season,year, species <strong>of</strong> herbivore, or geographic origin <strong>of</strong> <strong>the</strong>herbivore. Similar results were obtained for nine o<strong>the</strong>r saltmarsh plant species that were studied, and held true across asuite <strong>of</strong> herbivore taxa, indicating that <strong>the</strong> preference forhigh- versus low-latitude plants was general across <strong>the</strong> entireplant and herbivore community.What plant traits explain <strong>the</strong> preference for high-latitudeplants? A comparison <strong>of</strong> plant traits suggested that highlatitudeS. alterniflora plants had a higher nitrogen contentand were less tough than low-latitude plants (Siska et al.2002). In addition, given a choice between polar extracts <strong>of</strong>high- and low-latitude plants, consumers consistentlypreferred to eat <strong>the</strong> high-latitude extracts, suggesting that<strong>the</strong>re were latitudinal differences in polar chemistry.(Consumers did not show consistent preferences for nonpolarextracts.) Phenolics, which are polar compounds, werehigher in lowversus high-latitude plants. Thus, this studysuggested that all three plant traits that were examined—nitrogen, toughness and secondary chemistry—might makehigh-latitude plants more palatable than low-latitude plants.Differences in palatability <strong>of</strong> S. alterniflora plantsacross latitude appear to be constitutive ra<strong>the</strong>r than solely aplastic response to <strong>the</strong> environment. When S. alternifloraplants were grown in a common-garden greenhouse-10-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina Biologyenvironment over five clonal generations and two growingseasons, latitudinal differences in palatability, toughness andnitrogen content persisted undiminished (Salgado andPennings 2005). (Palatability <strong>of</strong> extracts was not examinedin this study.) This study did not rule out <strong>the</strong> possibility thatS. alterniflora has induced defenses against herbivores, butdid indicate that plastic responses alone, whe<strong>the</strong>r to <strong>the</strong>biotic or abiotic environment, would not be sufficient toexplain <strong>the</strong> latitudinal patterns in palatability. However,constitutive differences in defense are strong, and couldexplain part or all <strong>of</strong> <strong>the</strong> latitudinal gradient in palatability.The conclusion that <strong>the</strong>re is latitudinal variation in geneticresistance to herbivory is consistent with o<strong>the</strong>r studiesshowing constitutive latitudinal variation in S. alternifloratraits (Seneca 1974; Seliskar et al. 2002) and geneticdifferentiation <strong>of</strong> S. alterniflora populations across latitude(O'Brien and Freshwater 1999).The ultimate factors producing <strong>the</strong>se latitudinaldifferences in plant traits are yet to be determined; however,because all ten plant species studied showed similarlatitudinal differences in palatability (Pennings et al. 2001),it is probable that <strong>the</strong> ultimate explanations lie in generalecological and biogeographic processes ra<strong>the</strong>r thanidiosyncratic aspects <strong>of</strong> Spartina biology. Because <strong>the</strong>sestudies were all done at sites exposed to full-strengthseawater, latitudinal differences in soil edaphic conditionswere minor compared to <strong>the</strong> differences that occur acrosselevation within single marshes, and thus probably notimportant. It is possible that latitudinal differences insubstrate type (peat at high latitudes; mineral soils at lowlatitudes) might select for differences in plant palatability insome way. The most likely hypo<strong>the</strong>ses, however, are that<strong>the</strong> differences in plant traits are produced by latitudinaldifferences in herbivore pressure or in plant phenology.Considerable evidence indicates that herbivore pressureon S. alterniflora and o<strong>the</strong>r marsh plants is greater at lowversus high latitudes. In low-latitude marshes, gastropodscan strongly suppress growth <strong>of</strong> S. alterniflora (Silliman andZieman 2001), but gastropods have no effect on S.alterniflora growth in high-latitude marshes (Pennings andSilliman 2005). Omnivorous grasshoppers are larger and aremore likely to eat leaves ra<strong>the</strong>r than seeds or arthropods inlow- versus high-latitude marshes (Pennings and Silliman2005; Wason and Pennings 2008). Grazing damage to S.alterniflora from snails and grasshoppers (but not Prokelisiaspp.) is greater in lowversus high-latitude marshes(Pennings, unpublished data). Thus, it is possible thatgreater herbivore pressure at low versus high latitudes couldselect for increased plant defenses at low latitudes, creating ageographic difference in plant palatability. Alternatively,<strong>the</strong> shorter growing season at high latitudes might select forleaves that are higher in nitrogen but less tough (Reich andOleksyn 2004; Wright et al. 2004). Differences in <strong>the</strong>se leaftraits driven by phenology would likely affect palatability toherbivores even if herbivore pressure did not select for <strong>the</strong>traits.CONCLUSIONSSpartina alterniflora varies in palatability to herbivoresat both local and geographic scales. We have a basicunderstanding <strong>of</strong> how variation in leaf traits correlates withvariation in palatability at both spatial scales, but how muchdifferent leaf traits affect <strong>the</strong> feeding preferences <strong>of</strong> differen<strong>the</strong>rbivores remains to be determined. Similarly, we havesome understanding <strong>of</strong> how variation in soilbiogeochemistry, herbivore pressure and phenology may be<strong>the</strong> ultimate factors mediating local and geographic variationin leaf traits, but much remains to be learned about <strong>the</strong>relative importance <strong>of</strong> <strong>the</strong>se ultimate factors at differentspatial scales.Given that local and geographic patterns in palatability<strong>of</strong> S. alterniflora exist, <strong>the</strong>se patterns are important for atleast three reasons. First, <strong>the</strong>y provide a unifying frameworkto link studies <strong>of</strong> plant-herbivore interactions done indifferent zones <strong>of</strong> different marshes. Although differentresults may be obtained in different studies, much <strong>of</strong> <strong>the</strong>variation among studies may be reconciled by understanding<strong>the</strong> underlying differences in plant palatability at local andgeographic scales. Second, this variation allows tests <strong>of</strong>general <strong>the</strong>ories <strong>of</strong> plant-herbivore interactions, which positthat plant-herbivore interactions will vary predictably withphysical stress (Goranson et al. 2004; Huberty and Denno2004) and latitude (Pennings et al. 2001). <strong>Third</strong>, <strong>the</strong>sepatterns may lend insight into some aspects <strong>of</strong> introductions<strong>of</strong> S. alterniflora into new geographic locations. Forexample, plants introduced from relatively high-latitude sitesare likely to be more vulnerable to herbivores than plantsintroduced from low-latitude sites.ACKNOWLEDGMENTSMy work on local and geographic variation ininteractions between S. alterniflora and its herbivores hasbeen supported by <strong>the</strong> National Geographic Society, <strong>the</strong>National Science Foundation (DEB-0296160 and 0638796,OCE99-82133), and <strong>the</strong> Environmental Institute <strong>of</strong> Houston.This work is part <strong>of</strong> <strong>the</strong> Georgia Coastal Ecosystems LTERprogram.REFERENCESAnderson, C.M. and M. Treshow. 1980. A review <strong>of</strong> environmentaland genetic factors that affect height in Spartina alternifloraLoisel. (salt marsh cordgrass). Estuaries 3:168-176.Bortolus, A. and O. Iribarne. 1999. Effects <strong>of</strong> <strong>the</strong> SW Atlantic burrowingcrab Chasmagnathus granulata on a Spartina salt marsh.Marine Ecology Progress Series 178:79-88.Bowdish, T.I. and P. Stiling. 1998. The influence <strong>of</strong> salt and nitrogenon herbivore abundance: direct and indirect effects. Oecologia113:400-405.-11-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaBuchsbaum, R., I. Valiela, and T. Swain. 1984. The role <strong>of</strong> phenoliccompounds and o<strong>the</strong>r plant constituents in feeding by Canadageese in a coastal marsh. Oecologia 63:343-349.Denno, R.F., M.A. Peterson, C. Gratton, J. Cheng, G.A. Langellotto,A.F. Huberty, and D.L. Finke. 2000. Feeding-inducedchanges in plant quality mediate interspecific competition betweensap-feeding herbivores. Ecology 81:1814-1827.Denno, R.F., M.J. Raupp, D.W. Tallamy, and C.F. Reichelderfer.1980. Migration in heterogeneous environments: differences inhabitat selection between <strong>the</strong> wing forms <strong>of</strong> <strong>the</strong> dimorphicplanthopper, Prokelisia marginata (Homoptera: Delphacidae).Ecology 61:859-867.Denno, R.F., G.K. Roderick, M.A. Peterson, A.F. Huberty, H.G.Dobel, M.D. Eubanks, J.E. Losey, and G.A. Langellotto. 1996.Habitat persistence underlies intraspecific variation in <strong>the</strong> dispersalstrategies <strong>of</strong> planthoppers. Ecological Monographs 66:389-408.Drake, B.G. 1989. Photosyn<strong>the</strong>sis <strong>of</strong> salt marsh species. AquaticBotany 34:167-180.Gallagher, J.L., R.J. Reimold, R.A. Linthurst, and W.J. Pfeiffer.1980. Aerial production, mortality, and mineral accumulationexportdynamics in Spartina alterniflora and Juncus roemerianusplant stands in a Georgia salt marsh. Ecology 61:303-312.Gallagher, J.L., G.F. Somers, D.M. Grant, and D.M. Seliskar. 1988.Persistent differences in two forms <strong>of</strong> Spartina alterniflora: acommon garden experiment. Ecology 69:1005-1008.Goranson, C.E., C.-K. Ho, and S.C. Pennings. 2004. Environmentalgradients and herbivore feeding preferences in coastal saltmarshes. Oecologia 140:591-600.Hemminga, M.A. and J. van Soelen. 1988. Estuarine gradients and<strong>the</strong> growth and development <strong>of</strong> Agapanthia villosoviridescens,(Coleoptera), a stem-borer <strong>of</strong> <strong>the</strong> salt marsh halophyte Astertripolium. Oecologia 77:307-312.Ho, C.-K. and S.C. Pennings. 2008. Consequences <strong>of</strong> omnivory fortrophic interactions on a salt marsh shrub. Ecology 89:6.Howes, B.L. and D.D. Goehringer. 1994. Porewater drainage anddissolved organic carbon and nutrient losses through <strong>the</strong> intertidalcreekbanks <strong>of</strong> a New England salt marsh. Marine EcologyProgress Series 114:289-301.Howes, B.L., R. W. Howarth, J.M. Teal, and I. Valiela. 1981. Oxidation-reductionpotentials in a salt marsh: spatial patterns andinteractions with primary production. Limnology and Oceanography26:350-360.Huberty, A.F. and R.F. Denno. 2004. Plant water stress and itsconsequences for herbivorous insects: a new syn<strong>the</strong>sis. Ecology85:1838-1398.Levine, J.M., S.D. Hacker, C.D.G. Harley, and M.D. Bertness.1998. Nitrogen effects on an interaction chain in a salt marshcommunity. Oecologia 117:266-272.Mendelssohn, I.A. and J.T. Morris. 2000. Eco-physiological controlson <strong>the</strong> productivity <strong>of</strong> Spartina alterniflora Loisel. In:Weinstein, M.P. and D.A. Kreeger, eds. Concepts and Controversiesin Tidal Marsh Ecology. Kluwer Academic Publishers,Dordrecht.Moon, D.C. and P. Stiling. 2000. Relative importance <strong>of</strong> abioticallyinduced direct and indirect effects on a salt-marsh herbivore.Ecology 81:470-481.O'Brien, D.L. and D.W. Freshwater. 1999. Genetic diversity withintall form Spartina alterniflora Loisel. along <strong>the</strong> Atlantic and Gulfcoasts <strong>of</strong> <strong>the</strong> United States. Wetlands 19:352-358.Pennings, S.C. and M.D. Bertness. 1999. Using latitudinal variationto examine effects <strong>of</strong> climate on coastal salt marsh pattern andprocess. Current Topics in Wetland Biogeochemistry 3:100-111.Pennings, S.C., and M.D. Bertness. 2001. Salt marsh communities.Pages 289-316 In: Bertness, M.D., S.D. Gaines, and M.E. Hay,eds. Marine Community Ecology. Sinauer Associates, Sunderland.Pennings, S.C., and B. R. Silliman. 2005. Linking biogeographyand community ecology: latitudinal variation in plant-herbivoreinteraction strength. Ecology 86:2310-2319.Pennings, S.C., E.L. Siska, and M.D. Bertness. 2001. Latitudinaldifferences in plant palatability in Atlantic Coast salt marshes.Ecology 82:1344-1359.Pennings, S.C., M. Zimmer, N. Dias, M. Sprung, N. Davé, C.-K.Ho, A. Kunza, C. McFarlin, M. Mews, A. Pfauder, and C.Salgado. 2007. Latitudinal variation in plant-herbivore interactionsin European salt marshes. Oikos 116:543-549.Ponnamperuma, F.N. 1972. The chemistry <strong>of</strong> submerged soils.Advances in Agronomy 24:29-95.Pr<strong>of</strong>fitt, C.E., S.E. Travis, and K.R. Edwards. 2003. Genotype andelevation influence Spartina alterniflora colonization and growthin a created salt marsh. Ecological Applications 13:180-192.Rand, T.A. 1999. Effects <strong>of</strong> environmental context on <strong>the</strong> susceptibility<strong>of</strong> Atriplex patula to attack by herbivorous beetles.Oecologia 121:39-46.Rand, T.A. 2003. Herbivore-mediated apparent competition betweentwo salt marsh forbs. Ecology 84:1517-1526.Rand, T.A. 2004. Competition, facilitation, and compensation forinsect herbivory in an annual salt marsh forb. Ecology 85:2046-2052.Reich, P.B. and J. Oleksyn. 2004. Global patterns <strong>of</strong> plant leaf Nand P in relation to temperature and latitude. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong>National Academy <strong>of</strong> Science, USA 101:11001-11006.Richards, C.L., S.C. Pennings, and L.A. Donovan. 2005. Habitatrange and phenotypic variation in salt marsh plants. Plant Ecology176:263-273.Salgado, C.S. and S.C. Pennings. 2005. Latitudinal variation inpalatability <strong>of</strong> salt-marsh plants: are differences constitutive?Ecology 86:1571-1579.Seliskar, D.M., J.L. Gallagher, D.M. Burdick, and L.A. Mutz.2002. The regulation <strong>of</strong> ecosystem functions by ecotypic variationin a dominant plant: a Spartina alterniflora salt-marsh casestudy. Journal <strong>of</strong> Ecology 90:1-11.Seneca, E.D. 1974. Germination and seedling response <strong>of</strong> Atlanticand Gulf coasts populations <strong>of</strong> Spartina alterniflora. AmericanJournal <strong>of</strong> Botany 61:947-956.Silliman, B.R. and M.D. Bertness. 2002. A trophic cascade regulatessalt marsh primary production. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> NationalAcademy <strong>of</strong> Science, USA 99:10500-10505.Silliman, B.R. and A. Bortolus. 2003. Underestimation <strong>of</strong> Spartinaproductivity in western Atlantic marshes: marsh invertebrates eatmore than just detritus. Oikos 101:549-554.Silliman, B.R. and J.C. Zieman. 2001. Top-down control <strong>of</strong>Spartina alterniflora production by periwinkle grazing in a Virginiasalt marsh. Ecology 82:2830-2845.Siska, E.L., S.C. Pennings, T.L. Buck, and M.D. Hanisak. 2002.Latitudinal variation in palatability <strong>of</strong> salt-marsh plants: whichtraits are responsible? Ecology 83:3369-3381.Smalley, A.E. 1960. Energy flow <strong>of</strong> a salt marsh grasshopper population.Ecology 41:785-790.Srivastava, D.S. and R.L. Jefferies. 1996. A positive feedback:herbivory, plant growth, salinity, and <strong>the</strong> desertification <strong>of</strong> anArctic salt-marsh. Journal <strong>of</strong> Ecology 84:31-42.Teal, J.M. 1962. Energy flow in <strong>the</strong> salt marsh ecosystem <strong>of</strong> Georgia.Ecology 43:614-624.-12-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyValiela, I., J.M. Teal, and W.G. Deuser. 1978. The nature <strong>of</strong> growthforms in <strong>the</strong> salt marsh grass Spartina alterniflora. AmericanNaturalist 112:461-470.Wason, E.L. and S.C. Pennings. 2008. Grasshopper (Orthoptera:Tettigoniidae) species composition and size across latitude in AtlanticCoast salt marshes. Estuaries and Coasts 31:335-343.Wright, I.J., P.B. Reich, M. Westoby, D.D. Ackerly, Z. Baruch, F.Bongers, J. Cavender-Bares, T. Chapin, J.H.C. Cornelissen, M.Diemer, J. Flexas, E. Garnier, P.K. Groom, J. Gulias, K. Hikosaka,B.B. Lamont, T. Lee, W. Lee, C. Lusk, J.J. Midgley, M.-L.Navas, U. Niinements, J. Oleksyn, N. Osada, H. Poorter, P. Poot,L. Prior, V.I. Pyankov, C. Roument, S.C. Thomas, M.G.Tjoelker, E. Veneklass, and R. Villar. 2004. The worldwide leafeconomics spectrum. Nature 428:821-827.-13-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologySPECIATION,GENETIC AND GENOMIC EVOLUTION IN SPARTINAM.L. AINOUCHE 1 ,A.BAUMEL 2 ,R.BAYER 3 ,K.FUKUNAGA 4 ,T.CARIOU 1 AND M.T. MISSET 11 UMR CNRS 6553 Ecobio. University <strong>of</strong> Rennes 1. Campus de Beaulieu. 35 042 Rennes Cedex (France);Malika.Ainouche@univ-rennes1.fr2 Institut Méditerranéen d'Ecologie et de Paléoécologie, Université Aix-Marseille – Faculté des Sciences de St. Jérôme.Avenue Escadrille Normandie-Niemen Boite 461, F 13397 Marseille cedex 20 (France)3 University <strong>of</strong> Memphis, 201A Life Sciences Building, Memphis, TN 38152 (USA)4 Faculty <strong>of</strong> Life and Environmental Sciences, Prefectural University <strong>of</strong> Hiroshima, 562 Nanatsuka-Cho, Shobara,Hiroshima 727-0023 (Japan)The genus Spartina <strong>of</strong>fers several examples <strong>of</strong> reticulate evolution through interspecifichybridization and polyploidy. These processes appear to have critical impact on adaptation andinvasive abilities. Here we examine how molecular analyses have helped our undertsanding <strong>of</strong> <strong>the</strong>evolutionary patterns <strong>of</strong> Spartina, with particular focus on Spartina anglica, which is a well-knownexample <strong>of</strong> recent and successful polyploid species <strong>of</strong> hybrid origin that has now colonized severalcontinents. Molecular phylogenies have provided new insights on relationships and genomicdivergence among species. S. anglica is characterised by morphological plasticity and largeecological amplitude, contrasting with a weak inter-individual genetic variation in both its native(western Europe) and introduced (e.g., Australia) ranges. However, <strong>the</strong> homeologous sub-genomes<strong>of</strong> S. anglica exhibit consistent epigenetic and expression plasticity, which would represent keyprocesses explaining <strong>the</strong> ecological success <strong>of</strong> this species.INTRODUCTIONHybridization and polyploidy are among <strong>the</strong> mostprominent evolutionary processes involved in diversificationand speciation in plants. This is particularly well illustratedin <strong>the</strong> genus Spartina where reticulate events and genomeduplication (allopolyploidy) have occurred recurrently(Ainouche et al. 2004a). Recent hybridization andpolyploidisation events have resulted in <strong>the</strong> expansion <strong>of</strong>particularly successful genotypes with notorious ecologicalimpacts, and represent excellent opportunities to explore <strong>the</strong>early evolutionary processes that accompany <strong>the</strong>establishment and expansion <strong>of</strong> a new species.In this paper, we will examine how recent molecularanalyses have helped our understanding <strong>of</strong> <strong>the</strong> evolutionarypatterns in Spartina, with particular focus on Spartinaanglica which is a well-known example <strong>of</strong> recent andsuccessful polyploid species <strong>of</strong> hybrid origin that has nowcolonized several continents.HYBRIDIZATION AND POLYPLOIDY AS MAJORPROCESSES IN THE EVOLUTION OF SPARTINAAll Spartina species are polyploids; although anextensive screening <strong>of</strong> <strong>the</strong> chromosome numbers at <strong>the</strong>population level still needs to be performed in variousspecies, no diploid species are known in <strong>the</strong> genus, where<strong>the</strong> main ploidy levels recorded in <strong>the</strong> existing literature aretetraploid (2n = 40), hexaploid (2n = 60, 62) or dodecaploid(2n = 10, 122, 14), with a basic chromosome number x = 10(Marchant 1963, 1968). Spartina is a member <strong>of</strong> <strong>the</strong> tribeChloridoideae <strong>of</strong> <strong>the</strong> Grass family and it is composed <strong>of</strong> 17perennial species that are usually salt tolerant and thuscolonize coastal or inland salt marshes in both <strong>the</strong> Nor<strong>the</strong>rnand Sou<strong>the</strong>rn hemispheres. Most <strong>of</strong> <strong>the</strong> species originatefrom <strong>the</strong> New World (Mobberley 1956). Only four taxa arenative to <strong>the</strong> Old-world: S. maritima, S. x neyrautii, S. xtownsendii and S. anglica, <strong>the</strong> three latter being <strong>of</strong> recent(19 th century) hybrid origin.Using nuclear (ITS and Waxy) and chloroplast (trnTtrnL)DNA sequences, Baumel et al. (2002a) have shownthat <strong>the</strong> genus has evolved through two well-supportedlineages. The first lineage comprises <strong>the</strong> American tetraploidspecies S. patens, S. bakeri, S. cynusoroides, S. gracilis, and<strong>the</strong> endemic S. arundinacea from <strong>the</strong> South-Atlantic andIndian oceans, that appears closely related to <strong>the</strong>sou<strong>the</strong>astern American S. ciliata. The second lineage iscomposed <strong>of</strong> <strong>the</strong> hexaploid species including <strong>the</strong> eastern-American S. alterniflora, that appears weakly divergent fromits sister species S. foliosa from California, <strong>the</strong> AtlanticEuro-African S. maritima that is more differentiated at both<strong>the</strong> molecular and morphological levels. The tetraploid S.argentinensis is basal to this hexaploid lineage (Baumel etal. 2002a).Recent and well-documented hybridization eventsinvolve S. alterniflora that has been introduced in Californiaand in western Europe, and it has, in both cases, hybridizedwith native species (Fig. 1). The patterns and outcomes <strong>of</strong><strong>the</strong>se hybridizations agree with <strong>the</strong> phylogeneticrelationships and <strong>the</strong> molecular divergence found betweenspecies: fertile introgressant hybrids involving <strong>the</strong> sisterparental species S. alterniflora and S. foliosa on one hand,and sterile hybrids (S. x neyrautii and S. x townsendii)-15-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinafollowed by alloploid speciation (S. anglica), involvingrelated, but more divergent parental species (S. alternifloraand S. maritima) on <strong>the</strong> o<strong>the</strong>r hand.In California, S. alterniflora was deliberately introducedin <strong>the</strong> mid-1970s in San Francisco Bay where it now cooccurswith <strong>the</strong> native S. foliosa (Daehler and Strong 1997).D. Strong and his co-workers have extensively studied <strong>the</strong>ecological and evolutionary consequences <strong>of</strong> thisintroduction (Ayres and Strong 2010). Hybridizationbetween <strong>the</strong>se two outcrossing, wind-pollinated speciesoccurs during <strong>the</strong> overlap <strong>of</strong> <strong>the</strong>ir flowering periods and hasbeen shown to occur bi-directionally (Antilla et al. 2000).Recurrent backcrosses have resulted in hybrid swarms thatdisplay most frequently <strong>the</strong> chloroplast haplotype <strong>of</strong> S.foliosa and up to 90% <strong>of</strong> <strong>the</strong> nuclear markers specific to S.alterniflora (Ayres et al. 1999; Antilla et al. 2000). Thesehybrids are rapidly spreading, and <strong>the</strong>y are now consideredas a conservation threat to <strong>the</strong> native S. foliosa populations.Reticulate events recently recorded in California alsoinvolve S. densiflora that has been introduced from Chile:Baumel et al. (2002a) reported an unexpected phylogeneticincongruence between different molecular data sets for <strong>the</strong>phylogenetic placement <strong>of</strong> a S. densiflora sample fromHumboldt Bay (California), and have consequentlyinterpreted this incongruence as possibly resulting fromhybridization with S. foliosa or S. alterniflora. Although <strong>the</strong>history <strong>of</strong> S. densiflora in both its native and introducedrange needs fur<strong>the</strong>r molecular investigation, <strong>the</strong> recentdiscovery <strong>of</strong> hybrids between S. densiflora and S. foliosa in<strong>the</strong> San Francisco Bay (Ayres and Lee 2010) confirms thathybridization may take place even between distantly relatedSpartina species.In western Europe, S. alterniflora has been accidentallyintroduced by shipping ballast at <strong>the</strong> end <strong>of</strong> <strong>the</strong> 19 th century.In Southampton Bay (England) hybridization with S.maritima resulted in a first generation hybrid S. townsendii,that is still growing vegetatively near Hy<strong>the</strong>. Chromosomedoubling gave rise to a new fertile allopolyploid species, S.anglica that has rapidly expanded in range (Gray andRaybould 1997). Spartina anglica and S. x townsendii have35403S. foliosa1S. alternifloraS. maritimaBidirectional introgressionSterile F1 hybrids&Allopolyploid speciationS. foliosaS. alternifloraS. maritimaFig. 1. Phylogenetic relationships <strong>of</strong> three Spartina species involved in recent hybridizations. Branch lengths are proportional to <strong>the</strong> number <strong>of</strong> nucleotidechanges (above <strong>the</strong> branches) recorded from <strong>the</strong> analysis <strong>of</strong> two nuclear (ITS and Waxy) and one chloroplast (trnT-trnL spacer) DNA sequences(data from Baumel et al. 2002a). The map displays <strong>the</strong> natural range <strong>of</strong> <strong>the</strong>se species and <strong>the</strong> two arrows indicate <strong>the</strong> recent introductions <strong>of</strong> Spartinaalterniflora.-16-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina Biology<strong>the</strong> same chloroplast genome as S. alterniflora, which wasconsidered <strong>the</strong> maternal genome donor when it hybridizedwith S. maritima (Ferris et al. 2001).Ano<strong>the</strong>r sterile hybrid has been described in southwestFrance (Basque region) on <strong>the</strong> Spanish border and has beencalled S. x neyrautii (Foucault, 1897). As <strong>the</strong> two hybrids S.x neyrautii and S. x townsendii display markedmorphological differences, <strong>the</strong>y were first considered asreciprocal hybrids (Marchant 1977). However, Baumel et al.(2003) found that S. x neyrautii also has <strong>the</strong> same chloroplastgenome as S. alterniflora, which indicates that hybridizationoccurred in <strong>the</strong> same direction (with S. maritima as <strong>the</strong> maleparent) in England and in southwest France. Both parentalspecies (S. maritima and S. alterniflora) have been foundgenetically depauperate in western Europe (Raybould et al.1991a; Baumel et al. 2003; Yannic et al. 2004), so <strong>the</strong> twohybrids have inherited similar parental genotypes that maybe distinguished by a few molecular markers (Baumel et al.2003). Moreover, Salmon et al. (2005) have found that <strong>the</strong>setwo taxa display very similar genetic and epigeneticdynamics.THE YOUNG INVASIVE S. ANGLICAThe neoallopolyploid S. anglica is a perfect example <strong>of</strong><strong>the</strong> fitness that can be immediately gained followingduplication <strong>of</strong> a hybrid genome: This species displays largerecological amplitude than its parents; as a pioneer species, itis able to colonize a vacant niche fur<strong>the</strong>r down <strong>the</strong> shore,and to tolerate several hours <strong>of</strong> immersion under sea water(Thompson, 1991). Spartina anglica is characterized byhigher physiological tolerance to anoxic soil conditions, andit is able to enhance <strong>the</strong> sediment oxygenation (Lee 2003).This species now has a worldwide distribution (northwesternEurope, China, Australia, North America), as aresult <strong>of</strong> both natural propagation and deliberateintroductions for land reclamation. This propagation isfacilitated by high seed production and impressivevegetative features such as strong rhizomes, that increase <strong>the</strong>sediment accumulation. The rapid spread <strong>of</strong> <strong>the</strong> introducedplants has lead to various attempts to control or eradiate <strong>the</strong>species (Hedge et al. 2010; Hacker and Dethier 2010).However, after a variable period <strong>of</strong> rapid expansion, S.anglica populations seem to experience “dieback” in oldercolonized sites, which is probably caused by age-relateddecline <strong>of</strong> vigor or by competition with o<strong>the</strong>r species afterniche elevation (Gray 2010; An et al. 2010).Spartina anglica benefits from <strong>the</strong> reunion <strong>of</strong> twodivergent (homeologous) genomes from S. alterniflora andS. maritima (Baumel et al. 2002a), <strong>the</strong>refore it ischaracterized by high levels <strong>of</strong> intergenomic heterozygosityat nuclear loci (Guenegou et al. 1988; Raybould et al.1991b). However, contrasting with most o<strong>the</strong>r known cases<strong>of</strong> recent alloploid formation that have been explored withmolecular markers (Soltis et al. 2004; Abbott and Lowe2004), S. anglica has undergone a genetic bottleneck as aresult <strong>of</strong> ei<strong>the</strong>r a unique hybridization event, or multipleevents involving similar parental genotypes (Ainouche et al.2004b). The populations <strong>of</strong> S. anglica display consistentmorphological variation, particularly in differentsuccessional stages where pioneer populations have beenfound to exhibit phenotypes with smaller plants and smallerinflorescence than in mature populations (Thompson et al.1991a). This variation has been found to be caused mainlyby phenotypic plasticity ra<strong>the</strong>r than by geneticdifferentiation (Thompson et al. 1991b, c).The British populations appear genetically depauperatefor allozyme markers (Raybould et al. 1991b). Baumel et al.(2001) have explored <strong>the</strong> populations <strong>of</strong> Spartina anglica inFrance using RAPD (Random Amplification <strong>of</strong> PolymorphicDNA) and ISSR (InterSimple Sequence Repeat) markers,and have found that most populations are composed <strong>of</strong> <strong>the</strong>same major genotype that is identical to <strong>the</strong> first generationhybrid S. x townsendii. Table 1 summarizes molecularinvestigations that have been conducted in populations fromboth <strong>the</strong> native range <strong>of</strong> <strong>the</strong> species (western Europe) andmore recently colonized continents (Australia). Most <strong>of</strong> <strong>the</strong>introduced populations <strong>of</strong> S. anglica in <strong>the</strong> world arebelieved to have originated from Pool Harbor, sou<strong>the</strong>rnEngland (Hubbard 1965). In sou<strong>the</strong>rn England, S. xtowsendii has been sampled in Hy<strong>the</strong> (for comparison withS. anglica genotypes), and <strong>the</strong> populations <strong>of</strong> S. anglica havebeen sampled in Lymington, <strong>the</strong> former site where S. anglicawas recorded, Fawley (near Hy<strong>the</strong>), Keyhaven, Sand Bank,and Studland (Pool Harbour). Samples from Ireland (BullIsland) were also analyzed. Forty-three out <strong>of</strong> 45 S. anglicasamples from England and Ireland have <strong>the</strong> same multilocusRAPD genotype (Table 1); only two individuals exhibitdifferent genotypes: one individual, sampled in Fawleydiffers from <strong>the</strong> major genotype by <strong>the</strong> absence <strong>of</strong> oneRAPD marker, whereas two o<strong>the</strong>r individuals fromLymington are distinguished by <strong>the</strong> loss <strong>of</strong> two DNAfragments. A few ISSR, IRAP and REMAP markers arepolymorphic (Table 1), which allowed Baumel et al. (2002b)to distinguish 13 genotypes that were encountered at lowfrequency in some populations; <strong>the</strong>se genotypes are verysimilar to <strong>the</strong> major one, as most <strong>of</strong> <strong>the</strong>m differ by only one(lost) marker.The French populations have been sampled in variouslocations <strong>of</strong> <strong>the</strong> Atlantic coast (Baumel et al. 2001),including Baie des Veys (Normandy) <strong>the</strong> first colonized sitein France where S. anglica was recorded in 1906. Mostindividuals sampled in France also belong to <strong>the</strong>predominant genotype. One small population from Kerdruc(sou<strong>the</strong>rn Brittany) was found fixed for two mutations, with<strong>the</strong> absence <strong>of</strong> one RAPD (Baumel et al. 2001) and oneISSR marker (this study). The few slightly different o<strong>the</strong>rgenotypes (Table 1) were found to be randomly distributed,and at low frequency, among <strong>the</strong> o<strong>the</strong>r French populations.In Saint-Armel (sou<strong>the</strong>rn Brittany), S. anglica growssympatrically with its parental species S. maritima.-17-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaTable 1: Inter-individual molecular variation in native and introduced Spartina anglica populations investigated using three multilocus methods:RAPD (Random Amplification <strong>of</strong> Polymorphic DNA, 14 primers), ISSR (Inter Simple Sequence Repeats, 4 primers), IRAP (Inter-Retrotransposon Amplified Polymorphism, 7 primer pairs), REMAP (Retrotransposon Microsatellite Amplified Polymorphism, 4 primer pairs).The mean individual number sampled per population = 10. N= number <strong>of</strong> samples analyzed per region; pm = polymorphic markers; m=number<strong>of</strong> markers recorded; mg = number <strong>of</strong> genotypes that differ from <strong>the</strong> “major” genotype identified by Baumel et al. (2001) and identical to S. xtownsendii.RAPD ISSR IRAP/REMAPN pm/ m mg/N N pm/m mg/N N pm/m mg/NEngland &Ireland45 3/146 2/45 20 0/1440 b 3/920/204/40 40 b 10/296 9/40France129 a 6/146 6/12920129 a 2/921/142/201/1295 2/296 2/5Australia 20 0/146 0/20 160 0/21 0/160 - - -a = data from Baumel et al. (2001); b = data from Baumel et al. (2002b).Intermediate, morphologically variable individuals may beencountered (Guenegou and Levasseur, personalcommunication), but no genetic exchange was foundbetween <strong>the</strong> two species according to cytological, allozyme(M.T. Misset and G. Allard, unpublished data) andmolecular (Baumel et al. 2001) data.In Australia, 160 samples were collected in various sitesfrom Victoria (Corner Inlet including <strong>the</strong> mouth <strong>of</strong> <strong>the</strong>Albert River and Anderson Inlet, Gippsland) and Tasmania(Franklin River in Port Sorell Bay, Duck River nearSmithton, Perkins Bay, Tamar River and Little Swanportestuary). In Victoria, various attempts at eradication havebeen performed using <strong>the</strong> herbicide Fusillade; we havesampled non-treated zones and also collected rare survivingplants from treated areas. Morphological variation is alsoencountered in <strong>the</strong> Australian populations where two distinctphenotypes (short plants <strong>of</strong> 10 – 20 cm high on <strong>the</strong> one handand taller plants measuring up to 30 cm on <strong>the</strong> o<strong>the</strong>r hand)were collected in <strong>the</strong> Albert River (Victoria). In LittleSwanport, 20 seedlings have been collected in order toanalyse samples resulting unambiguously from sexualreproduction. We found that all <strong>the</strong> individuals investigated inAustralia and Tasmania display <strong>the</strong> same multilocus RAPD orISSR genotype (Table 1), which is identical to <strong>the</strong> majorgenotype encountered in Europe and to S. x townsendii.It appears that populations <strong>of</strong> S. anglica in both itsnative range and more recently colonized areas are mainlycomposed <strong>of</strong> <strong>the</strong> major genotype that is identical to <strong>the</strong> firstgeneration hybrid S. x townsendii. Some variation may beencountered, however, that result from a few mutationsoccurring in local populations and corresponding in mostcases to fragment loss. Ayres and Strong (2001) also notedfragment loss in S. anglica. They encountered variation inRAPD and ISSR analyses including samples <strong>of</strong> S. anglicafrom England and Australia, and observed variable bandlossat five S. maritima-specific loci. However, Salmon et al.(2005) have reported preferential fragment loss <strong>of</strong> S.alterniflora-specific markers using AFLP (AmplifiedFragment Length Polymorphism) in both S. x townsendii andS. anglica. It should be noted that <strong>the</strong> estimation <strong>of</strong> diversitymay be ei<strong>the</strong>r underestimated or overestimated according toboth sampling (number <strong>of</strong> individuals analysed and number<strong>of</strong> markers examined over <strong>the</strong> genome) and <strong>the</strong> techniqueused: We can see (Table 1) that although globally limited,more variation can be detected from ISSR and IRAP–REMAP markers, that target repetitive (Simple SequenceRepeats and Retrotransposons), potentially more variable,parts <strong>of</strong> <strong>the</strong> genome. Moreover, considering that <strong>the</strong>se PCRbasedmarkers are dominant, variation resulting fromsegregating heterozygous genotypes cannot be ruled out.Ano<strong>the</strong>r potential source <strong>of</strong> variation is <strong>the</strong> structuraldynamic that may affect recently formed allopolyploidgenomes (Wendel 2000), but in <strong>the</strong> case <strong>of</strong> S. anglica, thisdynamic does not seem to affect <strong>the</strong> genotype initiallyformed in England to <strong>the</strong> extent that was reported in <strong>the</strong>literature in o<strong>the</strong>r allopolyploid model systems (Baumel etal. 2002b; Ainouche et al. 2004b).We have recently explored <strong>the</strong> epigenetic alterationsthat may affect <strong>the</strong> allopolyploid genome <strong>of</strong> S. anglica(Salmon et al. 2005). Epigenetic mechanisms causeexpression changes that do not result from <strong>the</strong> modification<strong>of</strong> <strong>the</strong> DNA sequence, and have been shown to result fromvarious, non-exclusive processes including DNA or histonemethylation, histone acetylation, and chromatin compaction-18-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyTable 2: Comparisons <strong>of</strong> cDNA AFLP patterns in S. maritima, S. alternilflora, <strong>the</strong> hybrid S. x townsendii and <strong>the</strong> allopolyploid S. anglia. The followingselective primer combinations have been employed: EcoRI + ACG / MseI + CAA, EcoRI + ACG / MseI + CAC, EcoRI + ACA / MseI +CAA EcoRI + ACA / MseI + CAC EcoRI + AAC / MseI + CAA EcoRI + AAC / MseI + CA, EcoRI + AGA / MseI + CAG, EcoRI + AGA / MseI +CCA, EcoRI + ACC / MseI + CAA.Fragments inheritedin <strong>the</strong> hybridand <strong>the</strong> allopolyploïdLost fragments in<strong>the</strong> hybrid andpresent in <strong>the</strong>allopolyploidLost fragments in<strong>the</strong> allopolyploidonlyLost fragments in<strong>the</strong> hybrid and <strong>the</strong>allopolyploidMonomorphic fragments(173)127 30 14 2Polymorphicfragments(61)S. maritimaS. alteniflora28 2 13 410 0 4 0TOTAL 165 (70.5%) 32 (13.7%) 31 (13.2%) 6 (2.5%)(Liu and Wendel 2003). Epigenetic changes have been foundassociated with phenotypic instability in experimentally resyn<strong>the</strong>sizedallopolyploids (Comai et al. 2000; Kashkush etal. 2002). Using Methylation Sensitive AFLP (MSAP),Salmon et al. (2005) have compared <strong>the</strong> methylation patterns<strong>of</strong> <strong>the</strong> parental genomes <strong>of</strong> S. maritima and S. alterniflora, tothose <strong>of</strong> <strong>the</strong> first generation hybrid S. x townsendii and <strong>the</strong>resulting allopolyploid S. anglica. Thirty percent <strong>of</strong> <strong>the</strong>parental methylation patterns were found altered in <strong>the</strong>hybrid and <strong>the</strong> allopolyploid, indicating that hybridization,ra<strong>the</strong>r than genome doubling have triggered most <strong>of</strong> <strong>the</strong>methylation changes <strong>of</strong> S. anglica. This high level <strong>of</strong>epigenetic changes might explain <strong>the</strong> morphologicalplasticity <strong>of</strong> S. anglica and calls for fur<strong>the</strong>r investigations <strong>of</strong><strong>the</strong> potential regulation <strong>of</strong> duplicate gene expression.In this perspective, transcriptomic changes in <strong>the</strong> hybridand <strong>the</strong> allopolyploid have been first analyzed using cDNAAFLP in order to examine whe<strong>the</strong>r <strong>the</strong> parental genomeshave different expression patterns in S. x townsendii and S.anglica. RNA was extracted from young leaves collected inindividual plants cultivated in <strong>the</strong> greenhouse, in S. maritima(collected in Guerande and Saint Armel, Brittany France), S.alterniflora (collected in Landerneau, Brittany, France), S. xtowsendii (Hy<strong>the</strong>, England) and S. anglica (collected inSaint Lunaire and Saint Armel, Brittany, France). Interindividualor inter-population variation <strong>of</strong> cDNA AFLPpatterns was checked for each species and only stable,species-specific cDNA fragments were taken into accountfor comparisons. Of <strong>the</strong> over 234 unambiguous fragmentsscored, 173 were found monomorphic and 61 werepolymorphic (i.e., present or absent) between <strong>the</strong> parentalspecies. As AFLP markers are dominant, polymorphicfragments are expected to be present in <strong>the</strong> hybrid and <strong>the</strong>allopolyploid if expression <strong>of</strong> genome additivity isrespected. Loss <strong>of</strong> parental fragment is interpreted as a result<strong>of</strong> gene silencing, and appearance <strong>of</strong> fragments that areabsent in both parental species is interpreted as novel geneexpression. This genome-wide screening <strong>of</strong> duplicate geneexpression has been previously successfully employed to testgenome expression plasticity in various allopolyploids(Comai et al. 2000; Lee and Chen 2001; Kashkush et al.2002; Soltis et al. 2004; Adams et al. 2005). About seventypercent <strong>of</strong> <strong>the</strong> parental fragments are inherited by both <strong>the</strong>hybrid and <strong>the</strong> allopolyploid (Table 2). Among <strong>the</strong> 61polymorphic fragments, 38 (28 from S. maritima and 10from S. alterniflora) were found to be additive in <strong>the</strong> hybridand <strong>the</strong> allopolyploid. Seventeen fragments (13 from S.maritima and 4 from S. alterniflora) are lost in S. anglica.All <strong>the</strong> fragments present in <strong>the</strong> hybrid and <strong>the</strong> allopolyploidare already present in one (or both) <strong>the</strong> parental species.When considering both monomorphic and polymorphicfragments between <strong>the</strong> parental species, <strong>the</strong> allopolyploidhas lost 37 parental fragments <strong>of</strong> which 31 were still presentin <strong>the</strong> F1 hybrid S. x townsendii and 6 were lost in <strong>the</strong> hybridand <strong>the</strong> allopolyploid, which indicates an overall silencing inS. anglica <strong>of</strong> 15.8 % <strong>of</strong> <strong>the</strong> examined loci, that occurredmostly following genome duplication in <strong>the</strong> polyploid.Although more loci have to be screened in Spartina to verify<strong>the</strong> extent <strong>of</strong> <strong>the</strong> changes overall <strong>the</strong> genome, it appears fromthis first screening that 34.4 % <strong>of</strong> <strong>the</strong> loci are silenced in S.anglica, when considering only <strong>the</strong> parental fragments thatare polymorphic. This represents a consistent amount <strong>of</strong>changes compared to those recorded in o<strong>the</strong>r polyploidsusing similar methods. Kashkush et al. (2002) have reportedabout 2% alteration <strong>of</strong> parental transcripts in experimentallyre-syn<strong>the</strong>sized allopolyploid wheat. Expression alterationhas also been found in allopolyploid Arabidopsis (Comai etal. 2000; Lee and Chen 2001). Recently, Soltis et al. (2004)found about 5% cDNA parental fragment loss and 4 % novelexpression in <strong>the</strong> allotetraploid Tragopogon mirus andTragopogon miscellus. Adams et al. (2003) showeddifferential expression patterns between homeologous genes-19-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinain different tissues <strong>of</strong> allotetraploid cotton (Gossypium),suggesting that sub-functionalization was consequent toallopolyploid formation. Over 2000 transcripts werescreened by cDNA AFLP (Adams et al. 2004) and about 5%<strong>of</strong> <strong>the</strong> duplicated genes were inferred to have been silencedor down-regulated in <strong>the</strong> experimentally re-syn<strong>the</strong>sizedallotetraploid Gossypium.SUMMARY AND CONCLUSIONSAllopolyploid species represent two important outcomes<strong>of</strong> hybridization and genome duplication that are critical for<strong>the</strong>ir evolutionary success: Hybridization leads to <strong>the</strong> merger<strong>of</strong> two more or less differentiated genomes that havepreviously evolved independently and polyploidizationentails an immediate duplication and functional redundancyat all loci. These events entail various molecular interactionsand adjustments that have received much attention in <strong>the</strong>recent literature, as <strong>the</strong>y are now considered as <strong>the</strong> keyprocesses that explain <strong>the</strong> evolutionary success <strong>of</strong> polyploidyin Eucaryotes (reviewed in Liu and Wendel 2003; Osborn etal. 2003). Heterosis and dosage effect in duplicated genomesare likely to increase <strong>the</strong> metabolic plasticity <strong>of</strong> <strong>the</strong>duplicated genes, <strong>the</strong>reby affecting <strong>the</strong> fitness <strong>of</strong> <strong>the</strong> newlyformed species (Riddle and Birchler 2003).The particularly successful Spartina anglica ischaracterised by morphological plasticity and largeecological amplitude, contrasting with a weak interindividualgenetic variation, but consistent epigenetic andexpression plasticity <strong>of</strong> <strong>the</strong> duplicated homeologousgenomes. Fur<strong>the</strong>r investigations and identification <strong>of</strong> <strong>the</strong>sequences that are affected by epigenetic regulation andexpression changes should deepen our understanding <strong>of</strong> <strong>the</strong>ecological success <strong>of</strong> this young species. Spartina <strong>of</strong>fersparticular opportunities to learn about <strong>the</strong> evolutionary andecological consequences <strong>of</strong> polyploid speciation andreticulate evolution at both <strong>the</strong> short term (in nascent speciessuch as S. anglica) and long term (in <strong>the</strong> parental, olderpolyploid species) <strong>of</strong> <strong>the</strong> evolutionary time scale.ACKNOWLEDGMENTSThis work is supported by CNRS funds (UMR CNRS6553 Ecobio), PICS 932 CNRS – CSIRO Canberra, NSF –CNRS funds (project 12 978), and by a postdoctoral grantfrom <strong>the</strong> French Ministry <strong>of</strong> Research to K. Fukunaga. Weare particularly grateful to Les Leunig and Paul Hedge for<strong>the</strong>ir assistance in sampling Australian populations <strong>of</strong> S.anglica in Victoria and Tasmania, respectively, to MikeMcCorry, for sending samples from Ireland, to TravisColumbus and Debra Ayres for sending various AmericanSpartina samples. Jonathan Wendel and Keith Adams (IowaState University, Ames) are thanked for helpful assistance in<strong>the</strong> cDNA AFLP approach and stimulating discussions onpolyploidy. Neil Bagnall (CSIRO Canberra) and SylvieBaloche (CNRS Rennes) are thanked for technical help.REFERENCESAbbott, R.J. and A.J. Lowe. 2004. Origins, establishment andevolution <strong>of</strong> two new polyploid species <strong>of</strong> Senecio in <strong>the</strong> BritishIsles. Biological Journal <strong>of</strong> <strong>the</strong> Linnean Society 82: 467-474.Adams, K.L., R. Cronn, R. Percifield, and J.F. Wendel. 2003.Genes duplicated by polyploidy show unequal contributions to<strong>the</strong> transcriptome and organ-specific reciprocal silencing.<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> National Academy <strong>of</strong> Sciences <strong>of</strong> <strong>the</strong> USA100, 4649–4654.Adams, K.L., R. Percifield, and J.F. Wendel. 2004. Organ-specificsilencing <strong>of</strong> duplicated genes in a newly syn<strong>the</strong>sized cottonallotetraploid. Genetics 168: 2217-2226.Ainouche, M. L., A. Baumel, A. Salmon and G. Yannic. 2004a.Hybridization, polyploidy and speciation in Spartina (Poaceae).New Phytologist, 161: 165-172.Ainouche, M.L., A. Baumel, and A. Salmon. 2004b. Spartinaanglica C.E. Hubbard: a natural model system for analysingearly evolutionary changes that affect allopolyploid genomes. In:Biological relevance <strong>of</strong> polyploidy: ecology to genomics (A.R.Leitch, D.E. Soltis, P.S. Soltis, I.J. Leitch and J.C. Pires Eds).Bioogical Journal <strong>of</strong> <strong>the</strong> Linnean Society. 82: 475-484.An, S., S. Zhou, Z. Wang, Z. Deng, Y. Zhi, and L. Chen. 2010.Spartina in China: Introduction, history, current status and recentresearch. In: Ayres, D.R., D.W. Kerr, S.D. Ericson and P.R.Ol<strong>of</strong>son, eds. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong>on Invasive Spartina. 2004 Nov 8-10, San Francisco, CA, USA.San Francisco Estuary Invasive Spartina Project <strong>of</strong> <strong>the</strong> CaliforniaState Coastal Conservancy: Oakland, CA. (this volume).Anttila, C.K., R.A. King, C. Ferris, D.R. Ayres, and D.R. Strong.2000. Reciprocal hybrid formation <strong>of</strong> Spartina in San FranciscoBay. Molecular Ecology 9: 765-770.Ayres, D.R., D. Garcia-Rossi, H.G. Davis, and D.R. Strong. 1999.Extent and degree <strong>of</strong> hybridization between exotic (Spartinaalterniflora) and native (S. foliosa) cordgrass (Poaceae) inCalifornia, USA determined by randomly amplified polymorphicDNA (RAPDs). Molecular Ecology 8: 1179-1186.Ayres, D.R. and D.R. Strong. 2001. Origin and genetic diversity <strong>of</strong>Spartina anglica (Poaceae) using nuclear DNA markers.American Journal <strong>of</strong> Botany 88: 1863-1867.Ayres, D. and D. Strong. 2005. Evolution <strong>of</strong> Invasive Spartinahybrids in San Francisco Bay. In: Ayres, D.R., D.W. Kerr, S.D.Ericson and P.R. Ol<strong>of</strong>son, eds. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong><strong>International</strong> <strong>Conference</strong> on Invasive Spartina. 2004 Nov 8-10,San Francisco, CA, USA. San Francisco Estuary InvasiveSpartina Project <strong>of</strong> <strong>the</strong> California State Coastal Conservancy:Oakland, CA. (this volume).Ayres, D.R. and A.K.F. Lee. 2005. Spartina densiflora x foliosahybrids found in San Francisco Bay. In: Ayres, D.R., D.W. Kerr,S.D. Ericson and P.R. Ol<strong>of</strong>son, eds. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong><strong>International</strong> <strong>Conference</strong> on Invasive Spartina. 2004 Nov 8-10,San Francisco, CA, USA. San Francisco Estuary InvasiveSpartina Project <strong>of</strong> <strong>the</strong> California State Coastal Conservancy:Oakland, CA. (this volume).Baumel, A., M.L. Ainouche, and J.E. Levasseur. 2001. Molecularinvestigations in populations <strong>of</strong> Spartina anglica C.E. Hubbard(Poaceae) invading coastal Brittany (France). Molecular Ecology10: 1689-1701.Baumel, A, M.L. Ainouche, R.J. Bayer, A.K. Ainouche, and M.T.Misset. 2002a. Molecular phylogeny <strong>of</strong> hybridizing species from<strong>the</strong> genus Spartina Schreb. (Poaceae). Molecular Phylogeneticsand Evolution 22: 303-314.Baumel, A, M.L. Ainouche, R. Kalendar, and A.H. Schulman.2002b. Retrotransposons and genomic stability in populations <strong>of</strong><strong>the</strong> young allopolyploid species Spartina anglica C.E. Hubbard(Poaceae). Molecular Biology and Evolution 19: 1218-1227.-20-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyBaumel, A, M.L. Ainouche, M.T. Misset, J.P. Gourret, and R.J.Bayer. 2003. Genetic evidence for hybridization between <strong>the</strong>native Spartina maritima and <strong>the</strong> introduced Spartina alterniflora(Poaceae) in South-West France: Spartina x neyrautii re-examined.Plant Systematics and Evolution 237: 87-97.Comai, L, A.P. Tyagi, K. Winter, R. Holmes-Davis, S. Reynolds,Y. Stevens, and B. Byers. 2000. Phenotypic instability and rapidgene silencing in newly formed Arabidopsis allotetraploids.Plant Cell 12 : 1551-1567.Daehler, C.C. and D.R. Strong. 1997. Hybridization betweenintroduced smooth cordgrass (Spartina alterniflora; Poaceae)and native California cordgrass (S. foliosa) in San Francisco Bay,California, USA. American Journal <strong>of</strong> Botany 81: 307-313.Gray, A.J. and A.F. Raybould. 1997. The history and evolution <strong>of</strong>S. anglica in <strong>the</strong> British Isles. In: Patten, K., ed. <strong>Proceedings</strong> <strong>of</strong><strong>the</strong> Second <strong>International</strong> Spartina <strong>Conference</strong>. Washington StateUniversity, Cooperative extension, 39-45.Gray, A. 2005. Will Spartina anglica invade northwards withchanging climate? In: Ayres, D.R., D.W. Kerr, S.D. Ericson andP.R. Ol<strong>of</strong>son, eds. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong><strong>Conference</strong> on Invasive Spartina. 2004 Nov 8-10, San Francisco,CA, USA. San Francisco Estuary Invasive Spartina Project <strong>of</strong> <strong>the</strong>California State Coastal Conservancy: Oakland, CA. (thisvolume).Hubbard, J.C.E. 1965. Spartina marshes in Sou<strong>the</strong>rn England. Patterns<strong>of</strong> invasions in Poole Harbour. Journal <strong>of</strong> Ecology 53: 799-813.Ferris, C, R.A. King, and A.J. Gray. 1997. Molecular evidence for<strong>the</strong> maternal parentage in <strong>the</strong> hybrid origin <strong>of</strong> Spartina anglicaC.E. Hubbard. Molecular Ecology 6: 185-187.Foucaud, 1897. Un Spartina inédit. Annales de la Société deSciences Naturelles Charente Inférieure 32: 220-222.Guénégou, M.C., J. Citharel, and J.E. Levasseur. 1988. The hybridstatus <strong>of</strong> Spartina anglica (Poaceae). Enzymatic analysis <strong>of</strong> <strong>the</strong>species and <strong>the</strong> presumed parents. Canadian Journal <strong>of</strong> Botany66: 1830-1833.Hacker, S.D. and M.N. Dethier. 2010. Where do we go from here?Alternative control and restoration trajectories for a marine grass(Spartina anglica) invader in different habitat types. In: Ayres,D.R., D.W. Kerr, S.D. Ericson and P.R. Ol<strong>of</strong>son, eds.<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on InvasiveSpartina. 2004 Nov 8-10, San Francisco, CA, USA. SanFrancisco Estuary Invasive Spartina Project <strong>of</strong> <strong>the</strong> CaliforniaState Coastal Conservancy: Oakland, CA. (this volume).Hedge, P., C. Shepherd, and K. Dyke. 2010. Implementing <strong>the</strong>strategy for <strong>the</strong> management <strong>of</strong> rice grass, Spartina anglica inTasmania, Australia. In: Ayres, D.R., D.W. Kerr, S.D. Ericsonand P.R. Ol<strong>of</strong>son, eds. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong><strong>Conference</strong> on Invasive Spartina. 2004 Nov 8-10, San Francisco,CA, USA. San Francisco Estuary Invasive Spartina Project <strong>of</strong> <strong>the</strong>California State Coastal Conservancy: Oakland, CA. (thisvolume).Kashkush, K., M. Feldman, A. Levy. 2002. Gene loss, silencingand activation in newly syn<strong>the</strong>sized wheat allotetraploid.Genetics 160: 1651-1659.Lee, R.W. 2003. Physiological adaptations <strong>of</strong> <strong>the</strong> invasivecordgrass Spartina anglica to reducing sediments: rhizomemetabolic gas fluxes and enhanced O 2 and H 2 S transport. MarineBiology 143: 9-15.Lee, H.S. and Z.J. Chen. 2001. Protein-coding genes areepigenetically regulated in Arabidopsis polyploids. <strong>Proceedings</strong><strong>of</strong> <strong>the</strong> National Academy <strong>of</strong> Sciences <strong>of</strong> <strong>the</strong> USA 98, 6753–6758.Liu, B. and J.F. Wendel. 2003. Epignetic phenomena and <strong>the</strong>evolution <strong>of</strong> plant allopolyploids. Molecular Phylogenetics andEvolution 29: 365-379.Marchant, C.J. 1963. Corrected chromosome numbers for Spartinax townsendii and its parent species. Nature 31: 929.Marchant, C.J. 1968. Evolution in Spartina (Gramineae). II.Chromosomes, basic relationships and <strong>the</strong> problem <strong>of</strong> Spartina xtownsendii agg. Botanical Journal <strong>of</strong> <strong>the</strong> Linnean. Society. 60:381-409.Marchant ,C.J. 1977. Hybrid characteristics in Spartina x neyrautiiFouc., a taxon rediscovered in nor<strong>the</strong>rn Spain. Botanical Journal<strong>of</strong> <strong>the</strong> Linnean Society 74: 289-296.Mobberley, D.G. 1956. Taxonomy and distribution <strong>of</strong> <strong>the</strong> genusSpartina. Iowa State College Journal <strong>of</strong> Science 30: 471-574.Osborn, T.C., J.C. Pires, J.A. Birchler, D.L. Auger, K.J. Chen, H.Lee, L. Comai, A. Madlung, R.W. Doergue, V. Colot, and R.A.Martissen. 2003. Understanding mechanisms <strong>of</strong> novel geneexpression in polyploids. Trends in Genetics 19, 3: 141 –147.Raybould, A.F., A.J. Gray, M.J. Lawrence, and D.F. Marshall.1991a. The evolution <strong>of</strong> Spartina anglica C.E. Hubbard(Gramineae): genetic variation and status <strong>of</strong> <strong>the</strong> parental species inBritain. Biological Journal <strong>of</strong> <strong>the</strong> Linnean Society 44: 369-380.Raybould, A.F., A.J. Gray, M.J. Lawrence, and D.F. Marshall.1991b. The evolution <strong>of</strong> Spartina anglica C.E. Hubbard(Gramineae): origin and genetic variability. Biological Journal <strong>of</strong><strong>the</strong> Linnean Society 43: 111-126.Riddle, N.C. and J.A. Birchler. 2003. Effects <strong>of</strong> reunited divergedregulatory hierarchies in allopolyploids and species hybrids.Trends in Genetics 19, 597-600.Salmon, A., M. Ainouche and J. Wendel. 2005. Genetic andepigenetic consequences <strong>of</strong> recent hybridization and polyploidyin Spartina (Poaceae). Molecular Ecology 14, 1163-1175.Soltis, D.E., P.S. Soltis, J.C. Pires, A. Kovarik, J.A. Tate, and E.Mavrodiev. 2004. Recent and recurrent polyploiyd in Tragopogon(Asteraceae): cytogenetic, genomic and genetic comparisons.Biological Journal <strong>of</strong> <strong>the</strong> Linnean Society 82, 485-501.Thompson, J.D. 1991. The biology <strong>of</strong> an invasive plant. Whatmakes Spartina anglica so successful? Bioscience 41, 393-401.Thompson, J.D., T. McNeilly, and A.J. Gray. 1991a. Populationvariation in Spartina anglica C.E. Hubbard. I. Evidence from acommon garden experiment. New Phytologist 117, 115-128.Thompson, J.D., T. McNeilly, and A.J. Gray. 1991b - Populationvariation in Spartina anglica C.E. Hubbard. II. Reciprocaltransplants among three successional populations. NewPhytologist 117, 129–139.Thompson, J.D., T. McNeilly, and A.J. Gray. 1991c - Populationvariation in Spartina anglica C.E. Hubbard. III. Response tosubstrate variation in a glasshouse experiment. New Phytologist117, 141–152.Wendel, J.F. 2000. Genome evolution in polyploids. PlantMolecular Biology 42, 225–249.Yannic, G., A. Baumel, M.L. Ainouche. 2004. Uniformity <strong>of</strong> <strong>the</strong>nuclear and chloroplast genomes <strong>of</strong> Spartina maritima (Poaceae)a salt marshes species in decline along <strong>the</strong> western EuropeanCoast. Heredity 93, 182-188.Malika L. Ainouche works on <strong>the</strong> evolution <strong>of</strong>polyploid genomes in <strong>the</strong> grass family (Spartina andBromus), and leads <strong>the</strong> research group “Genome Evolutionand Speciation” in <strong>the</strong> Department <strong>of</strong> Evolutionary Ecology<strong>of</strong> <strong>the</strong> UMR CNRS 6553 at University <strong>of</strong> Rennes1 (France).-21-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyEVOLUTION OF INVASIVE SPARTINA HYBRIDS IN SAN FRANCISCO BAYD.R. AYRES 1 AND D.R. STRONG 2Dept. <strong>of</strong> Evolution and Ecology, University <strong>of</strong> California, Davis, One Shields Avenue, Davis, CA 956161drayres@ucdavis.edu, 2 drstrong@ucdavis.eduRapid evolution in contemporary time occurs when genetically variable individuals face strong selectionpressures. Just such a conjunction is taking place in <strong>the</strong> intertidal marshes <strong>of</strong> <strong>the</strong> San Franciscoestuary. Three decades ago, exotic smooth cordgrass, Spartina alterniflora, was planted in <strong>the</strong> Bayfor erosion control and marsh restoration. The natural structure <strong>of</strong> <strong>the</strong> Pacific estuaries leaves broadexpanses <strong>of</strong> open mud flats, critical foraging grounds for millions <strong>of</strong> birds. Fringing <strong>the</strong> upper mudflat margin is <strong>the</strong> native cordgrass S. foliosa that, due to small stature and sparse growth, is unableto colonize <strong>the</strong> mud flats or modify <strong>the</strong>ir geomorphology, unlike <strong>the</strong> alien congener. Shortly after<strong>the</strong> introduction, <strong>the</strong> two species hybridized; in <strong>the</strong> last 20-odd years a broad array <strong>of</strong> genotypeshas arisen through reciprocal hybridization and introgression. Our research has found that somehybrids are taller, have more rapid rates <strong>of</strong> lateral expansion, have higher tolerance to salinity, havehigher rates <strong>of</strong> self-pollinated seed set, and are better sires on <strong>the</strong> native species than ei<strong>the</strong>r parentalspecies. We predict that <strong>the</strong>se traits will result in 1) rampant colonization <strong>of</strong> open mud — both in <strong>the</strong>intertidal and in restoration sites, 2) invasion <strong>of</strong> <strong>the</strong> middle elevation Salicornia-dominated salinemarsh plains, 3) isolated self-compatible plants founding new populations; and 4) hybrid pollensiring <strong>the</strong> lion’s share <strong>of</strong> seed in surrounding S. foliosa plants in native marshes. Natural selectionwill favor <strong>the</strong>se traits in a positive feedback that will result in an accelerating population growth rate<strong>of</strong> <strong>the</strong> fittest, most invasive hybrid genotypes. Indirect evidence that evolution has ready occurredis our finding <strong>of</strong> super-exponential growth <strong>of</strong> hybrid cordgrass cover in <strong>the</strong> Bay. Effective managementand prediction <strong>of</strong> hybrid cordgrass invasion must incorporate a dynamic viewpoint <strong>of</strong> Spartinapopulation parameters.Keywords: Invasive Spartina, hybridizationINTRODUCTIONSpartina alterniflora, smooth cordgrass, native to <strong>the</strong>Atlantic coast was introduced into San Francisco Bay in<strong>the</strong> 1970s. It hybridized with native California cordgrass, S.foliosa resulting in an interbreeding swarm <strong>of</strong> highly geneticallyvariable plants (Ayres et al. 1999; Anttila et al. 2000).One <strong>of</strong> <strong>the</strong> chief constraints on evolution is a lack <strong>of</strong> heritablevariation in fitness traits. Conversely, high variation in heritablefitness-related traits may fuel rapid evolution. Currentapproaches to understand successful plant invasion considerboth <strong>the</strong> invader and <strong>the</strong> invaded system; invasive plantspossess traits that allow <strong>the</strong>m to invade open niches in vulnerablesystems. With <strong>the</strong>se concepts in mind, we assessedgenetic variation in hybrid cordgrass, measured variation inphysical traits <strong>of</strong> cordgrass hybrids (S. alterniflora x foliosa)that would confer invasive ability in <strong>the</strong> open mud <strong>of</strong> Pacificestuaries and in <strong>the</strong> high salinity tidal plains dominated bypickleweed (Sarcocornia virginica, former name Salicornia),and evaluated rate <strong>of</strong> hybrid spread.METHODSGenetic variation. In previous work we documented<strong>the</strong> genetic variability <strong>of</strong> S. alterniflora x foliosa hybridsusing nuclear DNA markers and chloroplast DNA sequences(Ayres et al. 1999; Anttila et al. 2000).Phenotypic variation. We focused on several traits thatwe propose confer superior invasion ability. These traitswere 1) stem height, where tall stems allow plants to survivetidal inundation and grow fur<strong>the</strong>r down <strong>the</strong> intertidal plain;2) rapid lateral expansion that anchors plants in s<strong>of</strong>t mudsubstrates and allows encroachment into <strong>the</strong> space <strong>of</strong> o<strong>the</strong>rcordgrass plants; 3) high seed and pollen production to favorcolonization <strong>of</strong> open mud flat habitat and seed siring on o<strong>the</strong>rplants; and 4) tolerance to salinity and brackish water thatpermits growth in <strong>the</strong> Sarcocornia higher marsh (high salinity)and tidal creeks (brackish water).At Cogswell Marsh, a former salt pond restored to tidalaction in 1980, we compared variation in vegetative andsexual reproduction among established plants (Zaremba2001). Seventy-two plants were haphazardly selected froma 1998 aerial photograph; <strong>the</strong>ir genotypes were determinedusing nuclear DNA analysis (Ayres et al. 1999; Zaremba2000). Heights <strong>of</strong> <strong>the</strong> three tallest stems (excluding inflorescences)in three randomly placed quadrats were measuredin 1998 and 1999 (results reported for 1998 only). Diameterswere staked and re-measured in 1998, 1999 and 2000.Inflorescences were randomly collected (three per plant)and assessed for seed set and pollen viability and quantity(see Zaremba 2001 for a full description <strong>of</strong> methods). In agreenhouse at UC Davis, we compared growth <strong>of</strong> an array <strong>of</strong>-23-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinahybrids under three salinity regimes (Pakenham-Walsh et al.this volume; Pakenham-Walsh 2003).Spread rate. We developed physical indicators <strong>of</strong> hybridity(tall, wide, red, stems; late flowering) in greenhouse andfield studies to identify hybrids by sight in <strong>the</strong> field (Ayres etal. 2004). We combined aerial photographs, ground surveys,and genetic confirmations to estimate areal cover by hybridcordgrass and S. alterniflora in 2001. We estimated cordgrasshybrid cover from <strong>the</strong> original 1976 population, and in<strong>the</strong> 1990s from early studies (Callaway and Josselyn 1992;Daehler and Strong 1996) to plot cordgrass cover over time.RESULTSGenetic variation. Diverse hybrid nuclear DNA genotypeswere present (Ayres et al. 1999), evidence that widespreadcrossing has occurred among cordgrass genotypes.Using chloroplast sequences, we found that hybridizationwas bi-directional; that is, both S. alterniflora and S. foliosahave been seed parents to hybrids (Antilla et al. 2000). Wealso determined that S. foliosa has been <strong>the</strong> dominant seedparent <strong>of</strong> hybrids.Phenotypic variation. A diverse array <strong>of</strong> hybrid genotypeswas present in <strong>the</strong> sampled population at CogwesllMarsh. Nine <strong>of</strong> <strong>the</strong> 72 plants were genetically determinedto be (non-hybrid) parental species; four plants <strong>of</strong> native S.foliosa and five plants <strong>of</strong> S. alternilfora. (Note: plants in Figs.1-3 are arranged according to <strong>the</strong> magnitude <strong>of</strong> <strong>the</strong> trait and<strong>the</strong> order is NOT <strong>the</strong> same in all graphs.)Height. Individual plant height varied from < 30 cm to>120 cm (Fig. 1). Both S. foliosa and S. alterniflora plantsranged from 30 to 70 cm in height. While <strong>the</strong> tallest 23% <strong>of</strong><strong>the</strong> plants were hybrids, most hybrids fell within <strong>the</strong> heightrange <strong>of</strong> <strong>the</strong> parental species, and three plants were shorterthan <strong>the</strong> shortest S. foliosa plant.Lateral spread. Plant areal increase varied from plantsthat shrunk in area after two growing seasons (1998–1999,and 1999–2000) to a plant that expanded over 200 m 2 (Fig. 2;data are presented for 52 plants for which we had a completedata set). Plants <strong>of</strong> <strong>the</strong> parental species had low to moderategrowth rates while <strong>the</strong> fastest growing 30% <strong>of</strong> plants werehybrids.Salinity tolerance. We found that under conditions <strong>of</strong>high salinity several hybrids exceeded <strong>the</strong> growth <strong>of</strong> bothparental plants by two to three times, and by as much as sixtimes under low salinity conditions (Pakenham-Walsh 2003;Pakenham-Walsh, Ayres and Strong in this proceedingsvolume).Sexual reproduction. Hybrid inflorescences averagedover twice as many flowers as those <strong>of</strong> S. foliosa (460 vs200), but hybrid inflorescences varied from fewer than 125flowers to inflorescences with over 800 flowers (Fig. 3, Ayreset al. 2008). Fertile seed per inflorescence was one and ahalf times higher in hybrids than in S. foliosa, but varied inhybrid plants from zero seed set (including a plant with over800 flowers per inflorescence) to five times higher seed setthan <strong>the</strong> native. We determined from genetic analyses thatbetween 50% and 80% <strong>of</strong> <strong>the</strong> seeds from <strong>the</strong> four S. foliosaFig. 1. Stem height in 1998 <strong>of</strong> cordgrass plants (individual plants are on <strong>the</strong> x-axis) at Cogswell Marsh. Grey arrows are S. foliosa and black arrows are S.alterniflora plants (by genetic analyses); <strong>the</strong> rest are hybrids.-24-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina Biology250200Clonal Spread (m-squ)150100500-50Fig. 2. Clonal spread in area <strong>of</strong> individual plants (along <strong>the</strong> x-axis) at Cogswell Marsh from 1998 to 2000 (complete data was only available for 52plants). Grey arrows are S. foliosa and black arrow is S. alterniflora plants (by genetic analyses); <strong>the</strong> rest are hybrids.10008006004002000Fertile seedInfertile seedFig. 3. Sexual reproduction <strong>of</strong> 48 cordgrass plants at Cogswell Marsh in 1998 (many plants did not flower); individual plants are along <strong>the</strong> x-axis.Each bar is <strong>the</strong> average number <strong>of</strong> flowers on an inflorescence; <strong>the</strong> black portion represents those florets that contained a fertile seed, <strong>the</strong> open portionrepresents those flowers without a seed. The S. foliosa plants are along <strong>the</strong> left side, with grey bars; <strong>the</strong> single S. alterniflora did not flower.-25-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartina1.00.80.60.40.20.0S. foliosa Hybrids S. foliosa Hybrids S. foliosa HybridsPollen Viability Pollen/m 2 Pollen/PlantFig. 4. Relative pollen viability and production <strong>of</strong> S. foliosa and cordgrass hybrids; box-whisker plots show 25th, 50th and 75th per-centiles solidlines in <strong>the</strong> “box”, whiskers are <strong>the</strong> 5th and 95th percentile; solid circles are at <strong>the</strong> 2.5 and 95.6 percentile; dotted line is <strong>the</strong> mean. This figurefirst appeared in Ayres et al. 2008.San Francisco vs Willapa Invasions• S. alterniflora• “Open niche”• 25 YA• Hybridization• Evolution fueled bygenetic variation• Accelerating spread• S. altern iflora• “Open niche”• 100 YA• No native Spartina• Evolution constrainedby lack <strong>of</strong> variation• Constant spread rateFig. 5. Comparison <strong>of</strong> S. alterniflora introductions into 2 Pacific coast USAestuaries, San Francisco Bay, CA, and Willapa Bay, WA. We posit thatspread rates are a function <strong>of</strong> <strong>the</strong> amount <strong>of</strong> genetic variation.plants in <strong>the</strong> marsh were hybrids. The net result was that over95% <strong>of</strong> <strong>the</strong> seed produced from all 42 plants in our study (ca.40 million seeds in 1998) was hybrid.Hybrid pollen had 3.5 times <strong>the</strong> viability <strong>of</strong> S. foliosapollen (8.24% vs 2.38% viable grains, respectively), andhybrid plants produced twice <strong>the</strong> number <strong>of</strong> pollen grainsas native plants (9.5 million vs. 4.3 million grains per plant)(Fig. 4, Ayres et al. 2008). Viable pollen production rangedfrom a single hybrid individual (plant C1-5) that produced58% <strong>of</strong> <strong>the</strong> total viable pollen to zero. The cumulative effects<strong>of</strong> <strong>the</strong> scarcity <strong>of</strong> S. foliosa plants, and differences in pollenproduction and viability resulted in hybrid plants producingover 400 times more viable pollen than <strong>the</strong> native plants.Spread Rate. Combining <strong>the</strong> results <strong>of</strong> our 2000–2001cordgrass survey with literature-based estimates <strong>of</strong> populationssize, we found that spread rates <strong>of</strong> hybrid and S. alterniflorahave increased over time (Ayres et al. 2004), from10% (during 1973 to 1992) to 40% (from 1993 to 2001) peryear. This pattern is apparently not typical <strong>of</strong> S. alterniflorainvasion in general as analyses <strong>of</strong> spread in Willapa Bay,Washington found a constant growth rate <strong>of</strong> 12% over <strong>the</strong>past 100 years (Civille et al. 2005; Taylor et al. 2004, and inthis proceeding volume).DISCUSSION AND CONCLUSIONSWe have shown that hybrids are genetically variabledue to interbreeding among hybrids and backcrossing to <strong>the</strong>native (Ayres et al. 1999; Anttila et al. 2000). In <strong>the</strong> presentstudy, we demonstrate that hybrids are also phenotypicallyhighly variable in features <strong>of</strong> vegetative and sexual reproduction,with a subset <strong>of</strong> hybrid individuals having superiorgrowth and contributing disproportionately to seed andpollen production.We <strong>the</strong>orize that <strong>the</strong> open niche <strong>of</strong> native marshes,intertidal mud flats, and restoration sites provides strongselection for vigorous vegetative growth and seed production.Hybridization has provided large amounts <strong>of</strong> geneticand phenotypic variation for <strong>the</strong>se traits. We suggest that <strong>the</strong>combination <strong>of</strong> natural selection and variability in fitnesstraits sets up a scenario <strong>of</strong> positive feedback whereby strongnatural selection favors a subset <strong>of</strong> highly fit hybrids thatdrives growth rate, followed by fur<strong>the</strong>r selection on hybridplants for invasion ability and enhanced growth rates (Fig. 5).The accelerating growth rate we have observed in cordgrasspopulations in San Francisco Bay is indirect evidence thatthis has already occurred.This outcome contrasts with that in Willapa Bay, whereS. alterniflora was introduced into <strong>the</strong> same open niche <strong>of</strong>intertidal mudflat habitat 100 years ago. There is no nativeSpartina at Willapa, and <strong>the</strong>refore hybridization has notoccurred. We propose that evolution <strong>the</strong>re has been constrainedby a relative lack <strong>of</strong> genetic diversity in <strong>the</strong> popula--26-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina Biologytion <strong>of</strong> S. alterniflora), resulting in constant spread rates for<strong>the</strong> last 100 years (Fig. 5).We conclude that if cordgrass is rapidly evolving in itsinvasion ability, it will be difficult to precisely predict itseventual ecological range or where <strong>the</strong> tidal marsh system isor will be vulnerable.ACKNOWLEDGMENTSWe thank Katy Zaremba for allowing us to draw extensivelyupon her Master’s <strong>the</strong>sis data, Mark Taylor and <strong>the</strong>East Bay Regional Park District for providing our field siteand supporting our research efforts over many years (andstanding by to render aid if we got stuck in <strong>the</strong> mud), <strong>the</strong>many graduate students for <strong>the</strong>ir intellectual collaborations,team Spartina for outstanding technical help at both UCDavis and <strong>the</strong> Bodega Marine Laboratory, and our fundingsources — California Coastal Conservancy (CalFed grant#99-110), California Sea Grant #27CN to D.R. Strong, andNSF Biocomplexity DEB 0083583 to A. Hastings and D.R.Strong.REFERENCESAnttila, C.K., A.R. King, C. Ferris, D.R. Ayres, and D.R. Strong.2000. Reciprocal hybrid formation <strong>of</strong> Spartina in San FranciscoBay. Molecular Ecology 9:765-771.Ayres, D.R., D. Garcia-Rossi, H.G. Davis, and D.R. Strong. 1999.Extent and degree <strong>of</strong> hybridization between exotic (Spartinaalterniflora) and native (S. foliosa) cordgrass (Poaceae) inCalifornia, USA determined by random amplified polymorphicDNA (RAPDs). Molecular Ecology 8:1179-1186.Ayres, D.R., D.L. Smith, K. Zaremba, S. Klohr, and D.R. Strong.2004. Spread <strong>of</strong> exotic cordgrasses and hybrids (Spartina sp.)in <strong>the</strong> tidal marshes <strong>of</strong> San Francisco Bay. Biological Invasions6:221-231.Ayres, D.R., K. Zaremba, C.M. Sloop and D.R. Strong. 2008.Sexual reproduction <strong>of</strong> cordgrass hybrids (Spartina foliosax alterniflora) invading tidal marshes in San Francisco Bay.Diversity and Distributions 14:187-195.Callaway, J.C., and M.N. Josselyn. 1992. The introduction andspread <strong>of</strong> smooth cordgrass (Spartina alterniflora) in South SanFrancisco Bay. Estuaries 15:218-226.Civille, J.C., K. Sayce, S.D. Smith and D.R. Strong. 2005.Reconstructing a century <strong>of</strong> Spartina invasion with historicalrecords and contemporary remote sensing. Ecoscience 12:330-338.Daehler, C.C. and D.R. Strong. 1996. Status, prediction and prevention<strong>of</strong> introduced cordgrass Spartina spp. invasions in Pacificestuaries, USA. Biological Conservation 78:51-58.Pakenham-Walsh, M.R. 2003. Variation in salinity tolerance andcompetitive ability <strong>of</strong> invasive Spartina hybrids in San FranciscoBay. Master <strong>of</strong> Science <strong>the</strong>sis. University <strong>of</strong> California, Davis.Taylor, C.M., H.G. Davis, J.C. Civille, F.S. Grevstad, and A.Hastings. 2004. Consequences <strong>of</strong> an Allee effect on <strong>the</strong> invasion<strong>of</strong> a Pacific estuary by Spartina alterniflora. Ecology85:3254-3266.Zaremba, K. 2001. Hybridization and Control <strong>of</strong> a Native-Non NativeSpartina Complex in San Francisco Bay. Master <strong>of</strong> Arts <strong>the</strong>sis,San Francisco State University, San Francisco, California.-27-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyEVOLVING INVASIBILITY OF EXOTIC SPARTINA HYBRIDS IN UPPER SALT MARSH ZONESOF SAN FRANCISCO BAYM.R. Pakenham-Walsh 1 , D.R. Ayres 2 , and D.R. Strong 21 Regulatory Division, U.S. Army Corps <strong>of</strong> Engineers, 1325 J Street, Room 1480, Sacramento, CA 958142 Section <strong>of</strong> Evolution and Ecology, University <strong>of</strong> California, Davis, One Shields Avenue, Davis, CA 95616Keywords: Spartina, Sarcocornia virginica, invasive, hybridizationINTRODUCTIONInvasion by hybrid cordgrass (Spartina alterniflora xSpartina foliosa) is pr<strong>of</strong>oundly altering habitat structurewithin <strong>the</strong> intertidal zone <strong>of</strong> San Francisco Bay, California.Spartina hybrids exhibit wide ecological tolerance comparedto native S. foliosa and are invading <strong>the</strong> naturallyunvegetated lower intertidal zone. Heterogeneous hybridgenotypes exhibit traits <strong>of</strong> both salinity tolerance andcompetitive vigor, which may enable invasion <strong>of</strong> highermarsh zones historically dominated by <strong>the</strong> highly salttolerantnative pickleweed species Sarcocornia virginica.The hybrid cordgrass swarm possesses a high degree <strong>of</strong>genetic variation, with bi-directional introgression due tooverlapping flowering periods <strong>of</strong> hybrids and both parentalspecies (Ayres et al. 1999; Antilla et al. 2000). Hybridizationhas been shown to play a role in large and rapid adaptiveevolution (Rieseberg et al. 2003), promoting <strong>the</strong> ability forniche separation between hybrids and parental species(Rieseberg et al. 1999).Drawing on field observations indicating higher relativefitness <strong>of</strong> hybrid Spartina compared to native S. foliosa in<strong>the</strong> higher marsh zones, we conducted two experiments toaddress <strong>the</strong> threat <strong>of</strong> hybrid colonization <strong>of</strong> Sa. virginicahabitat. A greenhouse experiment investigated salinitytolerance <strong>of</strong> hybrids. A field experiment examinedcompetitive suppression <strong>of</strong> hybrids by Sa. virginica.METHODSFor <strong>the</strong> greenhouse experiment, clonal fragments <strong>of</strong>eighteen hybrid genotypes were collected from CogswellMarsh, Hayward, California. In addition to <strong>the</strong> hybrids, oneS. alterniflora and two S. foliosa genotypes (21 plants total)were grown at three salinity levels (10, 25 and 40 ppt) withthree replications for one growing season in Davis,California (Fig. 1). Total dry biomass (roots and abovegroundmaterial) was determined at <strong>the</strong> end <strong>of</strong> <strong>the</strong> season,and effects <strong>of</strong> genotype, salinity and <strong>the</strong>ir potentialinteraction determined with statistical analysis.The field experiment was conducted at Cogswell Marsh,a former salt pond opened to tidal action in 1980. Four plottreatments (pure Spartina sp., pure Sa. virginica, clipped Sa.virginica and unclipped Sa. virginica; five replicates) wereset up at each <strong>of</strong> eight hybrid clone locations and one S.foliosa clone (Fig. 2). End-<strong>of</strong>-season shoot density andabove-ground biomass were determined for each plot, andeffects <strong>of</strong> genotype and plot type were determined withstatistical analysis.Fig. 1., left.40 ppt. (left) vs.10 ppt. (right)in salinityexperiment.Fig. 2., right.Location <strong>of</strong>nine Spartinagenotypes atCogswell Marsh.-29-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartina180.00160.00140.00Total Dry Biomass (g)120.00100.0080.0060.0040.0020.000.002-15(foli)S. foli 1-9 S. alt 3-7 3-2 3-4 2-5 2-17 1-5 2-13 3-13 3-1 3-10 3-5 2-6 1-17 2-8 2-16 3-8 2-11Spartina GenotypeLow Medium HighFig. 3. Total biomass <strong>of</strong> Spartina genotypes at each salinity level, in increasing order <strong>of</strong> total biomass achieved by each genotype(low + medium + high salinities).RESULTSHybrid Spartina genotypes exhibited great variability inmorphological traits, response to salinity stress andcompetitive suppression by Sa. virginica. Salinity reducedtotal biomass <strong>of</strong> most Spartina genotypes. Several hybridcordgrass genotypes exhibited more robust growth than <strong>the</strong>parental species at both low and high salinity levels (Fig. 3).Results <strong>of</strong> <strong>the</strong> field experiment indicated that shootdensity <strong>of</strong> hybrid cordgrasses was highly sensitive to Sa.virginica removal, increasing an average <strong>of</strong> 105% (range =39-211%; Table 1). End-<strong>of</strong>-season biomass <strong>of</strong> cordgrassgenotypes increased an average <strong>of</strong> 64% in response to Sa.virginica removal (range = 2-144%; Table 1). Genotypes inTable 1 are listed in order <strong>of</strong> appearance on Fig. 3.CONCLUSIONSCertain hybrids (e.g., genotypes 1-5 and 2-8) showed acombination <strong>of</strong> competitive vigor in <strong>the</strong> field and relativelystronger performance in higher salinity conditions,exhibiting <strong>the</strong> potential to successfully expand into highermarsh zones. The native S. foliosa has weaker competitiveabilities and tolerance for salinity. Our results support apositive association between hybridization and invasionability. The genetic heterogeneity <strong>of</strong> Spartina hybrids willlead to <strong>the</strong> evolution <strong>of</strong> even more invasive hybridpopulations.Table 1. Genotype means and percent change between plots with S. virginica "clipped" vs. unclipped, end-<strong>of</strong>-season shoot densityand aboveground dry biomass.Spartina Shoot Density (#/m 2 ) Dry Biomass (g/m 2 )Genotype Unclipped Clipped % Change Unclipped Clipped % Change2-15 (S. foli) 41 84 104 107 206 931-9 107 235 119* 212 330 562-17 84 117 39 234 243 41-5 67 136 103* 695 1360 96*3-13 77 173 126* 277 426 543-1 74 117 59 464 471 21-17 65 104 60 512 1251 144*2-8 62 139 124* 501 556 113-8 46 142 211* 159 342 115* Significant difference (P < 0.05)-30-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyACKNOWLEDGMENTSWe would like to thank <strong>the</strong> staff <strong>of</strong> <strong>the</strong> East BayRegional Shoreline, East Bay Regional Parks District, forsupport <strong>of</strong> field work at Cogswell Marsh, and <strong>the</strong> support <strong>of</strong>many individuals <strong>of</strong> <strong>the</strong> U.C. Davis Spartina Lab for field,greenhouse and laboratory assistance. This work wassupported by <strong>the</strong> California Sea Grant #27CN to DRS.REFERENCESAntilla, C.K., R.A. King, C. Ferris, D.R. Ayres, and D.R. Strong.2000. Reciprocal hybrid formation <strong>of</strong> Spartina in San FranciscoBay. Molecular Ecology 9:765-770.Ayres, D.R., D. Garcia-Rossi, H.G. Davis, and D. R. Strong. 1999.Extent and degree <strong>of</strong> hybridization between exotic (Spartina alterniflora)and native (S. folisa) cordgrass (Poaceae) in California,USA determined by random amplified polymorphic DNA(RAPDs). Molecular Ecology 8: 1179-1186.Rieseberg, L.H., M.A. Archer, and R.K. Wayne. 1999. Transgressivesegregation, adaptation and speciation. Heredity 83:363-372.Rieseberg, L.H., O. Raymond, D.M. Rosenthal, Z. Lai, K. Livingstone,T. Nakazato, J.L. Durphy, A.E. Schwarzbach, L.A. Donovan,and C. Lexer. 2003. Major ecological transitions in wildsunflowers facilitated by hybridization. Science 301: 1211-1216.-31-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyVARYING SUCCESS OF SPARTINA SPP. INVASIONS IN CHINA:GENETIC DIVERSITY OR DIFFERENTIATION?S. AN 1 , Y. XIAO 1 , H. QING 1 , Z. WANG 1 , C. ZHOU 1 , B. LI 2 , S. SHI 3 , D. YU 1 , Z. DENG 1 , AND L. CHEN 11School <strong>of</strong> Life Science, Nanjing University, Nanjing 210093, China; anshq@nju.edu.cn2Ministry Of Education Key Laboratory For Biodiversity Science And Ecological Engineering, Institute Of BiodiversityScience, Fudan University, Shanghai 200433, China; Bool@fudan.nju.edu.cn3School <strong>of</strong> Life Science, Zhongshan University, Guangzhou 510275, China; Sshlss@zsu.edu.cnBiological invasions are widely envisaged as a component <strong>of</strong> global change, which not only threatennative biodiversity but also cause a considerable economic loss to <strong>the</strong> invaded areas. However, forsome species little is known about why <strong>the</strong>y are successful invaders. At <strong>the</strong> molecular level, manyresearchers have reported that high genetic diversity contributes to <strong>the</strong> success <strong>of</strong> plant invasionswhereas o<strong>the</strong>r studies have shown that reduced genetic variation makes invasive species more successful.Obviously, <strong>the</strong>se are contradictory explanations. Our studies <strong>of</strong> Spartina in China showthat S. alterniflora with higher population differentiation (Fst) was more successful in invadingcoastal China than S. anglica which had low Fst, although <strong>the</strong> latter species, having fixed heterozygosity,had much higher genetic diversity than <strong>the</strong> former. Greater number <strong>of</strong> chromosomes orhigher genetic diversity (P and H) does not necessarily mean higher adaptability and more successfulinvasion for exotic species, while higher Fst was associated with higher invading ability in S.alterniflora. However, factors o<strong>the</strong>r than genetic variation and population genetic structure may bemore important in determining invasion success.Keywords: Average heterogeneity, genetic diversity, plant invasion, population differentiation,SpartinaInvasive plant species threaten <strong>the</strong> integrity <strong>of</strong> naturalecosystems and reduce <strong>the</strong> growth <strong>of</strong> economic cropsthroughout <strong>the</strong> world by displacing native plant communities(Kennedy et al. 2002), establishing monoculturesand competing with economic species in new habitats(Callaway 2002). At <strong>the</strong> population level, <strong>the</strong> leading <strong>the</strong>oriesfor <strong>the</strong> successful invasion <strong>of</strong> plants are <strong>the</strong>ir escapefrom <strong>the</strong> natural enemies that hold <strong>the</strong>m in check, freeing<strong>the</strong>m to utilize <strong>the</strong>ir full potential for resource competition(Keane & Crawley 2002); allelopathic effects, where <strong>the</strong>phytotoxins released by exotic plants damage native species(Callaway & Aschehoug 2000; Bais et al. 2003); and <strong>the</strong>occurrence <strong>of</strong> a suitable niche existing in <strong>the</strong> new location(Sakai et al. 2001). At <strong>the</strong> molecular level, most researchershave reported that high genetic diversity contributes to <strong>the</strong>success <strong>of</strong> plant invasions (Ellstrand & Schierenbeck 2000;Novak & Mack 2001), although reduced genetic variationcan make invasive species more successful (Tsutsui et al.2000). Here we suggest that <strong>the</strong> different molecular parametersare correlated with success or failure <strong>of</strong> Spartinainvasions in China.Historically, no native species <strong>of</strong> <strong>the</strong> genus Spartinaexisted in China. For <strong>the</strong> purpose <strong>of</strong> ecological engineering,two Spartina species, S. anglica and S. alterniflora, wereintroduced to China in <strong>the</strong> last century. However, <strong>the</strong>ir fatesdiffered. Spartina anglica originated in 1870s in east coastalEngland and is a hybrid between S. maritima as paternalspecies and S. alterniflora as <strong>the</strong> seed parent (Ferris et al.1997). In 1963, 44 individuals (actually ramets) <strong>of</strong> S. anglicawere produced by seed in China; 21 <strong>of</strong> <strong>the</strong>m were planted in<strong>the</strong> field and <strong>the</strong> o<strong>the</strong>rs were used to produce more <strong>of</strong>fspring,three <strong>of</strong> which had marvelous reproductive capacity (Chung1985). From 3 individuals, 9,100,000 clones were produced in1966, and were planted in 40 hectares (ha) <strong>of</strong> costal regionsin China. Thirteen years later, about 31,600 ha <strong>of</strong> S. anglicalow marshes were established in coastal China, and 36,000ha in 83 counties along coastal China (Fig. 1a) in 1985 (Qin& Chung 1992). Its sou<strong>the</strong>rn limit in China was at 21º27’Nwhereas <strong>the</strong> sou<strong>the</strong>rn limit in its native range is locatedbetween 42ºand 43º N in Europe (Chung 1993).The populations <strong>of</strong> S. anglica in China have generallylower height, lower biomass and lower net production (Chung1985) than <strong>the</strong>ir counterparts in its native range (Chater &Jones 1951; Gray & Benham 1991; Hubbard 1969) (Table 1).Starting in 1993, S. anglica became shorter in height, andlower in biomass and net biomass production. Pronounceddieback occurred in Chinese populations <strong>of</strong> S. anglica like<strong>the</strong> older populations in England (Gray et al. 1991). In 2000,S. anglica was found only in six counties (Fig. 1b). In 2002,<strong>the</strong> populations <strong>of</strong> S. anglica were found only in three counties,and <strong>the</strong> total area was less than 50 ha. Plant performancecontinued to be suppressed, e.g., height, biomass and netproduction were reduced; and <strong>the</strong> suppression still continues.Most <strong>of</strong> <strong>the</strong> S. anglica populations have been competitivelyreplaced by S. alterniflora, Phragmites australis, Typha spp.and Scirpus spp.-33-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaTable 1. Variation <strong>of</strong> Spartina anglica between native range and Chinaai s i r is s 1iri i1 2 4 6 H d s86 10 0.8 0.8 1.6 0.8 0.8 0.126 0.4202 0.0 84 0.0 20rs 1 1 1 1 1 66 1 8 2000 2002i ss2 41 .8 12 1. 2 24 . 8 8 1.0 21 .28 02.6r di2 yr44 00 110. 601.8 6. 112.4H i 0 0 4 1 80 0d r r ri s s r s 1 . 20012 H bb rd 1 6 r y d 1 1 d 4 r d s 1 1 .bFig. 1. Distribution <strong>of</strong> Spartina anglica and Spartina alterniflora in 1985 (a)and 2000 (b) along coastal China. Circles represent Spartina anglica andstars denote Spartina alterniflora.Seven multilocus genotypes (G1-G7) were identifiedusing RAPD markers for S. anglica in its native Europe(Baumel et al. 2001). G1 is <strong>the</strong> basic type that accounts for86% <strong>of</strong> <strong>the</strong> samples (Table 1). G2 – G7 are varieties <strong>of</strong> G1and only account for 10% - 0.8% <strong>of</strong> <strong>the</strong> total. The studyshowed that S. anglica has low genetic variation due to agenetic bottleneck. As an exotic species, S. anglica in Chinahas even lower genetic variation with lower percentage <strong>of</strong>polymorphical loci (P), mean coefficient <strong>of</strong> dissimilarity(Gd) and coefficient <strong>of</strong> population differentiation (Fst) (Table1) although AFLP markers can provide higher diversity thanRAPD markers. Lower variation may indicate that <strong>the</strong> speciespartially lost some <strong>of</strong> its variation during <strong>the</strong> human-aidedinvasions in China, since only small parts <strong>of</strong> native populationwere collected in <strong>the</strong> introduction process.S. alterniflora is a native species <strong>of</strong> north AtlanticAmerica. In 1979, 60 individuals and hundreds <strong>of</strong> seeds werecollected from Florida, Georgia and North Carolina, consisting<strong>of</strong> three distinct populations that differed in height,biomass production and protein (Chung 1985; Qin & Chung1992), namely F-type (Florida-type), G-type (Georgia-type)and N-type (North Carolina-type). In 1981, about 400 squaremeters (m 2 ) <strong>of</strong> planted area were established for each population,and in 1985 <strong>the</strong> area expanded to 260 ha with humanhelp (Qin & Chung 1992) (Fig. 1a). In 1990, <strong>the</strong>re were 1,300ha <strong>of</strong> S. alterniflora monoculture along coastal China, butonly G-type remained as <strong>the</strong> two o<strong>the</strong>r types were replacedby <strong>the</strong> G-type through intraspecific competition (Guan et al.2003). In 2000, <strong>the</strong> total area reached 120,000 ha (Guan et al.2003) (Fig. 1b).Unlike S. anglica, <strong>the</strong> exotic S. alterniflora has highreproductive and dispersal capacity, and is competitive; ittook only 19 years to increase its area from 0.08 ha to 21,300ha in Jiangsu Province (Shen & Liu 2002). It excluded almostall <strong>the</strong> o<strong>the</strong>r native plants that were originally dominantin wetlands, including Phragmites australis, Typha spp.,Scirpus spp., Suadae spp., and even invaded fishponds andyoung mangrove swamps (Qian & Ma 1995). Many nativespecies, including plants, some endangered birds (Ma etal. 2003), and molluscs <strong>of</strong> economic importance in coastalChina are threatened by S. alterniflora invasions (Qian &Ma 1994). Spartina alterniflora is one <strong>of</strong> nine notoriousinvasive pest plants in China since <strong>the</strong> species directly causesmillions <strong>of</strong> dollars <strong>of</strong> economic loss per year (Chen 1998).Spartina alterniflora is still rapidly replacing <strong>the</strong> nativeplants, although <strong>the</strong> Chinese government and scientists aredoing <strong>the</strong>ir best to control or eradiate <strong>the</strong> species by physical,chemical, biological and integrated methods (Lin 1997; Liu& Huang 2000).S. alterniflora populations in China have similar percent<strong>of</strong> polymorphic loci (P) to native populations (Travis et al.2002), but we found that <strong>the</strong> species in China has muchhigher average heterogeneity (H), coefficient <strong>of</strong> populationdifferentiation (Fst), and much lower mean coefficient <strong>of</strong>dissimilarity (Gd) (Table 2). The introduced S. alterniflorahas accumulated much population genetic differentiation,-34-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyTable 2. Genetic variation <strong>of</strong> Spartina alterniflora in native range andChina based on AFLP markers. percentage <strong>of</strong> polymorphical loci (P), heterogeneity(H), mean coefficient <strong>of</strong> dissimilarity (Gd) and coefficient <strong>of</strong>population differentiation (Fst).i r iH d s H d s0. 0 0 0.11 4 0.2 0.064 0.280 0. 4 0.1 4 0.20 6d r r r is . 2002 .and considerably increased its invading capacity into coastalChina since its arrival in 1979. Rapid and genetic differentiationmay aid S. alterniflora in being a successful invasivespecies.Although S. anglica arrived in China 16 years earlierthan its seed parent S. alterniflora, <strong>the</strong> species has muchlower genetic differentiation in coastal China’s environments(Tables 1 and 2). Spartina anglica experienced a geneticbottleneck during its formation, which might have contributedto its dieback in China as in its native Europe (Gray etal. 1991; Thompson 1991).S. anglica has 120-124 chromosomes while S. alterniflorahas 62 (Gray & Benham 1991). Spartina anglica shouldhave high heterosis and potentially higher adaptability thanei<strong>the</strong>r parental species. Spartina anglica has replaced S.maritima in coastal England (Gray et al. 1991). However, S.alterniflora has rapidly occupied previous S. anglica habitatsin China. We suggest that <strong>the</strong> species with low population differentiation(Fst) may be outcompeted by a species with highdifferentiation in changing environments like coastal areas<strong>of</strong> China, where accidental typhoons, irregular tidal cyclesand human-caused disturbances <strong>of</strong>ten happen. Fur<strong>the</strong>r, agreater number <strong>of</strong> chromosomes or higher genetic diversity(P and H) does not guarantee a higher adaptability and moresuccessful invasion for exotic species. However, <strong>the</strong> generality<strong>of</strong> this finding needs to be tested fur<strong>the</strong>r in more taxa.ACKNOWLEDGMENTSThis study was financially supported by NaturalScientific foundation <strong>of</strong> China (No. 30400054), by NKBRSFoundation, China (No. G2000046803) and KSJYXRCFoundation, Ministry <strong>of</strong> Education, China (No. 0208002002).We thank to Dr. Chung-Hsin Chung for his valuable suggestionsand comments.REFERENCESBais, H.P., R. Vepachedu, S. Gilroy, R.M. Callaway, and J.M.Vivanco. 2003. Allelopathy and exotic plant invasion: from moleculesand genes to species interactions. Science 301:1377-1380.Baumel, A., M.L. Ainouche, and J.E. Levasseur. 2001. Molecularinvestigations in populations <strong>of</strong> Spartina anglica C.E. Hubbard(Poaceae) invading coastal Brittany (France). Molecular Ecology10(7):1689-1701.Callaway, R.M. 2002. The detection <strong>of</strong> neighbors by plants. Trendsin Ecology and Evolution 17(3):104-105.Callaway, R.M. and E.T. Aschehoug. 2000 . Invasive plants versus<strong>the</strong>ir new and old neighbors: A Mechanism for exotic invasion.Science 290:521-523.Chater, E.H. and J.H. Jones. 1951. New form <strong>of</strong> Spartina townsendii.Nature 168:126-128.Chen, C.D. 1998. China’s Biodiversity: A Country Study. EnvironmentalScience Press.Chung, C.H. 1985. Twenty-two years <strong>of</strong> Spartina anglica in China.Journal <strong>of</strong> Nanjing University, Special Issue.Chung, C.H. 1993. Thirty years <strong>of</strong> ecological engineeringwith Spartina plantation in China. Ecological Engineering2:261-289.Ellstrand, N.C. and K.A. Schierenbeck. 2000. Hybridization as astimulus for <strong>the</strong> evolution <strong>of</strong> invasiveness in plant. <strong>Proceedings</strong><strong>of</strong> <strong>the</strong> National Academy <strong>of</strong> Sciences <strong>of</strong> <strong>the</strong> United States <strong>of</strong>America 97(13):7043-7050.Ferris C., R. A. King and A.J. Gray. 1997. Molecular evidence for<strong>the</strong> maternal parentage in <strong>the</strong> hybrid origin <strong>of</strong> Spartina anglicaC. E. Hubbard. Molecular Ecology 6:185-187.Gray, A. J., D.F. Marshall and A.F. Raybould. 1991. A century <strong>of</strong>evolution in Spartina anglica. Advances in Ecological Research21:1-62.Gray, A.J. and P.E.M. Benham 1991. Spartina anglica — a researchreview. HMSO.Guan Y.J., S.Q. An, Y.H Liu., et al. 2003. Ecological benefit andecological invasions <strong>of</strong> Spartina in China. Journal <strong>of</strong> NanjingForestry University 27:95-101.Hedge, P., L. Kriworken, and A. Ritar. 1997. Ecology, impact andcontrol <strong>of</strong> Spartina in Little Swanport, Tasmania. In: Patten,K., ed., <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> Second <strong>International</strong> Spartina<strong>Conference</strong>, Olymipia, WA. Washington State UniveristyCooperative Extension, pp. 46-47.Hubbard, J.C.E. 1969 . Light in relation to tidal immersion and <strong>the</strong>growth <strong>of</strong> Spartina townsendii. Journal <strong>of</strong> Ecology 57: 795-804.Kennedy, T.A. et al. 2002. Biodiversity as a barrier to ecologicalinvasion. Nature 417:636-638.Keane, R.M. and M.J. Crawley. 2002. Exotic plant invasions and<strong>the</strong> enemy release hypo<strong>the</strong>sis. Trends in Ecology and Evolution17:164-170.Lin, Q.R. 1997. Damage and management <strong>of</strong> Spartina anglica andSpartina alterniflora. Fujian Geography 12(1):16-20.Liu, J. and J.H. Huang. 2000. Management <strong>of</strong> Spartina alterniflora.Bulletin <strong>of</strong> Oceanography 10:68-72.Ma, Z.J., B. Li, and K. Jing. 2003. Effects <strong>of</strong> tidewater on <strong>the</strong> feedingecology <strong>of</strong> hooded crane (Crane monacha) and conservation<strong>of</strong> <strong>the</strong>ir wintering habitats at Zhongming Dongtai, China. EcologicalResearch 18(3):321-329.Novak, S.J. and R.N. Mack. 2001. Tracing plant introduction andspread: genetic evidence from Bromus tectorun (cheatgrass).Bioscience 51(2): 114-122.Qian, Y.Q. and K.P. Ma. 1994. Bio-techniques, protection and wiseuses <strong>of</strong> biodiversity. In: Qian Y.Q. and K.P. Ma, eds. Principlesand Methods <strong>of</strong> Biodiversity Studies. China’s Science andTechnique Press, pp. 217-224.Qian, Y.Q. and K.P. Ma. 1995. Bio-techniques and bio-safety.Journal <strong>of</strong> Natural Resources 10(4):322-331.Qin, P. and C.H. Chung. 1992. Utilization <strong>of</strong> Spartina. Ocean Press,pp. 3-20.Sakai, A.K., F.W. Allendorf, J.S. Holt, et al. 2001. The population-35-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinabiology <strong>of</strong> invasive species. Annual Review <strong>of</strong> Ecology andSystematics 32:305-332.Shen, Y.M. and Y.M. Liu. 2002. Remote sensing <strong>of</strong> distributionchanges <strong>of</strong> Spartina alterniflora marshes in coastal Jiangsu.Journal <strong>of</strong> Plant Resources and Environments 11(2):33-38.Thompson, J.D. 1991. The biology <strong>of</strong> an invasive plant: what makesSpartina anglica so successful? Bioscience 41(6):393-400.Travis, S.E., C.E. Pr<strong>of</strong>fitt, R.C. Lowenfeld, et al. 2002. A comparativeassessment <strong>of</strong> genetic diversity among differently-agedpopulations <strong>of</strong> Spartina alterniflora on restored versus naturalwetlands. Restoration Ecology 10(1):37-42.Tsutsui, N.D., A.V. Suarez, D.A. Holway, et al. 2000. Reducedgenetic variation and <strong>the</strong> success <strong>of</strong> an invasive species. <strong>Proceedings</strong><strong>of</strong> <strong>the</strong> National Academy <strong>of</strong> Sciences <strong>of</strong> <strong>the</strong> UnitedStates <strong>of</strong> America 97(11):5948-5953.-36-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologySPARTINA DENSIFLORA X FOLIOSA HYBRIDS FOUND IN SAN FRANCISCO BAYD.R. Ayres 1 and A.K.F. Lee 2Dept. <strong>of</strong> Evolution and Ecology, University <strong>of</strong> California, Davis, One Shields Avenue, Davis, CA 956161 drayres@ucdavis.edu;2 alexkinlee@gmail.comKeywords: Invasive Spartina, hybridization, polyploidyINTRODUCTIONIn <strong>the</strong> 1970s, Spartina densiflora and S. foliosa wereplanted during <strong>the</strong> restoration <strong>of</strong> Creekside Park in MarinCounty to a tidal salt marsh. In 2001 we discoveredcordgrass plants that spread by rhizomes like S. foliosa, buthad dense, evergreen stems like S. densiflora. Spartinafoliosa, California cordgrass, is native to <strong>the</strong> state. Plantsgrow laterally by rhizomes, creating meadows <strong>of</strong> sparse,evenly-spaced, deciduous stems. The species occupies lowertidal environments (above mean sea level to mean highwater). Spartina densiflora, dense-flowered cordgrass, isnative to South America. Lack <strong>of</strong> rhizomes create a bunchtypegrass, with dense, largely evergreen stems. The speciesoccupies higher tidal areas than S. foliosa, occurring withSarcocornia viginica. The intermediate appearance <strong>of</strong> <strong>the</strong>Creekside Park plants suggested that <strong>the</strong> two species hadhybridized.MOLECULAR AND CYTOLOGICAL DYNAMICS OF S.DENSIFLORA X FOLIOSA HYBRIDSWe developed and used RAPD (Random AmplifiedPolymorphic DNA) nuclear DNA markers specific to ei<strong>the</strong>rS. foliosa or S. densiflora to identify and type hybrids (F1 orintrogressed). We used species-specific chloroplast DNAsequences (Anttila et al. 2000) to determine <strong>the</strong> maternalparent <strong>of</strong> hybrid plants, as chloroplasts are maternallyinherited in Spartina (Ferris et al. 1997). Chromosomenumbers in root tips were counted in <strong>the</strong> parental species andin seven hybrid plants. We estimated genome size in mostplants using flow cytometry (Grotkopp 2004; Galbraith1982), compared <strong>the</strong> genome sizes with <strong>the</strong> correspondingchromosome counts, and used genome size to rapidly assess<strong>the</strong> ploidy <strong>of</strong> hybrids (see Ayres et al. 2008 for details).We found 35 hybrid plants. All exhibited a F1 pattern<strong>of</strong> nuclear bands; that is, generally <strong>the</strong>y contained all 13 S.densiflora-specific bands and all nine S. foliosa-specificbands. A few plants lacked one or two bands. Most plants(17 out <strong>of</strong> 20 analyzed) had S. densiflora cpDNA. Mosthybrids were intermediate between S. densiflora and S.foliosa in chromosome number and genome size (Table 1);both chromosome number and genome size are consistentwith haploid gametes <strong>of</strong> each parental species (31 from S.foliosa + 35 from S. densiflora) uniting to form a F1 hybridwith 66 chromosomes. However, two plants were triploids,with <strong>the</strong> cp DNA <strong>of</strong> S. foliosa. Chromosome counts andgenome size assessments are consistent with a 2ncontribution by S. foliosa and a 1n contribution by S.densiflora, with loss <strong>of</strong> one and three chromosomes,respectively, in <strong>the</strong> two triploid individuals. Due tochromosomal mis-matching, viable gamete formation isprobably rare in all hybrids. Even so, <strong>the</strong> presence <strong>of</strong> triploidplants is important as it indicates that several avenues existTable 1. Molecular and cytological dynamics <strong>of</strong> S. densiflora x foliosa hybrids.Type/number <strong>of</strong> hybrids(RAPDs)S. foliosa S. densiflora 2n hybrids 3n hybrids33 F1 2 F1cp DNA Sf Sd 17 Sd: 1 Sf 2 Sf : 0 SdChromosome number 62 70 65/66 94/96Genome size- pg (SD) 4.46 (SD= 0.10) 5.16 (SD = 0.06) 4.83 (SD = 0.06) 7.0 (SD = 0.01)Chromosome math 31 (S.f. 1n) + 35 (S.d. 1n) = 66 (S.d x f 2n)Triploid math 62 (S.f. 2n) + 35 (S.d. 1n) - (1 or 3 chromosomes) = 94 or 96 (S.d x f 3n)Genome size math {0.5 *4.46 (S.f. 2n)} + {0.5 *5.16 (S.d. 2n)} = 4.81 pg (S.d x f 2n)Triploid math {1*4.46 (S. f. 2n)} + {0.5 *5.16 (S.d. 2n)} = 7.0 pg (S.d x f 3n)-37-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaFig. 1. Final aboveground biomass <strong>of</strong> some plants in greenhouse salinityexperiment. CS-17 and CS-19 (diploid plants) are some <strong>of</strong> <strong>the</strong> transgressivehybrids, while CS-15 (triploid plant) has performance comparable to <strong>the</strong>parental species under <strong>the</strong> same high salinity stress.which may give rise to a new alloployploid species. Thus,this new hybridization possibly <strong>of</strong>fers us a chance to observe<strong>the</strong> origin <strong>of</strong> a new species.ECOLOGY OF S. DENSIFLORA X FOLIOSA HYBRIDSThis is <strong>the</strong> second Spartina hybridization in <strong>the</strong> SanFrancisco estuary in <strong>the</strong> past three decades; <strong>the</strong> o<strong>the</strong>r,between S. alterniflora x foliosa, has resulted in abackcrossing swarm <strong>of</strong> invasive hybrids. Introductions <strong>of</strong>Spartina have resulted in major biological invasions in saltmarshes around <strong>the</strong> world. Two <strong>of</strong> <strong>the</strong>se invaders arehybrids between native and introduced species (S.alterniflora x foliosa; S. anglica). Given <strong>the</strong> history <strong>of</strong>Spartina invasions and hybridizations, we investigatedwhe<strong>the</strong>r S. densiflora x foliosa hybrids have <strong>the</strong> potential tospread and invade surrounding marshes in <strong>the</strong> Bay. Thiswould require that hybrids tolerate marsh salinity and tidalinundation, and produce viable seed. A greenhouseexperiment was performed with ten hybrid genotypes toassess <strong>the</strong>ir salinity tolerance against <strong>the</strong> parental species (S.foliosa – low elevation, low salinity; and S. densiflora –higher elevation, higher salinity). Salinity was increased by10 parts per thousand per week (ppt/week), and wemeasured several fitness indicators for 10 weeks. In <strong>the</strong>field, we combined mapping data with elevationalmeasurements (using a Trimble Total Station) to determine<strong>the</strong> relative tidal elevations <strong>of</strong> hybrids, native species, and S.densiflora. We also collected inflorescences from <strong>the</strong> fieldto measure <strong>the</strong>ir reproduction (seed set).We found that some hybrids have high salinity tolerancebased on final aboveground biomass and flower productionat <strong>the</strong> end <strong>of</strong> 10 weeks (Fig. 1). Tolerance to high salinitycould allow hybrids to grow in higher marsh environments,largely occupied by Sarcocornia virginica. Hybrids had anelevational range similar to S. densiflora at Creekside Park.Spartina foliosa occupies <strong>the</strong> lowest range, while <strong>the</strong>dominant Sarcocornia virginica is found in <strong>the</strong> highest. Thehybrids occupy <strong>the</strong> same “middle” range as <strong>the</strong>ir S.densiflora parent, which suggests that <strong>the</strong> hybrids might nottolerate tidal inundation as well as <strong>the</strong>ir S. foliosa parent.Finally, hybrids set no seed.SUMMARY AND CONCLUSIONSWe found that <strong>the</strong> salinity tolerance <strong>of</strong> some <strong>of</strong> hybridgenotypes exceeded that <strong>of</strong> both parental species; thathybrids occurred higher in <strong>the</strong> marsh than S. foliosa andwithin a narrower elevation range than ei<strong>the</strong>r parentalspecies; and hybrids were apparently sterile as <strong>the</strong>yproduced no seed in <strong>the</strong> field and produced only shriveledan<strong>the</strong>rs in <strong>the</strong> greenhouse experiment. We conclude thatdespite hybrid superiority in salinity tolerance, S. densiflorax foliosa hybrids will not be invasive due to sterility. Evenso, sterility may be overcome if fertile tetraploids orhexaploids evolve from hybrid individuals.ACKNOWLEDGMENTSWe would like to thank Krista Callinan (field andgreenhouse experiment), Carina Anttila (cpDNA), JohnBailey (cytology), Eva Grotkopp (flow cytometry), PabloRosso (GIS) and acknowledge <strong>the</strong> financial support from <strong>the</strong>California Coastal Conservancy (CalFed grant #99-110),California Sea Grant #27CN to D.R. Strong, and NSFBiocomplexity DEB 0083583 to A. Hastings and D.R.Strong.REFERENCESAnttila, C.K., A.R. King, C. Ferris, D.R. Ayres and D.R. Strong.2000. Reciprocal hybrid formation <strong>of</strong> Spartina in San FranciscoBay. Molecular Ecology 9: 765-771.Ayres, D.R., E. Grotkopp, K. Zaremba, et al. 2008. Hybridizationbetween invasive Spartina densiflora (Poaceae) and native S. foliosain San Francisco Bay, California, USA. American Journal<strong>of</strong> Botany 95(6): 713–719.Ayres, D.R., D. Garcia-Rossi, H.G. Davis and D.R. Strong. 1999.Extent and degree <strong>of</strong> hybridization between exotic (Spartina alterniflora)and native (S. foliosa) cordgrass (Poaceae) in California,USA determined by random amplified polymorphic DNA(RAPDs). Molecular Ecology 8: 1179-1186.Ferris, C., R.A. King and A.J. Gray. 1997. Molecular evidence for<strong>the</strong> maternal parentage in <strong>the</strong> hybrid origin <strong>of</strong> Spartina anglicaC.E. Hubbard. Molecular Ecology 6: 185-187.Grotkopp, E., M. Rejmanek, M.J. Sanderson and T.L. Rost. 2004.Evolution <strong>of</strong> genome size in pines (Pinus) and its life-history correlates:supertree analyses. Evolution 58:1705-1729.Galbraith, D.W., K.R. Harkins, J.M. Maddox, N.M. Ayres, D.P.Sharma and E. Firoozabady. 1983. Rapid flow cytometric analysis<strong>of</strong> <strong>the</strong> cell cycle in intact plant tissues. Science 220: 1049-1051.-38-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyFUNGAL SYMBIOSIS:APOTENTIAL MECHANISM OF PLANT INVASIVENESSR.J. RODRIGUEZ 1 ,R.S.REDMAN 2,3 ,M.HOY 2 , AND N. ELDER 41 U.S. Geological Survey, 6505 NE 65th Street, Seattle, WA 98115; Rusty_Rodriguez@usgs.gov2 Department <strong>of</strong> Biology, University <strong>of</strong> Washington, Seattle, WA 981953 Department <strong>of</strong> Microbiology, Montana State University, Bozeman, MT 597174 U.S. Geological Survey, Marrowstone Station, Nordland, WA 98358We propose that fungal endophytes provide a mechanism for <strong>the</strong> habitat expansion <strong>of</strong> invasiveplants including Spartina species. We have determined that Spartina spp. are symbiotic wi<strong>the</strong>ndophytic fungi and have begun to assess <strong>the</strong> role <strong>of</strong> fungal endophytes in <strong>the</strong> invasiveness <strong>of</strong> <strong>the</strong>seplants. Preliminary data indicate that <strong>the</strong> fungal endophytes in Spartina spp. in Puget Sound arenative to that region. The role <strong>of</strong> symbiosis in <strong>the</strong> invasion <strong>of</strong> invasive species and <strong>the</strong> adaptation <strong>of</strong>plants to high stress habitats is discussed.Keywords: symbiosis, fungi, lifestyle, plant ecology, fungal ecology, Colletotrichum, Curvularia,SpartinaINTRODUCTIONIt is estimated that thousands <strong>of</strong> plants are introducedinto non-native habitats every year, however, only a smallpercentage <strong>of</strong> <strong>the</strong>se plants become invasive. Interestingly,invasions may result from intercontinental orintracontinental plant movements and may involve ei<strong>the</strong>rvast or very short distances. Several <strong>the</strong>ories have beenproposed to explain <strong>the</strong> invasiveness <strong>of</strong> plants (Daehler2003; Callaway and Ridenour 2004). However, noneadequately explain why plants can achieve high densities innew habitats (Daehler, 2003; Silvertown 2004). It doesappear that invasion is context dependent and may involve anumber <strong>of</strong> biotic and abiotic factors (Daehler 2003).Although <strong>the</strong>re have been numerous studies on invasiveplants, few include fungal symbiosis as a component <strong>of</strong> plantbiology. Yet, all plants in natural ecosystems are thought tobe symbiotic with mycorrhizal and/or endophytic fungi.These fungi differ in distribution, biology and physiology.Mycorrhizal fungi are limited to colonizing roots and growout into <strong>the</strong> rhizosphere effectively expanding root systemsby transporting nutrients and water that would o<strong>the</strong>rwise beunavaiable to root systems (Read 1999).Endophytic fungi reside entirely within plant tissues andmay occur in specific tissues (roots, crowns, stems, leaves,seed coats or seeds) or throughout <strong>the</strong> plant. Althoughmycorrhizal fungi do not colonize all plants, it is thoughtthat all plants are colonized by endophytic fungi. Fungalendophytes can be divided into two major groups(Rodriguez et al. 2009b): 1) a small number <strong>of</strong> fastidiousspecies that are restricted to a small number <strong>of</strong> monocothosts (Clay and Schardl 2002) and 2) a large number <strong>of</strong>tractable species with broad host ranges (Stone et al. 2004).Both groups <strong>of</strong> fungal endophytes are known to be importantto <strong>the</strong> structure, function, and health <strong>of</strong> plant communities.In fact, without fungal symbioses, plant communities do notappear to survive many environmental stresses (Arnold et al.2003; Dingle and McGee 2003; Ernst et al., 2003; Redmanet al. 2002b).FUNGAL SYMBIOTIC LIFESTYLESCollectively, fungal symbionts express a variety <strong>of</strong>symbiotic lifestyles including mutualism, commensalism,and parasitism. These lifestyles are based on positive,neutral or negative fitness benefits to <strong>the</strong> host and symbiont(Table 1, Lewis 1985). Mutualistic fungi have been shown toincrease plant growth and productivity (Marks and Clay1996; Varma et al. 1999; Read 1999; Redman et al. 2002a;Rodriguez et al. 2009a), and confer tolerance to abiotic andbiotic stresses including drought, metals, salt, temperature,pathogens and herbivores (Bacon and Hill 1996; Bacon1993; Read 1999; Carroll 1986; Redman et al. 2001;Redman et al. 2002b; Latch 1993; Rodriguez et al. 2008).We have demonstrated that endophytic fungi also have <strong>the</strong>ability to switch lifestyles in response to host genotypes andthat <strong>the</strong> host range <strong>of</strong> fungal endophytes is typically greaterthan previously thought (Redman et al. 2001). For example,fungi from <strong>the</strong> genus Colletotrichum are classified as plantpathogenic fungi. However, several species have <strong>the</strong> abilityto asymptomatically colonize plants not previously known tobe hosts and express non-pathogenic lifestyles includingmutualism. When Colletotrichum species expressmutualistic lifesytles <strong>the</strong>y confer disease resistance againstfungi that express pathogneic lifestyles in <strong>the</strong> respective hostplant. Therefore, one fungal isolate may be a virulentpathogen on one host and a disease protecting mutualist onano<strong>the</strong>r.Symbiotically conferred disease resistance is correlatedwith <strong>the</strong> activation <strong>of</strong> host defense systems (Redman et al.1999). When exposed to virulent pathogens, non-symbioticplants slowly activated defense systems over a four-dayperiod and by day five <strong>the</strong> plants were dead. Symbioticplants activated defense systems within 24 hours <strong>of</strong> exposureto a virulent pathogen and completely terminated pathogen-39-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaTable 1. Fitness impacts <strong>of</strong> Symbiotic LifestylesLifestyleFitness ImpactHost SymbiontParasitism - +Commensalism 0 +Mutualism + +- = fitness decreased, + = fitness increased,0 = fitness not affectedingress. It appears that when Colletotrichum species areexpressing non-pathogenic lifestyles <strong>the</strong>y are acting asbiological triggers allowing plants to recognize virulentpathogens more quickly (Rodriguez et al. 2004). Theeffectiveness <strong>of</strong> an endophyte as a biological trigger appearsto be a function <strong>of</strong> currently undefined genetic compatibilitywith <strong>the</strong> host.ADAPTIVE SYMBIOSISMutualistic fungi are known to confer a variety <strong>of</strong>fitness benefits. However, it is not known if symbioticallyconferred stress tolerance reflects an adaptive response by<strong>the</strong> host and/or symbiont. We have studied <strong>the</strong> ecologicalsignificance <strong>of</strong> endophytic fungi in plants thriving ingeo<strong>the</strong>rmal soils and found that <strong>the</strong> symbiosis is responsiblefor <strong>the</strong> ability <strong>of</strong> both host and symbiont to survive <strong>the</strong>rmalstress (Redman et al. 2002b; Marquez et al. 2007). In over200 individuals analyzed, <strong>the</strong> plant Dichan<strong>the</strong>liumlanuginosum is colonized by one fungal species (Curvulariaprotuberata) that is known to be a pathogen in o<strong>the</strong>r plantspecies (Farr et al. 1989). The fungal endophyte colonizesroots, crowns, stems, leaves and seed coats but not seeds.When seed coats are removed and seeds’ surfaces aresterilized it is possible to propagate plants devoid <strong>of</strong> <strong>the</strong>fungus. This approach allowed us to compare <strong>the</strong> ability <strong>of</strong>symbiotic and non-symbiotic plants to mitigate <strong>the</strong> impacts<strong>of</strong> <strong>the</strong>rmal stress. Nei<strong>the</strong>r <strong>of</strong> <strong>the</strong> symbiotic partners toleratedtemperatures above 40°C, however, <strong>the</strong> symbiosis allowedboth to survive root temperatures up to 70°C. Recently wehave observed that isolates <strong>of</strong> C. protuberata from nongeo<strong>the</strong>rmalplants do not confer temperature tolerancesuggesting that this is an adaptive response by fungi in <strong>the</strong>geo<strong>the</strong>rmal habitat (Rodriguez et al. 2008).Comparative studies with Colletotrichum andCurvularia species support <strong>the</strong> hypo<strong>the</strong>sis that symbiosesadapt to habitat stresses (Rodriguez et al. 2008). These fungiwere assessed for <strong>the</strong> ability to confer heat tolerance (anabiotic stress in geo<strong>the</strong>rmal habitats) and disease resistance(a biotic stress in agricultural habitats) to host plants.Colletotrichum isolates expressing non-pathogenic lifestylesconferred disease resistance but not heat tolerance.Conversely, Curvularia isolates conferred heat tolerance butnot disease resistance. This observation suggests that at leastsome <strong>of</strong> <strong>the</strong> benefits conferred by mutualistic endophytesappear to reflect habitat-specific adaptation.Colletotrichum species that express a mutualisticlifestyle typically are able to asymptomatically colonizegenetically divergent plant species (Redman et al. 2001). Forexample, C. magna is a virulent pathogen on cucurbits but isa disease protecting mutualist on tomato (Lycopersiconesculentum), beans (Phaseolus vulgaris) and strawberry(Fragaria ananassa), three genetically divergent species.Therefore, it appears that genetically distant plant speciescan gain novel biological function simply by formingsymbioses with different fungi. This could allow individualplant species to make quantum evolutionary leaps andexpand into new habitats or become dominant members <strong>of</strong>existing communities.SYMBIOSIS AS A POTENTIAL MECHANISM OFINVASIVENESSBased on <strong>the</strong> broad host range <strong>of</strong> class 2 fungalendophytes, <strong>the</strong>ir ability to confer abiotic and biotic stresstolerance, and <strong>the</strong> apparent adaptive nature <strong>of</strong> symbioses, wepropose <strong>the</strong> following hypo<strong>the</strong>sis: The conversion <strong>of</strong> at leastsome non-native plants into invasive species requires <strong>the</strong>establishment <strong>of</strong> symbioses with endophytic fungi thatconfer tolerance to abiotic and biotic stresses. Thesesymbioses allow invasive species to circumvent <strong>the</strong>ecological factors that keep native species under control andplant communities balanced. In order to properly challengethis hypo<strong>the</strong>sis (designated <strong>the</strong> “symbiotic communicationhypo<strong>the</strong>sis”) several questions must be addressed such as:Do non-native species carry non-native endophytes whentransported? Are <strong>the</strong> non-native endophytes capable <strong>of</strong>releasing non-native plants from ecological controls? If <strong>the</strong>non-native plants are devoid <strong>of</strong> endophytes what are <strong>the</strong>dynamics <strong>of</strong> colonization by native fungi? If an adaptiveresponse is required before non-native plants becomeinvasive, what are <strong>the</strong> temporal dynamics <strong>of</strong> adaptation?Ano<strong>the</strong>r mystery <strong>of</strong> plant invasions is that <strong>the</strong>re is a timelag time between introduction and spread which may bemore than 100 years (Binggeli 2000). Several hypo<strong>the</strong>seshave been proposed to explain invasive lag time but nonehave adequately addressed this phenomenon. We havedemonstrated that subtle differences in host genotypes aresufficient to eliminate symbiotically conferred stresstolerance and may result in <strong>the</strong> expression <strong>of</strong> pathogenicra<strong>the</strong>r than mutualistic lifestyles by endophytic fungi(Redman et al. 2001). Therefore, we propose that <strong>the</strong> lagphase commonly observed between introduction andinvasion may reflect one or more <strong>of</strong> <strong>the</strong> following:1) Endophytes carried with non-native plants adapt tonew habitat stresses and confer stress tolerance allowingnon-native plants to outcompete native plants;-40-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina Biology2) Non-native plants establish symbioses with nativefungi that communicate more efficiently than with nativeplants;3) Non-native plants bring fungi that switch lifestylesand are pathogenic on native plants;4) Non-native plants alter <strong>the</strong> rhizosphere such thatnative root pathogens normally present in low abundancebecome abundant and are more virulent on native plants than<strong>the</strong> non-native species.To begin testing <strong>the</strong> symbiotic communicationhypo<strong>the</strong>sis we have begun to assess <strong>the</strong> fungal endophytes inseveral native and invasive plant species in Puget Sound andnative habitats <strong>of</strong> <strong>the</strong> invasive species. The list <strong>of</strong> plantspecies we are analyzing includes Spartina anglica, S.alterniflora and S. patens. A minimum <strong>of</strong> three populationsfor each plant species was analyzed and at least 10 plants/perpopulation were sampled. All <strong>of</strong> <strong>the</strong> Spartina plantsanalyzed (N=100) contained class 2 endophytes whichappear to be native fungi based on a biogeographic study(Rodriguez et al. in preparation). These endophytes arecurrently being taxonomically defined by sequence analysis<strong>of</strong> nuclear ribosomal DNA and classical morphologicalmethods. The host range <strong>of</strong> <strong>the</strong> endophytes is also beingassessed and culture conditions identified for fungalconidiation. In addition, we have developed methods togerminate Spartina seeds devoid <strong>of</strong> endophytic fungi and arebeginning to perform comparative studies to evaluatesymbiotic and non-symbiotic plants. In 2005-2007 weperformed greenhouse experiments to understand <strong>the</strong>symbiotic dynamics <strong>of</strong> Spartina anglica in <strong>the</strong> presence andabsence <strong>of</strong> abiotic stress. These studies indicated thatSpartina, like native coastal plants, establish associationswith endophytes that confer salt tolerance Rodriguez et al.2008; Rodriguez et al., in preparation). Therefore, wesurmise that endophytes play an important role in <strong>the</strong>invasion <strong>of</strong> coastal areas by Spartina species. Studies areunderway to determine if endophytes contribute to <strong>the</strong>invasion processes in non-coastal habitats that imposedifferent degrees <strong>of</strong> abiotic stress.ACKNOWLEDGMENTSWe would like to thank <strong>the</strong> organizers <strong>of</strong> <strong>the</strong><strong>International</strong> Spartina <strong>Conference</strong> for <strong>the</strong> opportunity topresent <strong>the</strong>se hypo<strong>the</strong>ses and submit this manuscript. Wewould like to thank Edward Pr<strong>of</strong>fitt, David Heimer andRoger Fuller for assistance in collecting Spartina samples.This work was funded in part by grants from <strong>the</strong> U.S.Geological Survey and <strong>the</strong> National Science Foundation.REFERENCESArnold, E.A., L.C. Mejia, D. Kyllo, E. Rojas, Z. Maynard,N. Robbins, and E.A. Herre. 2003. Fungal endophyteslimit pathogen damage in a tropical tree. <strong>Proceedings</strong> <strong>of</strong><strong>the</strong> National Academy <strong>of</strong> Sciences 100: 15649-15654.Bacon, C.W. 1993. Abiotic stress tolerances (moisture, nutrients)and photosyn<strong>the</strong>sis in endophyte-infected tall fescue.Agriculture Ecosystems and Environment 44: 123-141.Bacon, C. W., and N.S. Hill. 1996. Symptomless grassendophytes: products <strong>of</strong> coevolutionary symbioses and<strong>the</strong>ir role in <strong>the</strong> ecological adaptations <strong>of</strong> grasses. In: Redkin,S.C. and L.M. Carris, eds. Endophytic fungi in grassesand woody plants. St. Paul, APS Press, pp. 155-178.Binggeli, P. 2000. Time-lags between introduction, establishmentand rapid spread <strong>of</strong> introduced environmentalweeds. <strong>Third</strong> <strong>International</strong> Weed Science Congress, Fozdo Iguassu, Brazil, Manuscript number 8, pp. 14.Callaway, R.M., and W.M. Ridenour. 2004. Novel weapons:invasive success and <strong>the</strong> evolution <strong>of</strong> increased competitiveability. Frontiers in Ecology and <strong>the</strong> Environment 2:436-443.Carroll, G.C. 1986. The biology <strong>of</strong> endophytism in plantswith particular reference to woody perennials. In: Fokkema,N.J. and J. Van Den Heuvel, eds. Microbiology <strong>of</strong><strong>the</strong> phyllosphere. Cambridge University Press, Cambridge,pp. 205-222.Clay, K., and C. Schardl. 2002. Evolutionary origins andecological consequences <strong>of</strong> endophyte symbiosis withgrasses. The American Naturalist 160: Supplement, S99-S127.Daehler, C.C. 2003. Performance comparisons <strong>of</strong> cooccurringnative and alien invasive plants: Implications forconservation and restoration. Annual Review Ecolology,Evolution and Systematics 34: 183-211.Dingle, J., and P.A. McGee. 2003. Some endophytic fungireduce <strong>the</strong> density <strong>of</strong> pustules <strong>of</strong> Puccinia recondita f. sp.tritici in wheat. Mycological Research 107: 310-316.Ernst, M., K.W. Mendgen, and S.G.R. Wirsel. 2003. Endophyticfungal mutualists: seed-borne Stagonospora spp.enhance reed biomass production in axenic microcosms.Molecular Plant Microbe Interactions 16: 580-587.Farr, D.F., G.F. Bills, G.P. Chamuris, and A.Y. Rossman.1989. Fungi on plants and plant products in <strong>the</strong> UnitedStates. APS Press, St. Paul, MN, pp. 1252.Latch, G.C.M. 1993. Physiological interactions <strong>of</strong> endophyticfungi and <strong>the</strong>ir hosts. Biotic stress tolerance impartedto grasses by endophytes. Agriculture, Ecosystemsand Environment 44: 143-156.Lewis, D.H. 1985. Symbiosis and mutualism: Crisp conceptsand soggy semantics. In: D.H. Boucher, ed. The Biology <strong>of</strong>Mutualism. London, Croom Helm Ltd., pp. 29-39.Marks, S., and K. Clay. 1990. Effects <strong>of</strong> CO 2 enrichment,nutrient addition, and fungal endophyte-infection on <strong>the</strong>growth <strong>of</strong> two grasses. Oecologia 84: 207-214.Marquez, L.M., R.S. Redman, R.J. Rodriquez and M.J.Roossinck. 2007. A virus in a fungus in a plant – threeway symbiosis required for <strong>the</strong>rmal tolerance. Science315:513-515.Read, D.J. 1999. Mycorrhiza - <strong>the</strong> state <strong>of</strong> <strong>the</strong> art. In: Varma,A. and B. Hock, eds. Mycorrhiza. Berlin, Springer-Verlag,pp. 3-34.Redman, R.S., D.D. Dunigan, and R.J. Rodriguez. 2001.Fungal symbiosis: from mutualism to parasitism, who controls<strong>the</strong> outcome, host or invader? New Phytologist 151:705-716.-41-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaRedman, R.S., S. Freeman, D.R. Clifton, J. Morrel, G.Brown, and R.J. Rodriguez. 1999. Biochemical analysis <strong>of</strong>plant protection afforded by a nonpathogenic endophyticmutant <strong>of</strong> Colletotrichum magna. Plant Physiology 119:795-804.Redman, R.S., M.J. Roossinck, S. Maher, Q.C. Andrews,W.L. Schneider, and R.J. Rodriguez. 2002a. Field performance<strong>of</strong> cucurbit and tomato plants colonized with anonpathogenic mutant <strong>of</strong> Colletotrichum magna (teleomorph:Glomerella magna; Jenkins and Winstead). Symbiosis32: 55-70.Redman, R.S., K.B. Sheehan, R.G. Stout, R.J. Rodriguez,and J.M. Henson. 2002b. Thermotolerance Conferred toPlant Host and Fungal Endophyte During MutualisticSymbiosis. Science 298: 1581.Rodriguez, R.J., D.C. Freeman, E.D. McArthur, Y.O. Kim,and R.S. Redman. 2009a. Symbiotic regulation <strong>of</strong> plantgrowth, development and reproduction. Communicative &Integrative Biology 2:141-143.Rodriguez R.J., J. Henson, E. Van Volkenburgh, M. Hoy, L.Wright, F. Beckwith, Y. Kim, and R.S. Redman. 2008.Stress Tolerance in Plants via Habitat-Adapted Symbiosis.<strong>International</strong> Society <strong>of</strong> Microbial Ecology 2: 404–416.Rodriguez, R.J., R.S. Redman, and J.M. Henson. 2004. TheRole <strong>of</strong> Fungal Symbioses in <strong>the</strong> Adaptation <strong>of</strong> Plants toHigh Stress Environments. Mitigation and AdaptationStrategies for Global Change 9: 261-272.Rodriguez, R.J., J.F. White, Jr., A.E. Arnold, and R.S.Redman. 2009b. Fungal endophytes: diversity and functionalroles. Tansley Review, New Phytologist 182:314-330.Silvertown, J. 2004. Plant coexistence and <strong>the</strong> niche. Trendsin Ecology and Evolution 19: 605-611.Stone, J.K., J.D. Polishook, and J.F. White, Jr. 2004. EndophyticFungi. In: Mueller, G.M., G.F. Bills, and M.S. Foster,eds. Biodiversity <strong>of</strong> Fungi: Inventory and MonitoringMethods. Elsevier Academic Press, Oxford, U.K. pp. 241-270.Varma, A., S. Verma, Sudha, N. Sahay, B. Butehorn, and P.Franken, 1999. Piriformospora indica, a cultivable plantgrowth-promotingroot endophyte. Applied and EnvironmentalMicrobiology 65: 2741-2744.-42-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyIS ERGOT A NATURAL COMPONENT OF SPARTINA MARSHES?DISTRIBUTION ANDECOLOGICAL HOST RANGE OF SALT MARSH CLAVICEPS PURPUREAA.J. FISHERDepartment <strong>of</strong> Plant Pathology, University <strong>of</strong> California, Davis, One Shields Ave., Davis, CA 95616; alijfisher@gmail.comThis study was undertaken to better characterize <strong>the</strong> geographic distribution and host range <strong>of</strong>Claviceps purpurea from grass hosts in salt marsh habitats. Claviceps purpurea contains threeintraspecific groups, each with a habitat association: G1 from terrestrial habitats, G2 from moisthabitats and G3 from salt marshes. Twenty-six G3 isolates, representing 11 distinct populationswere characterized based on <strong>the</strong> presence <strong>of</strong> an EcoRI restriction site in <strong>the</strong> 5.8S ribosomal DNA,and genetic similarity to isolates representing <strong>the</strong> o<strong>the</strong>r two C. purpurea intraspecific groups (G1and G2). Distichlis spicata was identified as <strong>the</strong> first non-Spartina host to G3 C. purpurea. Inaddition, all isolates originating from Spartina densiflora, S. foliosa, S. alterniflora, and S. anglicawere identified as belonging to G3 based on genetic analysis. Salt marsh Claviceps purpurea can befound along <strong>the</strong> Atlantic and Pacific Coasts <strong>of</strong> North America, Argentina, Ireland and England. Aglobal distribution suggests that salt marsh (G3) C. purpurea is a natural component <strong>of</strong> Spartinadominatedsalt marsh habitats.Keywords: Ergot, Claviceps purpurea, Spartina, host rangeINTRODUCTIONClaviceps purpurea (Fr.) Tul, <strong>the</strong> cause <strong>of</strong> ergot, is awell known pathogen <strong>of</strong> cereal grains and forage. Thepathogen has a global distribution, and a wide host rangewithin <strong>the</strong> Poaceae. This broad host range has led manyresearchers to seek evidence <strong>of</strong> taxonomic substructurebased on morphology (Loveless 1971), alkaloid production(Eleuterius and Meyers 1977; Kobel and Sanglier 1978), andgenetic characteristics (Jungehülsing and Tudzynski 1997).In a recent study, Pazoutová et al. (2000) syn<strong>the</strong>sizedprevious research on C. purpurea morphology, alkaloidchemistry and genetics, and identified three distinct groupswithin <strong>the</strong> species. Ra<strong>the</strong>r than divisions based on hostrange, <strong>the</strong> groups were defined by habitat association. Thelargest group, G1, was associated with terrestrial grasses,and G2 with grasses in moist environments, whereas G3 wasfound only in salt marsh habitats. Isolates associated withG1 and G3 share an EcoRI restriction site in <strong>the</strong> 5.8Sribosomal DNA (rDNA), which G2 isolates lack. The threegroups can be differentiated by RAPD analysis as well(Pazoutová et al. 2000).The ecological host range <strong>of</strong> maritime C. purpurea (G3)appears very narrow, compared to G1 and G2. G3 has beenisolated exclusively from cordgrass species <strong>of</strong> <strong>the</strong> genusSpartina. Two host species to G3 have been identified, eachin only one location: common cordgrass, S. anglica, in <strong>the</strong>United Kingdom and smooth cordgrass, S. alterniflora, from<strong>the</strong> state <strong>of</strong> New Jersey on <strong>the</strong> Atlantic coast <strong>of</strong> <strong>the</strong> UnitedStates (Pazoutová et al. 2000, 2002b).Whereas only two populations <strong>of</strong> G3 have beenidentified using molecular markers, <strong>the</strong> host to thisintraspecific group, Spartina, is a widespread genusincluding both terrestrial and halophytic salt marsh species.Most members <strong>of</strong> <strong>the</strong> genus, including S. alterniflora, sharea common native geographic range along <strong>the</strong> Atlantic Coast<strong>of</strong> North and South America. Introduced populations <strong>of</strong> S.alterniflora have established on <strong>the</strong> Pacific Coast as well, in<strong>the</strong> San Francisco Bay, California (Daehler and Strong 1994)and Willapa Bay, Washington (Stiller and Denton 1995).Thus, <strong>the</strong> geographic range <strong>of</strong> G3 may be much greater thanis presently recognized and it is reasonable to think thatdiversity in populations <strong>of</strong> <strong>the</strong> pathogen may reflect <strong>the</strong>diversity <strong>of</strong> <strong>the</strong> host species. The objectives <strong>of</strong> this studywere to assess <strong>the</strong> geographic distribution and ecologicalhost range <strong>of</strong> G3 C. purpurea.MATERIALS AND METHODSIsolates included in this study, and <strong>the</strong>ir origins, arelisted in Table 1. Isolates were sterilized and culturedaccording to <strong>the</strong> methods described in Pazoutová et al.(2000). Isolates not collected in <strong>the</strong> field were obtained aspure cultures from S. Pazoutová, including G1 isolates 374and 428 (Pazoutová et al. 2000), G2 isolates 236 and 434(Pazoutová et al. 2000), G3 isolates 500 and 538 (Pazoutováet al. 2002b), and G1 isolates165, 204 and 478. DNA wasextracted, and RAPD and AFLP markers were developedaccording to <strong>the</strong> methods described in Pazoutová et al.(2000). EcoRI analysis was also performed using <strong>the</strong>methods described in Pazoutová et al. (2000). Allmonomorphic and polymorphic RAPD and AFLP markerswere scored using a binary system (zero = absent and one =present). Percent similarity between isolates was calculatedby dividing <strong>the</strong> number <strong>of</strong> shared markers by <strong>the</strong> totalnumber <strong>of</strong> markers and multiplying by 100.-43-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaRESULTSDistichlis spicata was identified as <strong>the</strong> first non-Spartina host to G3 C. purpurea. In addition, all isolatesoriginating from Spartina densiflora, S. foliosa, S.alterniflora, and S. anglica were identified as belonging toG3 based on genetic analysis. G1 and G3 isolates containedan EcoRI restriction site within <strong>the</strong> 5.8S rDNA, while G2isolates lacked this site, consistent with distinctions among<strong>the</strong>se groups described by Pazoutová et al. (2000). UsingRAPD and AFLP markers, between-group diversity wasextraordinarily high with only less than 5 % <strong>of</strong> markersshared by all members <strong>of</strong> G1, G2 and G3. Within-groupsimilarity for G3 isolates was approximately 40%.DISCUSSIONThe maritime group <strong>of</strong> C. purpurea (G3) has beendistinguished from o<strong>the</strong>r intraspecific groups in this speciesby having an EcoRI site in <strong>the</strong> 5.8S ribosomal DNA, and aunique banding pattern based on RAPD analysis <strong>of</strong> nuclearDNA. Based on <strong>the</strong>se characters, and AFLP analysis, <strong>the</strong>field collections examined in this study from Spartina spp.and Distichlis spicata are representative <strong>of</strong> <strong>the</strong> maritimegroup. The results <strong>of</strong> this study revealed G3 C. purpurea onthree new species, S. foliosa, S. densiflora, and D. spicata.In addition, this study has identified salt marsh (G3) C.purpurea on S. alterniflora where it is invasive: in <strong>the</strong> SanFrancisco Bay, CA and Willapa Bay, WA. Spartinaalterniflora was intentionally introduced into <strong>the</strong> SanFrancisco Bay as seed in <strong>the</strong> 1970s; S. alterniflora wasunintentionally introduced to Willapa Bay, Washington overa century ago (Stiller and Denton 1995; Civille et al. 2005).The impacts <strong>of</strong> C. purpurea on S. foliosa populations are notknown but <strong>the</strong>re is evidence that <strong>the</strong> pathogen can reduceseed production in this species (A. Fisher, 2007). In WillapaBay, <strong>the</strong> impacts <strong>of</strong> C. purpurea on seed production are notknown. However, rates <strong>of</strong> infection are so low, that even if<strong>the</strong> pathogen reduces seed production on individualinfloresences, C. purpurea is unlikely to be a significantfactor for S. alterniflora fecundity.Spartina anglica, host to maritime C. purpurea on <strong>the</strong>coast <strong>of</strong> <strong>the</strong> United Kingdom and now identified as a host inIreland, has a hybrid origin, being derived from native S.maritima and introduced S. alterniflora (Raybould et al.1991). The first incidence <strong>of</strong> C. purpurea on S. anglica andS. townsendii, <strong>the</strong> sterile F1 hybrid <strong>of</strong> S. maritima and S.alterniflora, was reported in Ireland in 1975 (Boyle 1976).While unsure <strong>of</strong> its origin, Boyle argued that this and aherbarium specimen <strong>of</strong> C. purpurea on S. anglica collectedin 1971 were <strong>the</strong> first definitive reports <strong>of</strong> C. purpurea onsalt marsh Spartina spp. in Great Britain and Ireland.Distichlis spicata has been previously identified as ahost to C. purpurea (Sprague 1950), and is currently <strong>the</strong>only known host to G3 outside <strong>the</strong> genus Spartina. Perhapsnot surprisingly, Distichlis and Spartina are closely relatedgenera, belonging to <strong>the</strong> same subfamily (Chloridoideae)Table 1. Origin and host plant <strong>of</strong> C. purpurea isolates.a GroupOriginHostIDBřezno, Czech Republic Dactylis glomerata G1Altamont, AL, USA F. arundinacea G1Lauderdale, AL, USAFestucaarundinaceaG1L’Anse aux meadows,Newfoundland, CanadaLeymus mollis G1Zubri, Czech Republic Poa pratensis G1MacDoel, CA, USA Secale cereale G1Aberdeen, ID, USA Secale cereale G1Hohenheim, Germany Secale cereale G1Phillipsreuth, Germany Dactylis spp. G2Vlčí Pole u Bousova,Czech RepublicMolinia coerulea G2Willapa River, WA, USA Distichlis spicata G3Palix River, WA, USA S. alterniflora G3Dolphin Island, AL, USA S. alterniflora G3Point Reyes NS, CA,USAS. alterniflora G3St. Augustine, FL, USA S. alterniflora G3Marsh Landing, GA, USA S. alterniflora G3Flax River, NY, USA S. alterniflora G3Rhode Island, USA S. alterniflora G3Southriver, NJ, USA S. alterniflora G3Marchwood, UK S. alterniflora G3Dublin, Ireland S. anglica G3Argentina Celpa Marsh,ArgentinaS. densiflora G3Bolinas Lagoon, CA,USAS. foliosa G3Palo Alto, CA, USA S. foliosa G3Mountain View, CA, USA S. foliosa G3San Mateo, CA, USA S. foliosa G3Point Reyes NS, CA,USAS foliosaG3and tribe (Cynodonteae) (Peterson et al. 2001). This suggestsa phylogenetic correspondence between host and pathogen.The range <strong>of</strong> D. spicata includes <strong>the</strong> Atlantic Coast <strong>of</strong> NorthAmerica, <strong>the</strong> Gulf Coast states, Cuba, and <strong>the</strong> Pacific Coast<strong>of</strong> North America, from British Columbia south to Mexico,and South America (Hitchcock 1971). Distichlis spicata istypically found in salt marshes and seashores on moist andalkaline soils, and borders some S. alterniflora marshes inWillapa Bay, WA.In Argentina, G3 C. purpurea was collected from S.densiflora. Spartina densiflora is native to <strong>the</strong> Atlantic andPacific coasts <strong>of</strong> sou<strong>the</strong>rn South America (Mobberley 1956).Based on alkaloid pr<strong>of</strong>ile, Samuelson and Gjerstad (1966)proposed that C. purpurea from Spartina spp. in coastal-44-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyArgentina was a separate species, C. maritima, or ‘fea<strong>the</strong>rergot.’ The host at <strong>the</strong> time was misidentified as S. maritima(Eleuterius and Meyers 1977). Eleuterius and Meyers (1977)proposed it was more likely to have been S. alterniflora,based on floristic surveys <strong>of</strong> <strong>the</strong> region, but since S.densiflora is now known to be a host in South America, itcould be a candidate as well.Our results support <strong>the</strong> recognition <strong>of</strong> three discretegroups within C. purpurea. Results <strong>of</strong> RAPD and AFLPanalysis showed high intergroup variability between G1, G2,and G3. Using a collection <strong>of</strong> isolates from a different set <strong>of</strong>hosts, Jungehülsing and Tudzynski (1997) also found highintraspecific variability using RAPD markers. In <strong>the</strong>ir study,parsimony analysis grouped samples from <strong>the</strong> same hostspecies toge<strong>the</strong>r, suggesting some degree <strong>of</strong> host specificity.More sampling is necessary to reveal <strong>the</strong> ecological andphysiological range <strong>of</strong> G3, though <strong>the</strong>re is evidence that <strong>the</strong>pathogenicity pr<strong>of</strong>ile <strong>of</strong> G3, following greenhouseinoculations, is not limited to salt marsh species (Pazoutováet al. 2002a). Laboratory experiments notwithstanding, ourresults imply a degree <strong>of</strong> specificity in <strong>the</strong> ecological range <strong>of</strong>G3.Our results confirm <strong>the</strong> presence <strong>of</strong> three intraspecificgroups within C. purpurea and support from both RAPD andAFLP pr<strong>of</strong>iles strongly suggests widespread geographicdispersal <strong>of</strong> salt marsh C. purpurea throughout <strong>the</strong> range <strong>of</strong>both native and invasive Spartina species.ACKNOWLEDGMENTSThanks to S. Pazoutová for providing C. purpureaisolates; D. Strong for comments on manuscripts; A.Bartolis, K. Heck, D. Garcia-Rossi, T. Buck, A. Kolker, S.Orl<strong>of</strong>f, M. Burzynski, D. Wesenberg and E. Landy for fieldsamples; B. Aegerter, D. Ayres, G. Douhan and D. Rizzo forassistance and use <strong>of</strong> laboratory equipment.This study was financially supported by The GardenClub <strong>of</strong> America, Jastro Shields and Humanities awardsfrom <strong>the</strong> University <strong>of</strong> California, Davis and NSFBiocomplexity DEB 0083583 to A. Hastings.REFERENCESBoyle, P.J. 1976. Ergot epiphytotic on Spartina spp. in Ireland.Irish Journal <strong>of</strong> Agricultural Research 15: 419-424.Civille J.C., K. Sayce, S.D. Smith and D.R. Strong. 2005. Reconstructinga century <strong>of</strong> Spartina invasion with historical recordsand contemporary remote sensing. Ecoscience 12: 330-338.Daehler, C.C. and D.R. Strong. 1994. Variable reproductive outputamong clones <strong>of</strong> Spartina alterniflora (Poaceae) invading SanFrancisco Bay, California: <strong>the</strong> influence <strong>of</strong> herbivory, pollination,and establishment site. American Journal <strong>of</strong> Botany 81:307-131.Eleuterius, L.N. and S.P. Meyers. 1977. Alkaloids <strong>of</strong> Clavicepsfrom Spartina. Mycologia 69: 838-840.Fisher, A.J., J.M. DiTomaso, T.R. Gordon, B. Aegerter, and D.R.Ayres. 2007. Salt marsh Claviceps purpurea in native and invadedSpartina marshes in Nor<strong>the</strong>rn California. Plant Disease91:380-386.Hitchcock, A.S. 1971. Manual <strong>of</strong> <strong>the</strong> Grasses <strong>of</strong> <strong>the</strong> United States.Dover: New York.Jungehülsing, U. and P. Tudzynski. 1997. Analysis <strong>of</strong> genetic diversityin Claviceps purpurea by RAPD markers. MycologicalResearch 101: 1-6.Kobel, H. and J.J. Sanglier. 1978. Formation <strong>of</strong> ergotoxine alkaloidsby fermentation and attempts to control <strong>the</strong>ir biosyn<strong>the</strong>ses.FEMS Symposium 5: 233-242.Loveless, A. R. (1971). Conidial evidence for host restriction inClaviceps purpurea. Transactions <strong>of</strong> <strong>the</strong> British Mycological Society56: 419-434.Mobberley, D.G. 1956. Taxonomy and distribution <strong>of</strong> <strong>the</strong> genusSpartina. Iowa State Journal <strong>of</strong> Science 30: 471-574.Pazoutová, S., B. Cagas, R. Kolínská, and A. Honzátko. 2002a.Host specialization <strong>of</strong> different populations <strong>of</strong> ergot fungus(Claviceps purpurea). Czech Journal <strong>of</strong> Genetics and PlantBreeding 38: 75-81.Pazoutová, S., J. Olsovska, M. Linka, R. Kolínská, and M. Flieger.2000. Chemoraces and habitat specialization <strong>of</strong> Claviceps purpureapopulations. Applied and Environmental Microbiology 66:5419-5425.Pazoutová, S., A.F. Raybould, A. Honzátko, and R. Kolínská.2002b. Specialized populations <strong>of</strong> Claviceps purpurea from saltmarsh Spartina species. Mycological Research 106: 210-214.Peterson, P.M., R.J. Soreng, D. Davidse, T.S. Filgueiras, F.O. Zuloagaand E.J. Judziewicz. 2001. Catalogue <strong>of</strong> New WorldGrasses (Poaceae): II. Subfamily Chloridoideae. Contributions <strong>of</strong><strong>the</strong> U.S. National Herbarium 41: 1-255.Raybould A.F., A.J. Gray, M.J. Lawrence and D.F. Marshall. 1991.The evolution <strong>of</strong> Spartina anglica C. E. Hubbard (Gramineae);<strong>the</strong> origin and genetic variability. Biological Journal <strong>of</strong> <strong>the</strong> LinneanSociety 43: 111-126.Samuelson, R.J. and G. Gjerstad. 1966. Intermediary metabolism <strong>of</strong>ergot XIII. Fea<strong>the</strong>r ergot, a new, promising Claviceps species.Meddelelser Fra Norsk Farmaceutisk Selskap 28: 229-237.Sprague, R. 1950. Diseases <strong>of</strong> Cereals and grasses in North America.Ronald Press: New York.Stiller, J. W. and A.L. Denton. 1995. One hundred years <strong>of</strong>Spartina alterniflora (Poaceae) in Willapa Bay, Washington:random amplified polymorphic DNA analysis <strong>of</strong> an invasivepopulation. Molecular Ecology 4: 355-363.-45-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyMECHANISMS OF SULFIDE AND AOXIA TOLERANCE IN SALT MARSH GRASSESIN RELATION TO ELEVATIONAL ZONATIONB.R. MARICLE 1,2 AND R.W. LEE 11 School <strong>of</strong> Biological Sciences, Washington State University, Pullman, WA 99164-42362 Present address: Department <strong>of</strong> Biological Sciences, Fort Hays State University, Hays, KS 67601-4099; brmaricle@fhsu.eduSharply-defined ecotones commonly separate species living in high intertidal and low intertidalestuarine zones. Low intertidal regions are characterized by anoxic sediments and toxic levels <strong>of</strong>hydrogen sulfide. These conditions exclude high marsh species. In contrast, low marsh species arebelieved to possess physiological adaptations to resist <strong>the</strong> anoxia and sulfide. However, <strong>the</strong>seadaptations are poorly understood. One <strong>of</strong> <strong>the</strong> most important characteristics <strong>of</strong> waterloggedsediments is <strong>the</strong> lack <strong>of</strong> oxygen. Many wetland plants have been shown to transport atmosphericoxygen internally to support respiration in submerged tissues. This ability may allow plant survivalin low intertidal marsh areas and is <strong>of</strong>ten implicated as a factor in determining species zonation inestuaries. In this study, oxygen transport and metabolic characteristics related to anoxia toleranceand rhizosphere oxidation were investigated in <strong>the</strong> emergent estuarine species Spartina alterniflora,S. anglica, S. densiflora, S. patens, and Distichlis spicata (Poaceae). Plants were grown ingreenhouse experiments under simulated estuarine conditions. All species showed a strong ability torespire anaerobically. The high intertidal marsh species S. densiflora, S. patens, and D. spicata werefound to have high aerobic respiration rates, low oxygen transport rates, and an apparent highsensitivity to sulfide. The low intertidal marsh species S. alterniflora and S. anglica had loweraerobic respiration rates, moderate to high oxygen transport rates, and a lower sensitivity to sulfide.Spartina anglica appeared to have <strong>the</strong> greatest ability to transport oxygen and was more resistant tomudflat-related stressors compared to <strong>the</strong> o<strong>the</strong>r plants in this study. Evidence is presented thataerobic respiration rates and sulfide sensitivity may be important factors for differences in estuarinezonation between species.Keywords: Distichlis spicata, hypoxia, oxygen transport, respiration, sediment, Spartina, sulfideINTRODUCTIONThe introduction <strong>of</strong> four species <strong>of</strong> Spartina grasses(Poaceae) into Washington estuaries has led to manydevastating ecological and economic impacts. Nearly 8,100hectares (ha) <strong>of</strong> intertidal mudflat in Willapa Bay,Washington, USA, has been affected by introduced S.alterniflora (Hedge et al. 2003). Similarly, S. anglica wasintroduced in nor<strong>the</strong>rn Puget Sound, Washington in 1961 topresvent shoreline erosion, but quickly spread via tidalcurrents and now affects over 3,300 ha in <strong>the</strong> Puget Soundarea (Hacker et al. 2001). O<strong>the</strong>r Spartina introductions intoWashington estuaries include Spartina densiflora Brong.(WSDA news release 11 Jan 2002) and S. patens (Aiton)Muhl. (Frenkel 1987). These populations remain small andare closely monitored to prevent spread.Introduced Spartina flourishes in West Coast estuariesbecause it can occupy an open niche: low intertidal mudflatsand tidal channels characterized by highly reducingconditions (an oxidation reduction potential less than -300millivolts [Eh

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinain <strong>the</strong>se environments is thought to be oxygen transport to<strong>the</strong> roots and rhizosphere, facilitated by a system <strong>of</strong> gasspaces (aerenchyma) connecting leaves to root tissues (Tealand Kanwisher 1966, Hwang and Morris 1991, Arenovskiand Howes 1992, Howes and Teal 1994). The presence andfunctioning <strong>of</strong> aerenchyma is well documented in plantstolerant <strong>of</strong> flooded conditions, and generally results in asupply <strong>of</strong> oxygen for aerobic respiration as well as radialoxygen loss to <strong>the</strong> environment (reviewed by Jackson andArmstrong 1999).Once oxygen has reached submerged tissues inemergent plants like Spartina, it has at least three possiblefates (Fig. 1). Oxygen can be released into <strong>the</strong> rhizosphere,support mitochondrial respiration, or be used in sulfideoxidation processes. The strength <strong>of</strong> <strong>the</strong>se competing oxygensinks can be important in <strong>the</strong> ecophysiology <strong>of</strong> wetlandplants since it influences how oxygen is budgeted insubmerged tissues (Sorrell 1999).Anoxic estuarine sediments represent a strong externaloxygen sink and <strong>the</strong>y can overwhelm plant oxygen transportprocesses. Therefore, Spartina grasses also must exhibit astrong capability for anaerobic respiration to sustainmetabolism when oxygen supplies are low. The enzymealcohol dehydrogenase (ADH) catalyzes <strong>the</strong> final reaction infermentative ethanol syn<strong>the</strong>sis. The ability to respire underhypoxic conditions is important for life in waterlogged soils,so ADH activity in <strong>the</strong>se plants appears to be an adaptationfor anoxia tolerance (Crawford 1967).To understand <strong>the</strong> mechanisms conferring success inlow intertidal zones, aspects <strong>of</strong> oxygen transport andmetabolic characteristics related to anoxia tolerance andrhizosphere oxidation were investigated in greenhouseSpartina plants. The four Spartina species introduced intoFig. 1. Atmospheric oxygen can be transported through wetland plants tosubmerged tissues. Once oxygen reaches <strong>the</strong> roots (inset), it has severalpossible fates. Measurements <strong>of</strong> <strong>the</strong> indicated processes and enzymeactivities may indicate how oxygen is allocated in <strong>the</strong> roots <strong>of</strong> flood-tolerantplants.Washington estuaries provide a good system for studyingestuarine zonation since <strong>the</strong>y represent a range from highmarsh to low marsh species. The low marsh species Spartinaalterniflora and S. anglica were studied and compared to <strong>the</strong>high marsh species S. densiflora, S. patens, and <strong>the</strong> nativeDistichlis spicata.Plants are aerobes. Therefore, survival in waterloggedsoils requires a supply <strong>of</strong> oxygen to support tissuessubmerged in anoxic sediments (Crawford 1982). However,many additional physiological processes can be affected by<strong>the</strong> supply <strong>of</strong> oxygen to submerged tissues in wetland plants.Measures <strong>of</strong> <strong>the</strong> mechanisms shown in Fig. 1 may allow oneto estimate how much oxygen is used in aerobic respirationand how strong external oxygen sinks may be. In <strong>the</strong> presentstudy, rates <strong>of</strong> oxygen transport were measured andcompared to rates <strong>of</strong> aerobic respiration. Highly reducingmudflat conditions may inhibit aerobic respiration processesand induce alternative anaerobic respiration pathways.Therefore, rates <strong>of</strong> aerobic respiration were measured as wellas activities <strong>of</strong> <strong>the</strong> aerobic respiration enzyme cytochrome coxidase and <strong>the</strong> anaerobic respiration enzyme alcoholdehydrogenase. Tolerance <strong>of</strong> estuarine mudflat conditionsmay also require mechanisms to detoxify hydrogen sulfide, aphytotoxin produced under anaerobic conditions.Consequently, rates <strong>of</strong> sulfide oxidation processes were alsomeasured in root tissues. The results <strong>of</strong> this study areexpected to provide a physiological explanation to helpdefine differences between high marsh and low marshfunctional types and <strong>the</strong>ir relationship to estuarine zonation.MATERIALS AND METHODSSpartina plants were collected from field sites andsubsequently maintained under greenhouse conditions.Spartina alterniflora plants were collected in Willapa Bay,Washington and S. anglica was collected from nor<strong>the</strong>rnPuget Sound, Washington. Additionally, S. patens plantswere obtained from <strong>the</strong> Gulf Coast <strong>of</strong> northwest Florida andS. densiflora plants were obtained from <strong>the</strong> Odiel SaltMarshes, southwest Spain. The native Distichlis spicata wascollected in nor<strong>the</strong>rn Puget Sound, Washington.Greenhouse temperatures were 26°C during <strong>the</strong> day and18°C at night. Natural lighting provided a photosyn<strong>the</strong>ticphoton flux density (PPFD) averaging 200 micromolesquanta per square meter per second (200 μmol quanta m -2sec -1 ) during daylight hours with peaks around 1,100 μmolquanta m -2 sec -1 on sunny days. Daughter tillers from fieldcollectedplants were potted individually in a 50/50(vol./vol.) sand/potting soil mixture and were watered tosaturation twice weekly with modified Hoagland nutrientsolution (Epstein 1972). Freshly potted plants were selectedfor uniformity in size and randomized between flooded anddrained treatments. At least four replicate plants were grownin each treatment. Plants were allowed 60-80 days in <strong>the</strong>ir-48-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina Biologyrespective treatment (drained or flooded) before testing orharvest. All statistical analyses were performed betweenspecies and treatments with a two-factor (species andtreatment) analysis <strong>of</strong> variance (Statview 5; 1998 SASInstitute Inc., Cary, NC; α=0.05).Individual plants were tested for <strong>the</strong>ir ability to transportoxygen internally after <strong>the</strong> method <strong>of</strong> Maricle and Lee(2002). A fiber optic oxygen-sensing probe (FOXY-R probe;Ocean Optics Inc., Dunedin, FL) was used to measuredissolved oxygen concentrations in sealed flasks containingroots <strong>of</strong> intact plants suspended in water. Flask watercontained penicillin G and streptomycin sulfate at 1 gramper liter (1 g L -1 ) each and 50 mg L -1 chloramphenicol toprevent bacterial respiration. During testing, plants wereplaced under a 250 watt (W) metal halide light (Hydr<strong>of</strong>armgardening products; Petaluma, CA). At plant level, PPFDwas about 150 μmol quanta m -2 s -1 and air temperature was28˚C. Flask dissolved oxygen concentrations started at 100μM (±10 μM) and were measured over a period <strong>of</strong> 2-3 hoursto observe consumption or release <strong>of</strong> oxygen by plants to <strong>the</strong>surrounding medium.Parallel measurements <strong>of</strong> oxygen consumption wereconducted on plants where aerenchyma transport capabilitieswere blocked. Placing <strong>the</strong> shoot <strong>of</strong> <strong>the</strong> plant in a 100% N 2atmosphere prevents <strong>the</strong> entry <strong>of</strong> oxygen into <strong>the</strong>aerenchyma system (Armstrong 1964, Teal and Kanwisher1966). Plants were maintained in <strong>the</strong> dark during thismeasurement to prevent <strong>the</strong> release <strong>of</strong> photosyn<strong>the</strong>ticoxygen. Rates <strong>of</strong> oxygen consumption were comparedbetween plants under a 21% O 2 atmosphere and a 100% N 2atmosphere; <strong>the</strong> difference between <strong>the</strong> two flux rates equals<strong>the</strong> amount <strong>of</strong> oxygen transported internally through <strong>the</strong>plant’s aerenchyma system (Lee 2003). Rates <strong>of</strong> internaloxygen transport were standardized to g fresh root weight.Oxygen consumption in <strong>the</strong> dark under 100% N 2 representstotal oxygen demand by <strong>the</strong> plant and was taken to be <strong>the</strong>dark respiration rate (Maricle and Lee 2007).At harvest, root samples were obtained from each plant,flash-frozen in liquid nitrogen, and stored at –80°C.Cytochrome c oxidase (CytOx) and sulfide oxidase (SOx)activities were determined in extracts from root tissuesamples. Roots were ground in liquid nitrogen and coldextraction buffer was added at 2 milliliters per gram (2 mLg -1 ) (Maxwell and Bateman 1967). This mixture washomogenized with a mortar and pestle, filtered throughMiracloth (Calbiochem; San Diego, CA), and centrifuged at1,000 g for 20 minutes at 4°C. The supernatant was used inCytOx and SOx assays. Alcohol dehydrogenase (ADH) wasextracted from root tissue samples after John and Greenway(1976). Roots were ground in liquid nitrogen, and coldextraction buffer was added at 5 mL g -1 . The resultingmixture was homogenized with a mortar and pestle, filteredthrough Miracloth, and centrifuged at 10,000 g for 10 min at4°C. This supernatant was used in ADH assays. All enzymeassays were performed spectrophotometrically at 25°C(Maricle et al. 2006).CytOx activity was determined as <strong>the</strong> rate <strong>of</strong>cytochrome c oxidation, measured as a decrease inabsorbance at 550 nanometers (nm) (Smith 1955). Rates <strong>of</strong>CytOx activity were corrected for background rates <strong>of</strong>cytochrome c oxidation, <strong>the</strong>n standardized to g fresh rootweight. ADH activity was assayed spectrophotometrically in<strong>the</strong> ethanol-forming direction (John and Greenway 1976).Enzyme activity was determined as <strong>the</strong> rate <strong>of</strong> NADHoxidation, measured as a decrease in absorbance at 340 nm.Rates <strong>of</strong> NADH oxidation in <strong>the</strong> presence <strong>of</strong> acetaldehydewere corrected for background rates <strong>of</strong> ethanol formation,<strong>the</strong>n standardized to g fresh root weight (Maricle et al.2006).A colorimetric method was developed to measure <strong>the</strong>activity <strong>of</strong> sulfide oxidation processes. 50 microliter (μL)aliquots <strong>of</strong> extract were added to a series <strong>of</strong> 1 mL buffersolutions <strong>of</strong> 100 μM Na 2 S. 40 μL <strong>of</strong> Cline reagent (Cline1969) was added to <strong>the</strong> solutions after 0, 10, and 20 min todetermine <strong>the</strong> amount <strong>of</strong> sulfide present. Background rates<strong>of</strong> sulfide oxidation were measured by adding 50 μLphosphate buffer instead <strong>of</strong> enzyme extract to a series <strong>of</strong>tubes containing 100 μM Na 2 S as described above. SOxactivity was measured as a decrease in sulfide concentrationover time. Total rates <strong>of</strong> SOx activity were corrected forbackground rates <strong>of</strong> spontaneous sulfide oxidation, <strong>the</strong>nstandardized to g fresh root weight. Rates <strong>of</strong> nonenzymaticsulfide oxidation were determined using 50 μL aliquots <strong>of</strong>boiled enzyme extract. Enzymatic rates <strong>of</strong> sulfide oxidationwere calculated from <strong>the</strong> difference between <strong>the</strong> rates <strong>of</strong>total sulfide oxidation and nonenzymatic sulfide oxidation(Maricle et al. 2006).RESULTS AND DISCUSSIONSoil waterlogging is <strong>the</strong> most common cause <strong>of</strong> plantoxygen deficiency (Vartapetian and Jackson 1997). Impacts<strong>of</strong> flooding on plant productivity can also have significantimpacts on commodity crops. Excessive rains in <strong>the</strong> spring<strong>of</strong> 1993 resulted in a 33% reduction in Midwest cropproduction (Bray et al. 2000). Despite <strong>the</strong>se kinds <strong>of</strong>economic losses, many observers may have overlooked <strong>the</strong>cattails and o<strong>the</strong>r wetland plants flourishing in nearbyflooded ditches. Differences in flooding tolerance have longbeen recognized between plant species, but <strong>the</strong> specificphysiology governing <strong>the</strong> differences has remained largelyunknown.Internal oxygen transport is <strong>of</strong>ten regarded to be animportant factor conferring plant success in waterloggedareas, and <strong>the</strong>refore is thought to be important in estuarinezonation (e.g., Gleason and Zieman 1981, Bertness 1991).The oxygen transport rates measured in this study exhibited-49-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaFig. 2. (a.) Internal oxygen transport rates, and (b.) root respiration rates(μmol g -1 h -1 ) <strong>of</strong> Spartina and Distichlis grown under flooded soiltreatments. Shown is <strong>the</strong> mean <strong>of</strong> 5-14 plants ± SE. Species are labeled asfollows: S. alt = Spartina alterniflora, S. ang = S. anglica, S. den = S.densiflora, Dist. = Distichlis spicata, S. pat = S. patens.a variability among individuals, with some apparentdifferences between species. The highest rates <strong>of</strong> transportwere exhibited by S. anglica, with moderate to low rates inall o<strong>the</strong>r species (Fig. 2a). Compared to <strong>the</strong> o<strong>the</strong>r species inthis study, S. anglica showed <strong>the</strong> greatest ability to transportoxygen (analysis <strong>of</strong> variance [ANOVA], p

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyFig. 4. Alcohol dehydrogenase (ADH) activities (μmol g -1 min -1 ) <strong>of</strong>Spartina and Distichlis grown under drained and flooded soil treatments.Shown is <strong>the</strong> mean <strong>of</strong> 7-13 plants ± SE. Species are labeled as in Fig. 2.activities in S. anglica were significantly lower than all o<strong>the</strong>rspecies (ANOVA, p≤0.046). Flooded soil conditionsresulted in significantly higher ADH activities in all species(ANOVA, p≤0.018) expect <strong>the</strong> low marsh species S. anglica(ANOVA, p=0.564).Most plants can survive short term absence <strong>of</strong> oxygen(≤60 min) without cell death. In <strong>the</strong>se cases, normal ATPstores are quickly depleted in active cells, and mitochondrialswelling is usually observed within minutes (Drew 1997).Large reserves <strong>of</strong> stored carbon can help <strong>the</strong> cells respireanaerobically, but longer absences <strong>of</strong> oxygen can result incell death. Irreversible damage to mitochondria and cellviability generally occurs following 15 hours <strong>of</strong> anaerobiosis(Perata and Alpi 1993). However, a supply <strong>of</strong> oxygen frominternal aeration can allow marsh plants like Spartina torespire aerobically despite growing in waterloggedsubstrates. The superior oxygen transport abilities <strong>of</strong> S.anglica may have helped to account for its low root ADHactivities observed in this study.The plants studied showed varying degrees <strong>of</strong> sulfideoxidation capacity. Total sulfide oxidation was partitionedinto enzymatic and nonenzymatic processes. The meanenzymatic sulfide oxidase (SOx) activity ranged from 14.4to 97.2 nmol g -1 min -1 across species and waterloggingtreatments (Fig. 5a). Enzymatic SOx activities were highestin <strong>the</strong> high marsh species S. patens and D. spicata and weresignificantly lower in S. densiflora and <strong>the</strong> low marshspecies S. alterniflora and S. anglica (ANOVA, p≤0.050).This difference may suggest that high marsh species aremore sensitive to sediment sulfides and thus require greaterenzymatic protection. Enzymatic SOx activities did notchange in response to flooding across species (ANOVA,p≥0.960).Nonbiotic factors such as metal ions can contribute tosulfide oxidation (Lee et al. 1999). Such nonenzymaticprocesses were also found to be important in sulfideoxidation in <strong>the</strong> present study. Mean nonenzymatic sulfideFig. 5. Sulfide oxidase (SOx) activities (nmol g -1 min -1 ) <strong>of</strong> Spartina andDistichlis grown under drained and flooded soil treatments. Shown are (a.)enzymatic and (b.) nonenzymatic rates <strong>of</strong> sulfide oxidation. The mean <strong>of</strong>4-10 plants are shown ± SE. Species are labeled as in Fig. 2.oxidation rates ranged from 12.6 to 51.1 nmol g -1 min -1across species and waterlogging treatments (Fig. 5b).Nonenzymatic rates <strong>of</strong> sulfide oxidation were not differentbetween species or flooding treatment (ANOVA, p≥0.141).CONCLUSIONSThe upper regions <strong>of</strong> salt marshes are characterized byoxidized soils, since tidal flooding is rare. However, episodicflooding at <strong>the</strong> highest tides can result in occasional anoxicand sulfidic conditions. Therefore, plants from <strong>the</strong> highmarsh are not forced to withstand chronic anoxia. The highmarsh species S. patens, S. densiflora, and Distichlis spicatawere found to have high aerobic respiration rates and highaerobic enzyme activity. This aerobic oxygen demand maybe too high to allow survival in anoxic low marshconditions, where plants had lower aerobic demand.Anaerobic pathways (root ADH activities) increased afterflooding in all three high marsh species suggesting a highsensitivity to soil waterlogging. Internal oxygen transportrates were low in <strong>the</strong>se plants since <strong>the</strong>y are adapted to lifein sediments where soil oxygen is not normally limiting toroot respiration.Additionally, higher SOx activities were found in highmarsh species compared to low marsh species. These trendssuggested that high marsh species were more sensitive tosulfide and required greater protection <strong>of</strong> aerobic respiration.This idea is consistent with <strong>the</strong> finding that <strong>the</strong>se speciesexhibited higher activities <strong>of</strong> CytOx, <strong>the</strong> site <strong>of</strong> sulfideinhibition <strong>of</strong> aerobic respiration (Bagarinao 1992). Highrates <strong>of</strong> aerobic respiration and apparent sulfide sensitivitymay substantially account for <strong>the</strong> exclusion <strong>of</strong> <strong>the</strong>se speciesfrom low marsh zones.The low zones <strong>of</strong> salt marshes are characterized byfrequent tidal flooding. This leads to highly reducedsediments, <strong>of</strong>ten containing high levels <strong>of</strong> sulfides. Plantsinhabiting low marsh regions must be able to tolerate highlyreducing and sulfidic sediment conditions. Spartinaalterniflora is <strong>the</strong> dominant low marsh species in manyNorth American East- and Gulf Coast estuaries (Bertness-51-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartina1991). Spartina anglica can grow lower in <strong>the</strong> intertidalrange than S. alterniflora (Frenkel 1987, Sayce andMumford 1990), and <strong>the</strong>refore any o<strong>the</strong>r species in thisstudy. Low marsh species exhibited low aerobic respirationrates and CytOx activity, which when coupled to high rates<strong>of</strong> internal oxygen transport may pose a significantadvantage for survival in anoxic sediments. The ability tosupply oxygen to submerged tissue is crucial to survival in<strong>the</strong> low marsh, as no plant tissue can endure anoxiaindefinitely (Crawford 1982). Low marsh species must alsopossess an ability to respire anaerobically since demand foroxygen from highly reduced sediments may overwhelmtransport processes. ADH activities measured in Spartinaroots indicated a well-developed capacity for fermentation.However, increases in root ADH were not observed in <strong>the</strong>low marsh species S. anglica. High rates <strong>of</strong> oxygen transportin S. anglica may be adequate to supply oxygen to roots andexternal sinks, as suggested by its low root ADH activities.Low marsh species may be more resistant to sulfides whencompared to high marsh species. Lower aerobic respirationrates and lower CytOx activities may relax needs for intensesulfide oxidation requirements. Alternatively, higher rates <strong>of</strong>oxygen transport in some low marsh species may help tooxidize rhizosphere sulfides, resulting in reduced SOxactivity. However, <strong>the</strong> acute toxicity <strong>of</strong> dissolved sulfidesaround roots still necessitates moderate SOx activities in lowmarsh species.The results <strong>of</strong> this study suggest metabolic characteristicsrelated to respiration and sulfide tolerance may affect zonation<strong>of</strong> grasses in estuaries. While internal oxygen transport isimportant for survival in estuarine sediments, this study mayindicate that zonation within estuaries is dependent on morethan just oxygen transport. Aerobic respiration rates andsensitivity to sediment sulfides may play a large role ininfluencing estuarine zonation as well.ACKNOWLEDGMENTSThe authors thank Chuck Cody for greenhouse assistance;Kim Patten, Sally Hacker, Eric Hellquist, and M. Enrique Figueroafor providing plants; and <strong>the</strong> Washington State Department <strong>of</strong>Agriculture for a Spartina transport permit. This project waspartially funded from <strong>the</strong> Betty W. Higinbotham Trust, NSFIBN0076604, EPA grant R-82940601, and NSF DBI-0116203.REFERENCESAdam, P. 2002. Saltmarshes in a time <strong>of</strong> change. EnvironmentalConservation 29:39-61.Arenovski, A.L., and B.L. Howes. 1992. Lacunal allocation and gastransport capacity in <strong>the</strong> salt marsh grass Spartina alterniflora.Oecologia 90:316-322.Armstrong, W. 1964. Oxygen diffusion from <strong>the</strong> roots <strong>of</strong> someBritish bog plants. Nature 204:801-802.Bagarinao, T. 1992. Sulfide as an environmental factor and toxicant:Tolerance and adaptations in aquatic organisms. AquaticToxicology 24:21-62.Bertness, M.D. 1991. Zonation <strong>of</strong> Spartina patens and Spartinaalterniflora in a New England salt marsh. Ecology 72:138-148.Bray, E.A., J. Bailey-Serres, and E. Weretilnyk. 2000. Responses toabiotic stresses. In: Buchanan, B. B., W. Gruissem, and R. L.Jones, eds. Biochemistry and Molecular Biology <strong>of</strong> Plants.American Society <strong>of</strong> Plant Physiologists: Rockville, Maryland.pp. 1158-1203.Brix, H., and B.K. Sorrell. 1996. Oxygen stress in wetland plants:comparison <strong>of</strong> de-oxygenated and reducing root environments.Functional Ecology 10:521-526.Burdick, D.M., and I.A. Mendelssohn. 1987. Waterlogging responsesin dune, swale, and marsh populations <strong>of</strong> Spartina patensunder field conditions. Oecologia 74:321-329.Byrd, G.T., R.F. Sage, and R.H. Brown. 1992. A comparison <strong>of</strong>dark respiration between C 3 and C 4 plants. Plant Physiology100:191-198.Cline, J.D. 1969. Spectrophotometric determination <strong>of</strong> hydrogensulfide in natural waters. Limnology and Oceanography 14:454-458.Crawford, R.M.M. 1967. Alcohol dehydrogenase activity in relationto flooding tolerance in roots. Journal <strong>of</strong> Experimental Botany18:458-464.Crawford, R.M.M. 1982. Physiological responses to flooding. In:Lange, O.S. P.S. Nobel, C.B. Osmond, and H. Ziegler, eds. Encyclopedia<strong>of</strong> Plant Physiology, Physiolological Plant EcologyII. Water Relations and Carbon Assimilation. Springer-Verlag:Berlin. pp. 453-477Drew, M.C. 1997. Oxygen deficiency and root metabolism: Injuryand acclimation under hypoxia and anoxia. Annual Review <strong>of</strong>Plant Physiology and Plant Molecular Biology 48:223-250.Emmett, R., R. Llanso, J. Newton, R. Thom, M. Hornberger, C.Morgan, C. Levings, A. Copping, and P. Fishman. 2000. Geographicsignatures <strong>of</strong> North American West Coast estuaries. Estuaries23:765-792.Epstein, E. 1972. Mineral Nutrition <strong>of</strong> Plants: Principles and Perspectives.John Wiley and Sons, Inc.: New York.Frenkel, R.E. 1987. Introduction and spread <strong>of</strong> cordgrass (Spartina)into <strong>the</strong> Pacific Northwest. Northwest Environmental Journal3:152-154.Gleason, M.L., and J.C. Zieman. 1981. Influence <strong>of</strong> tidal inundationin internal oxygen supply <strong>of</strong> Spartina alterniflora andSpartina patens. Estuarine, Coastal and Shelf Science 13:47-57.Hacker, S.D., D. Heimer, C.E. Hellquist, T G. Reeder, B. Reeves,T.J. Riordan, and M.N. Dethier. 2001. A marine plant (Spartinaanglica) invades widely varying habitats: Potential mechanisms<strong>of</strong> invasion and control. Biological Invasions 3:211-217.Hedge, P., L.K. Kriwoken, and K. Patten. 2003. A review <strong>of</strong>Spartina management in Washington State, US. Journal <strong>of</strong>Aquatic Plant Management 41:82-90.Howes, B.L., and J.M. Teal. 1994. Oxygen loss from Spartina alternifloraand its relationship to salt marsh oxygen balance.Oecologia 97:431-438.Hwang, Y.-H., and J.T. Morris. 1991. Evidence for hygrometricpressurization in <strong>the</strong> internal gas space <strong>of</strong> Spartina alterniflora.Plant Physiology 96:166-171.Jackson, M.B., and W. Armstrong. 1999. Formation <strong>of</strong> aerenchymaand <strong>the</strong> process <strong>of</strong> plant ventilation in relation to soil floodingand submergence. Plant Biology 1:274-287.John, C.D., and H. Greenway. 1976. Alcoholic fermentation andactivity <strong>of</strong> some enzymes in rice roots under anaerobiosis. AustralianJournal <strong>of</strong> Plant Physiology 3:325-336.-52-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyLee, R.W. 2003. Physiological adaptations <strong>of</strong> <strong>the</strong> invasivecordgrass Spartina anglica to reducing sediments: rhizome metabolicgas fluxes and enhanced O 2 and H 2 S transport. Marine Biology143:9-15.Lee, R.W., D.W. Kraus, and J.E. Doeller. 1999. Oxidation <strong>of</strong> sulfideby Spartina alterniflora roots. Limnology and Oceanography44:1155-1159.Maricle, B.R., and R.W. Lee. 2002. Aerenchyma development andoxygen transport in <strong>the</strong> estuarine cordgrasses Spartina alternifloraand S. anglica. Aquatic Botany 74:109-120.Maricle, B.R., J.J. Crosier, B.C. Bussiere, and R.W. Lee. 2006.Respiratory enzyme activities correlate with anoxia tolerance insalt marsh grasses. Journal <strong>of</strong> Experimental Marine Biology andEcology 337:30-37.Maricle, B.R., and R.W. Lee. 2007. Root respiration and oxygenflux in saltmarsh grasses from different elevational zones. MarineBiology 151:413-423.Maxwell, D.P., and D.F. Bateman. 1967. Changes in <strong>the</strong> activities<strong>of</strong> some oxidases in extracts <strong>of</strong> Rhizoctonia-infected bean hypocotylsin relation to lesion maturation. Phytopathology 57:132-136.Mendelssohn, I.A., K.L. McKee, and W.H. Patrick, Jr. 1981. Oxygendeficiency in Spartina alterniflora roots: Metabolic adaptationto anoxia. Science 214:439-441.Naidoo, G., K.L. McKee, and I.A. Mendelssohn. 1992. Anatomicaland metabolic responses to waterlogging and salinity in Spartinaalterniflora and S. patens. American Journal <strong>of</strong> Botany 79:765-770.Perata, P., and A. Alpi. 1993. Plant responses to anaerobiosis. PlantScience 93:1-17.Pezeshki, S.R., J.H. Pardue, and R.D. DeLaune. 1993. The influence<strong>of</strong> soil oxygen deficiency on alcohol dehyrogenase activity,root porosity, ethylene production, and photosyn<strong>the</strong>sis inSpartina patens. Environmental and Experimental Botany33:565-573.Sayce, K., and T.F. Mumford, Jr. 1990. Identifying <strong>the</strong> Spartinaspecies. In: Mumford, T.F. Jr., ed. The Spartina Workshop Record.Washington Sea Grant Program: Seattle, Washington. pp.9-14.Smith, L. 1955. Spectrophotometric assay <strong>of</strong> cytochrome c oxidase.Methods in Biochemical Analysis 2:427-434.Sorrell, B.K. 1999. Effect <strong>of</strong> external oxygen demand on radialoxygen loss by Juncus roots in titanium citrate solutions. Plant,Cell and Environment 22:1587-1593.Teal, J., and M. Teal. 1969. Life and Death <strong>of</strong> <strong>the</strong> Salt Marsh. BallantineBooks, New York.Teal, J.M., and J.W. Kanwisher. 1966. Gas transport in <strong>the</strong> marshgrass, Spartina alterniflora. Journal <strong>of</strong> Experimental Botany17:355-361.Thompson, J.D. 1991. The biology <strong>of</strong> an invasive plant: Whatmakes Spartina anglica so successful? BioScience 41:393-401.Vartapetian, B.B., and M.B. Jackson. 1997. Plant adaptation toanaerobic stress. Annals <strong>of</strong> Botany 79 (Supplement A):3-20.-53-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyEFFECTS OF SALINITY ON PHOTOSYNTHESIS IN C 4 ESTUARINE GRASSESB.R. MARICLE 1,2 ,O.KIIRATS 1 ,R.W.LEE 1 , AND G.E. EDWARDS 11 School <strong>of</strong> Biological Sciences, Washington State University, Pullman, WA 99164-42362 Present address: Department <strong>of</strong> Biological Sciences, Fort Hays State University, Hays, KS 67601-4099; brmaricle@fhsu.eduThe effects <strong>of</strong> salinity on gross and net photosyn<strong>the</strong>sis rates were measured in estuarine C 4 grasses.Net CO 2 fixation was most affected by increasing salinity in Spartina patens and S. alterniflora,moderately affected in Distichlis spicata and S. densiflora, and unaffected in S. anglica. Spartinaanglica exhibited a decrease in internal carbon dioxide (CO 2 ) with increasing salinity, suggestingsome decrease in stomatal conductance. The results suggest S. anglica has a superior level <strong>of</strong> salttolerance compared to o<strong>the</strong>r species in <strong>the</strong> study, which could have implications for communitystructure in field sites or for invasive potential <strong>of</strong> S. anglica. The maximum quantum efficiency <strong>of</strong>CO 2 fixation, measured under limiting light, decreased with increasing salinity in S. alterniflora andS. patens, indicating an increase in leakage <strong>of</strong> CO 2 from <strong>the</strong> CO 2 pump. While carbon fixation decreasedunder increasing salinity in most species, fluorescence yield data showed <strong>the</strong>re was little effecton <strong>the</strong> use <strong>of</strong> solar energy in photochemistry. This indicates additional sinks are induced undersalinity for use <strong>of</strong> photochemically generated energy (e.g., increase in <strong>the</strong> CO 2 pump, photorespiration,or Mehler reaction). Also, in S. patens and D. spicata, which had moderate decreases in photosyn<strong>the</strong>sis,non-photochemical quenching (NPQ) mechanisms increased with salinity indicating some<strong>of</strong> <strong>the</strong> excess light energy was lost as heat. Therefore, excess excitation energy was diverted awayfrom <strong>the</strong> photosyn<strong>the</strong>tic reaction centers to prevent photoinhibition. Variable to maximal fluorescence(F V /F M ) ratios were not significantly decreased by increasing salinity, suggesting <strong>the</strong>re was nodamage to photosystem II (PSII) reaction centers in any species.Keywords: Spartina, Distichlis, chlorophyll fluorescence, gas exchange, salt stressINTRODUCTIONHigh and fluctuating salinity levels are characteristic <strong>of</strong>salt marshes. Elevated soil salinity can cause low waterpotentials, which can decrease stomatal conductance,reducing incoming CO 2 , and thus reducing photosyn<strong>the</strong>ticrates (Willmer 1983). One aim <strong>of</strong> <strong>the</strong> present study was todetermine how stomatal conductance and photosyn<strong>the</strong>sis inC 4 salt marsh grasses are affected by environmental salinity.Gross and net photosyn<strong>the</strong>sis rates were measured underthree salinity levels in <strong>the</strong> estuarine C 4 grasses Spartinaalterniflora, S. anglica, S. patens, S. densiflora, andDistichlis spicata in growth chamber studies. Differences in<strong>the</strong>se parameters were anticipated as a result <strong>of</strong> changes inphotosyn<strong>the</strong>sis as affected by salinity. Primary productionby Spartina is an important input into <strong>the</strong> estuarine system(Peterson et al. 1985), so understanding <strong>the</strong> photosyn<strong>the</strong>sisand hence production <strong>of</strong> this material may be important forunderstanding estuarine ecology and trophic interactions.MATERIALS AND METHODSPlants were collected at field sites in Washington (S.alterniflora, S. anglica, and D. spicata), Florida (S. patens),and Spain (S. densiflora). Tillers were potted in a 50/50volume to volume (v/v) sand/potting soil mixture, and plantswere watered to saturation twice weekly with modifiedHoagland solution (Epstein 1972). Growth chamberconditions were 14L/10D (light/dark hours) with 26°C daysand 18°C nights. Photosyn<strong>the</strong>tic photon flux density (PPFD)was 300 micromoles per square meter per second (μmol m -2s -1 ) at bench level. Flooded treatment plants were placed inlarge plastic tubs in a randomized block design. Water wasmaintained at 2 centimeters (cm) above <strong>the</strong> soil surface andwas completely replaced weekly. Salinity levels wereincreased 15 parts per thousand (‰) per week until floodedtreatments included 0, 15, and 30‰ salt (Instant Ocean salts;Aquarium Systems, Mentor, Ohio). Drained treatmentscontained 0‰ salt. Plants were held for 30 days under finaltreatment conditions before testing. There were at least threereplicate plants per species/treatment combination.Chlorophyll fluorescence was measured with an OS-500modulated fluorometer (Opti-Sciences, Inc.; Tyngsboro,MA). Gross photosyn<strong>the</strong>tic rates <strong>of</strong> O 2 evolution werecalculated from fluorescence yield measurements after Kralland Edwards (1992). The second-youngest leaf on each plantwas tested and light-response curves were generated forPPFD <strong>of</strong> 15-2000 μmol m -2 s -1 .A FastEst gas exchange system (Maricle et al. 2007) wasused to measure leaf gas exchange on high (30‰) and low(0‰) salinity plants. Intact leaves were enclosed in a leafchamber at 25°C and 25% relative humidity. Measures <strong>of</strong> CO 2uptake gave net photosyn<strong>the</strong>sis rates, and measures <strong>of</strong> externalwater vapor allowed calculations <strong>of</strong> stomatal conductance and-55-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartina<strong>the</strong> ratio <strong>of</strong> internal to atmospheric CO 2 concentration (c i /c a ).A CO 2 analyzer (Li-Cor 6251; Lincoln, Nebraska) was used tomeasure leaf photosyn<strong>the</strong>tic carbon uptake. Simultaneousmeasures <strong>of</strong> chlorophyll fluorescence (Walz PAM 101;Effeltrich, Germany) allowed calculations <strong>of</strong> grossphotosyn<strong>the</strong>tic rates <strong>of</strong> O 2 evolution. Light-response curveswere generated for PPFD <strong>of</strong> 0-1100 μmol m -2 s -1 . A 20-minutedark period was allowed before measuring F V /F M at 0 μmolm -2 s -1 . The initial slope <strong>of</strong> each plant’s light-response curve(under limiting light) was taken to be <strong>the</strong> quantum efficiency<strong>of</strong> CO 2 fixation or O 2 evolution (Genty et al. 1989).Light-response curves were compared between speciesand treatments using repeated measures analysis <strong>of</strong>covariance (ANCOVAR). Parameters like quantumefficiency, c i /c a , F V /F M , and NPQ were compared betweenspecies and treatments using analysis <strong>of</strong> variance (ANOVA).Treatments were blocked by tubs in all analyses.RESULTS AND DISCUSSIONMaximum rates <strong>of</strong> gross photosyn<strong>the</strong>sis (rates <strong>of</strong> O 2evolution) were quite high in this study (Fig. 1), consistentwith productivity data presented by Long and Woolhouse(1979) for Spartina species. There were no significantdifferences in gross photosyn<strong>the</strong>sis rates between species,treatment, or <strong>the</strong>ir interactions (ANCOVAR, p≥0.439).Maximum quantum efficiencies <strong>of</strong> O 2 evolution,measured under limiting light, were not significantlydecreased by salinity in any species (ANOVA, p≥0.107;Table 1). Values for gross quantum efficiencies in Spartinaand Distichlis were slightly higher than previously publishednet quantum efficiency measures for C 4 monocots.Values for quantum efficiencies <strong>of</strong> CO 2 fixation wereslightly lower than gross quantum efficiencies (Table 1) andwere similar to those reported by Ehleringer and Pearcy(1983) for C 4 monocots. Quantum efficiencies <strong>of</strong> CO 2fixation decreased in S. alterniflora and S. patens withincreased salinity (ANOVA, p≤0.058), but not in any o<strong>the</strong>rFig. 1. Light-response curves showing gross photosyn<strong>the</strong>sis rates (μmolO 2 evolved m -2 s -1 ) <strong>of</strong> Spartina and Distichlis plants in flooded or drainedsoil conditions and salt up to 30‰. Points are means <strong>of</strong> 3-23 plants ± SD.species (ANOVA, p≥0.161).There were large differences between gross and netphotosyn<strong>the</strong>sis rates in most species (Table 2). This resulted ina surplus <strong>of</strong> harvested light energy not used in CO 2 fixation.This energy must be dissipated in order to prevent damage tophotosyn<strong>the</strong>tic reaction centers. Net rates <strong>of</strong> photosyn<strong>the</strong>siswere lower in 30‰ salt compared to 0‰ salt in S. patens, S.alterniflora, and D. spicata (ANOVA, p≤0.063), but not in S.anglica or S. densiflora (ANOVA, p≥0.624).Maintaining gross photosyn<strong>the</strong>sis rates (as measured byfluorescence yield) in <strong>the</strong> light while rates <strong>of</strong> CO 2 fixationdecrease with increasing salinity indicates additionalTable 1. Photosyn<strong>the</strong>sis data collected for high- and low-salinity, flooded-treatment plants in <strong>the</strong> gas-exchange system. Shown is <strong>the</strong> net quantum efficiency <strong>of</strong> CO 2fixation, <strong>the</strong> gross quantum efficiency <strong>of</strong> O 2 evolution, <strong>the</strong> maximum amount <strong>of</strong> nonphotochemical quenching and c i/c a values at 1100 μmol m -2 s -1 , and <strong>the</strong> maximumF V/F M ratio <strong>of</strong> dark-adapted plants. The mean ± SD (n) is given for each species/treatment combination.Species TreatmentsalinityNet QE(CO 2 photon -1 )Gross QE(O 2 photon -1 )Max NPQ(unitless)c i/c a(unitless)Max F V/F M(unitless)S. alterniflora 0‰ 0.065 ± 0.020 (3) 0.066 ± 0.010 (3) 1.99 ± 0.24 (3) 0.53 ± 0.10 (3) 0.74 ± 0.01 (3)30‰ 0.026 ± 0.016 (3) 0.046 ± 0.010 (3) 1.80 ± 0.36 (3) 0.60 ± 0.15 (3) 0.67 ± 0.08 (3)S. anglica 0‰ 0.025 ± 0.018 (4) 0.049 ± 0.009 (4) 1.64 ± 0.33 (4) 0.61 ± 0.11 (4) 0.72 ± 0.04 (4)30‰ 0.027 ± 0.021 (4) 0.056 ± 0.009 (4) 1.63 ± 0.50 (4) 0.41 ± 0.08 (4) 0.72 ± 0.03 (4)S. densiflora 0‰ 0.049 ± 0.048 (4) 0.056 ± 0.014 (4) 1.78 ± 0.24 (4) 0.57 ± 0.17 (4) 0.74 ± 0.01 (4)30‰ 0.034 ± 0.025 (4) 0.063 ± 0.006 (4) 1.44 ± 0.12 (4) 0.35 ± 0.14 (4) 0.73 ± 0.01 (4)S. patens 0‰ 0.067 ± 0.018 (3) 0.068 ± 0.003 (3) 1.19 ± 0.08 (3) 0.56 ± 0.14 (3) 0.71 ± 0.01 (3)30‰ 0.033 ± 0.022 (3) 0.057 ± 0.007 (3) 1.70 ± 1.16 (3) 0.53 ± 0.19 (3) 0.72 ± 0.01 (3)D. spicata 0‰ 0.038 ± 0.002 (3) 0.050 ± 0.014 (3) 0.79 ± 0.23 (3) 0.53 ± 0.04 (3) 0.60 ± 0.10 (3)30‰ 0.023 ± 0.015 (4) 0.057 ± 0.006 (4) 1.39 ± 0.37 (4) 0.53 ± 0.19 (3) 0.69 ± 0.04 (4)-56-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 1: Spartina BiologyTable 2. Gross and net photosyn<strong>the</strong>tic rates collected for high- and lowsalinity,flooded-treatment plants at moderate PPFD (1100 μmol m -2 s -1 ).The mean gross photosyn<strong>the</strong>tic rate <strong>of</strong> O 2 evolution and net photosyn<strong>the</strong>ticrate <strong>of</strong> CO 2 fixation ± SD (n) is given for each species/treatmentcombination.Species SalinityGross PS rates Net PS rates(μmol O 2 m -2 s -1 ) (μmol CO 2 m -2 s -1 )S. alterniflora 0‰ 40.7 ± 1.0 (3) 11.0 ± 0.8 (3)30‰ 34.4 ± 9.0 (3) 2.9 ± 2.6 (3)S. anglica 0‰ 40.4 ± 5.5 (4) 2.2 ± 1.7 (4)30‰ 41.7 ± 2.6 (4) 5.9 ± 4.4 (4)S. densiflora 0‰ 38.3 ± 12.9 (4) 5.3 ± 7.2 (4)30‰ 46.3 ± 1.7 (4) 3.3 ± 11.4 (4)S. patens 0‰ 48.2 ± 2.5 (3) 16.1 ± 11.4 (3)30‰ 42.0 ± 1.6 (3) 3.0 ± 2.1 (3)D. spicata 0‰ 41.7 ± 13.2 (3) 10.3 ± 2.7 (3)30‰ 41.0 ± 3.3 (4) 6.3 ± 7.0 (4)electron sinks are induced. Some possibilities include(Demmig-Adams and Adams, 1992): CO 2 pump activitymay increase to compensate for bundle sheath CO 2 leakage;this would use additional ATP generated by <strong>the</strong> Mehlerreaction, and was potentially reflected in lower PSII yieldsin S. alterniflora and S. patens under high salinity (Table 1).Alternatively, an increase in photorespiration could helpsustain gross photosyn<strong>the</strong>sis rates if CO 2 levels drop in <strong>the</strong>bundle sheath allowing O 2 to react with RuBP. Aconsequence <strong>of</strong> photorespiration is producing products likePGA and ammonia that need reductive power fromphotochemistry. Increased reduction rates <strong>of</strong> nitrate, sulfate,or phosphate within chloroplasts could utilize excess lightenergy. Fur<strong>the</strong>r work will be needed to see if nitrate orsulfate reduction or photorespiration rates increase withincreasing salinity. Finally, energy not used inphotochemistry may be dissipated by nonphotochemicalquenching (NPQ) mechanisms, where excess light energy islost as heat. Maximum amounts <strong>of</strong> NPQ increased withsalinity in S. patens and D. spicata (ANOVA, p≤0.073), butin no o<strong>the</strong>r species (ANOVA, p≥0.570; Table 1).Under high light, rates <strong>of</strong> light-harvesting (grossphotosyn<strong>the</strong>sis) will invariably be larger than carbon fixation(net photosyn<strong>the</strong>sis) rates. Therefore excess energy must besafely dissipated before reaction centers are damaged(Demmig-Adams and Adams 1992). Dark-adapted F V /F Mratios <strong>of</strong> chlorophyll fluorescence can indicate damage to <strong>the</strong>PSII reaction center (Krause and Weis 1991). F V /F M ratioswere not significantly reduced by salinity in any species inthis study (ANOVA, p≥0.165; Table 1), suggesting excessenergy was efficiently dispersed before reaction centers weredamaged. Prevention <strong>of</strong> photoinhibition is likely to beimportant in determining plant resistance to environmentalstresses that reduce carbon fixation relative to lightharvesting rates (Demmig-Adams and Adams 1992).The c i /c a values <strong>of</strong> <strong>the</strong> plants in this study ranged from0.35 to 0.61 (Table 1). The c i /c a value <strong>of</strong> S. anglicasignificantly decreased with increasing salinity at moderatePPFD (ANOVA, p=0.032). The c i /c a values <strong>of</strong> all o<strong>the</strong>rspecies did not change in response to increasing salinity(ANOVA, p≥0.421). In previous studies, c i /c a values did notchange with increasing salinity in C 4 grasses (Bowman et al.1989, Meinzer et al. 1994). C 3 leaf c i /c a values tend to behigher than corresponding C 4 values and are more sensitiveto increasing salinity (Brugnoli and Lauteri 1991).Increasing salinity may decrease C 3 c i /c a values by as muchas 0.48 (Farquhar et al. 1982). In <strong>the</strong> present study, c i /c adecreased by 0.20 in S. anglica, and by 0.22 in S. densiflora(nonsignificant change) with 30‰ salinity, but only as muchas 0.07 in <strong>the</strong> o<strong>the</strong>r species (Table 1). We believe this to be<strong>the</strong> first report <strong>of</strong> salinity-induced decreases in C 4 c i /c avalues. One factor contributing to salt sensitivity in C 4 plantsmay be <strong>the</strong> susceptibility <strong>of</strong> CO 2 leakage from bundle sheathcells under increasing salinity. Increasing salinity appearedto decrease c i /c a in S. anglica, but it was <strong>the</strong> most resistantspecies to salinity in terms <strong>of</strong> CO 2 fixation. This suggestsCO 2 pump activity does not increase in S. anglica withincreasing salinity. Rates <strong>of</strong> photorespiration or nitratereduction may increase instead to use excess light energy.Excess energy dissipation in response to salinity may be anarea for future investigation in marsh halophytes.This work illustrated how light harvesting and CO 2uptake relate to sediment salinity in C 4 marsh grasses.Excess energy dissipation may also increase in times <strong>of</strong>salinity stress. Additionally, NPQ mechanisms increase inmany plants as a result <strong>of</strong> external salinity. Portableflourometers can easily be used in <strong>the</strong> field to determine invivo NPQ and thus in vivo salt stress in some plants.However, photochemistry does not appear to be affected bysalt, so fluorescence yield measurements in <strong>the</strong> light or darkadaptedF V /F M measures will not reflect salt stress. Incontrast, gas-exchange methods appear to be far moresensitive to salinity.ACKNOWLEDGMENTSThe authors thank C. Cody, P. Rabie, K. Patten, S.Hacker and M. Figueroa for assistance and plants. Thisproject was partially funded by <strong>the</strong> Betty W. HiginbothamTrust and a Padilla Bay NERR Research Assistantship. Thisresearch was also supported by NSF IBN0076604, EPA R-82940601, and NSF DBI-0116203.REFERENCESBowman, W.D., K.T. Hubick, S. von Caemmerer, and G.D. Farquhar.1989. Short-term changes in leaf carbon isotope discriminationin salt and water stressed C 4 grasses. Plant Physiol. 90:162-166.Brugnoli, E., and M. Lauteri. 1991. Effects <strong>of</strong> salinity on stomatalconductance, photosyn<strong>the</strong>tic capacity, and carbon isotope discrimination<strong>of</strong> salt-tolerant (Gossypium hirsutum L.) and saltsensitive(Phaseolus vulgaris L.) C 3 non-halophytes. PlantPhysiol. 95:628-635.-57-

Chapter 1: Spartina Biology<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaDemmig-Adams, B., and W.W. Adams, III. 1992. Photoprotectionand o<strong>the</strong>r response <strong>of</strong> plants to high light stress. Ann. Rev. PlantPhysiol. Plant Mol. Biol. 43:599-626.Ehleringer, J., and R.W. Pearcy. 1983. Variation in quantum yieldfor CO 2 uptake among C 3 and C 4 plants. Plant Physiol. 73:555-559.Epstein, E. 1972. Mineral Nutrition <strong>of</strong> Plants: Principles and Perspectives.John Wiley and Sons, Inc.: New York.Farquhar, G.D., M.C. Ball, S. von Caemmerer, and Z. Roksandic.1982. Effect <strong>of</strong> salinity and humidity on δ 13 C value <strong>of</strong> halophytes—Evidencefor diffusional isotope fractionation determinedby <strong>the</strong> ratio <strong>of</strong> intercellular/atmospheric partial pressure <strong>of</strong>CO 2 under different environmental conditions. Oecologia52:121-124.Genty, B., J.-M. Briantais, and N.R. Baker. 1989. The relationshipbetween <strong>the</strong> quantum yield <strong>of</strong> photosyn<strong>the</strong>tic electron transportand quenching <strong>of</strong> chlorophyll fluorescence. Biochim. Biophys.Acta 990:87-92.Krall, J.P., and G.E. Edwards. 1992. Relationship between photosystemII activity and CO 2 fixation in leaves. Physiol. Plant.86:180-187.Krause, G.H., and E. Weis. 1991. Chlorophyll fluorescence andphotosyn<strong>the</strong>sis: <strong>the</strong> basics. Ann. Rev. Plant Physiol. Plant Mol.Biol. 42:313-349.Long, S.P., and H.W. Woolhouse. 1979. Primary production inSpartina marshes. In: R.L. Jefferies and A.J. Davy, eds. EcologicalProcesses in Coastal Environments, 333-352. Blackwell ScientificPublications: Oxford.Maricle, B.R., R.W. Lee, C.E. Hellquist, O. Kiirats, and G. Edwards.2007. Effects <strong>of</strong> salinity on chlorophyll fluorescence andCO 2 fixation in C 4 estuarine grasses. Photosyn<strong>the</strong>tica 45:433-440.Meinzer, F.C., Z. Plaut, and N.Z. Saliendra. 1994. Carbon isotopediscrimination, gas exchange, and growth <strong>of</strong> sugarcane cultivarsunder salinity. Plant Physiol. 104:521-526.Peterson, B.J., R.W. Howarth, and R.H. Garritt. 1985. Multiplestable isotopes used to trace <strong>the</strong> flow <strong>of</strong> organic matter flow inestuarine food webs. Science 227:1361-1363.Willmer, C.M. 1983. Stomata. Longman: London.-58-

CHAPTER TWOSpartina Distribution and Spread

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadATALE OF TWO INVADED ESTUARIES: SPARTINA IN SAN FRANCISCO BAY,CALIFORNIAAND WILLAPA BAY,WASHINGTOND.R. STRONGDepartment <strong>of</strong> Ecology and Evolution, University <strong>of</strong> California, Davis, Davis, CA 95616; drstrong@ucdavis.eduMaritime Spartina species grow lower on <strong>the</strong> tidal plane than o<strong>the</strong>r vascular plants and maintain <strong>the</strong>shoreline on temperate coasts where <strong>the</strong>y are native. All but two <strong>of</strong> <strong>the</strong> 14 known species are nativeto <strong>the</strong> Atlantic. Spartina alterniflora was introduced a century ago into Willapa Bay, Washington,far north <strong>of</strong> <strong>the</strong> native limit <strong>of</strong> this genus. This Atlantic native spread exponentially throughtidelands <strong>the</strong>re at a remarkably constant approximately 12% per year over <strong>the</strong> 55-year history <strong>of</strong>aerial photographs. In 2000, it covered approximately 1,670 <strong>of</strong> 6,000 hectares (ha) or 27% <strong>of</strong> <strong>the</strong>intertidal habitat <strong>of</strong> Willapa Bay. However, <strong>the</strong> rate <strong>of</strong> spread was slowed greatly by an Allee effectdue to poor pollen dispersal. Without <strong>the</strong> Allee effect, <strong>the</strong> invasion would have covered <strong>the</strong> lion’sshare <strong>of</strong> Willapa Bay long ago. Large-scale chemical control is now greatly reducing S. alterniflorain Willapa Bay. The San Francisco Bay region is <strong>the</strong> nor<strong>the</strong>rn limit <strong>of</strong> Spartina foliosa, one <strong>of</strong> twoSpartina species native to <strong>the</strong> Pacific. Introduced Spartina played virtually no role in San FranciscoBay until 1975, when <strong>the</strong> U.S. Army Corps <strong>of</strong> Engineers planted S. alterniflora, which hybridizedwith <strong>the</strong> native S. foliosa soon afterwards. While S. alterniflora has become virtually extinct, <strong>the</strong>hybrids have had a truly phenomenal rate <strong>of</strong> spread. The few hybrids that formed in <strong>the</strong> late 1970sspread to about 1,500 ha when <strong>the</strong> San Francisco Estuary Invasive Spartina Project (ISP) began <strong>the</strong>ircontrol effort. The rapid spread <strong>of</strong> hybrids is probably due to <strong>the</strong> evolution <strong>of</strong> self-pollination, thuseliminating <strong>the</strong> Allee effect. Ultimate success <strong>of</strong> <strong>the</strong> ISP will depend upon a sophisticatedcombination <strong>of</strong> biochemical systematics with ecological field research that determines dynamics <strong>of</strong>cryptic hybrids that could survive control efforts.Key Words: Hybrid, rate <strong>of</strong> spreadSpartinas are cordgrasses (Strong and Ayres 2009). Thespecies <strong>of</strong> estuarine cordgrasses that we have studied arewind-pollinated, largely self-incompatible and outbreeding.They are protogynous; <strong>the</strong> female flowers appear before <strong>the</strong>male flowers. In order to set much viable seed, each plantrequires pollen to be carried on <strong>the</strong> wind from a different,earlier-flowering plant that has progressed to <strong>the</strong> later stage<strong>of</strong> having male, pollen-bearing flowers.Cordgrasses are ecosystem engineers. Their tall densestems slow water movement and cause sediment to settle andbe bound by thick, fibrous roots. Roots grow upwardthrough <strong>the</strong> settling sediment to form thick peat that elevates<strong>the</strong> surface <strong>of</strong> <strong>the</strong> marsh. Cordgrasses that invade areas withno emergent vegetation greatly increase local photosyn<strong>the</strong>ticrates. Their roots greatly increase subsurface carbon whichremains in <strong>the</strong> anoxic sediment long after <strong>the</strong> plants havebeen removed. Invasive cordgrasses have transformed vastexpanses <strong>of</strong> open intertidal mudflat into meadows thatelevate with time. Estuarine cordgrasses disperse primarilyby floating seed that does not accumulate in soil. Noevidence indicates a seed bank more than one year old forestuarine cordgrasses.All but one <strong>of</strong> <strong>the</strong> 14 nominal species <strong>of</strong> cordgrass arenative to <strong>the</strong> Americas: Spartina maritima is endemic to <strong>the</strong>south <strong>of</strong> England and France (Daehler and Strong 1997).The Pacific has but two natives, Spartina densiflora in Chileand Spartina foliosa, California cordgrass, that thrives inBaja California and in San Francisco Bay. Peoplepurposefully and inadvertently spread cordgrasses, and fournon-native Spartinas have been introduced to San FranciscoBay.Spartina patens is <strong>the</strong> least successful San Franciscoinvader and is known from only two plants in Suisun Bay(Ayres et al. 2004). Native to Atlantic marshes <strong>of</strong> NorthAmerica, S. patens spread rapidly at one site in Oregonduring <strong>the</strong> mid-20th century and it has appeared recently inSpain. With somewhat greater success, S. densiflora spreadto several sites in San Francisco Bay after at least twointroductions in Marin County, where it was brought from<strong>the</strong> huge, century-old Humboldt Bay, California infestation.This species has also been introduced into Spain.Spartina anglica arose in England in <strong>the</strong> 19th century asa hybrid <strong>of</strong> S. maritima and S. alterniflora, after <strong>the</strong> latterspecies was introduced from its native Atlantic shores <strong>of</strong>North America. Spartina anglica spread widely afterintroduction to Puget Sound, Washington, <strong>the</strong> Ne<strong>the</strong>rlands,Tasmania, Australia, and New Zealand. However, S. anglicahas not spread to o<strong>the</strong>r places in San Francisco Bay postintroductionin 1977 to Creekside Park in Greenbrae, MarinCounty.In misguided attempts at marsh restoration in <strong>the</strong> mid-1970s, <strong>the</strong> fourth species, Spartina alterniflora, was-61-

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaintroduced twice into south San Francisco Bay. Seed from aMaryland marsh was sown in New Alameda Creek, locatedbetween Fremont and Union City (Faber 2000). The secondknown introduction, a planting sponsored by <strong>the</strong> ArmyCorps <strong>of</strong> Engineers at Alameda Island, occurred about 40kilometers (km) north <strong>of</strong> <strong>the</strong> first (Ayres et al. 2003). As aresult <strong>of</strong> lack <strong>of</strong> pollen from conspecific plants and pollenswamping from S. alterniflora x S. foliosa hybrids, S.alterniflora has become quite rare in San Francisco Bay in<strong>the</strong> two or three decades since introduction.One or both <strong>of</strong> <strong>the</strong> S. alterniflora introductions resultedin hybridization with <strong>the</strong> native California cordgrass toproduce a backcrossing swarm <strong>of</strong> S. alterniflora x S. foliosahybrids in San Francisco Bay. The chloroplast DNA <strong>of</strong> bothparental species is found among <strong>the</strong> plants in <strong>the</strong> swarmindicating that both parental species have served as seedparents <strong>of</strong> hybrids. While hybridization probably hasoccurred multiple times F1 hybrids are rare in <strong>the</strong> field. Withgreat effort in <strong>the</strong> greenhouse we have produced a few F1hybrids.A subset <strong>of</strong> hybrid genotypes are extremely fit in one ora combination <strong>of</strong> <strong>the</strong> following traits: vegetative growth rate,numbers <strong>of</strong> viable seed, volume <strong>of</strong> pollen, and/or selfcompatibility.These hybrid traits can be transgressive,which means that hybrids exceed both parental species in <strong>the</strong>magnitude <strong>of</strong> <strong>the</strong> trait. We entertain <strong>the</strong> hypo<strong>the</strong>sis that <strong>the</strong>most important transgressive trait <strong>of</strong> hybrids is selfcompatibility,which allows a plant to pollinate itself and setseed at low density after invasion.A great deal <strong>of</strong> hybrid S. alterniflora x S. foliosa seed iscarried on <strong>the</strong> currents and tides around San Francisco Bayand new marshes are invaded every year. Hybrid seed floatsto open mud flats, germinates, and grows rapidly. This seedalso spreads into vegetated marshes comprised <strong>of</strong> S. foliosaand o<strong>the</strong>r native marsh plants. Hybrid invasion <strong>of</strong> vegetativemarshes leads to severe ecological and genetic competitionwith native S. foliosa. Hybrid cordgrass is increasing atgreater than exponential rates in San Francisco Bay. In 2002,approximately 1,500 ha <strong>of</strong> salt marsh was dominated by <strong>the</strong>swarm in San Francisco Bay (Ayres et al. 2004). The mostrecent pair <strong>of</strong> censuses yields a doubling time <strong>of</strong> about threemonths for coverage <strong>of</strong> <strong>the</strong> hybrid swarm in <strong>the</strong> Bay.The loss <strong>of</strong> native cordgrass due to competition andinterbreeding is accelerating. We can conceive <strong>of</strong> no naturallimitation to this loss, which is a runaway, unregulatedprocess that could lead to <strong>the</strong> extinction <strong>of</strong> S. foliosa in SanFrancisco Bay. Similarly, dispersal <strong>of</strong> hybrids to saltmarshes in Baja California could lead to <strong>the</strong> extinction <strong>of</strong> S.foliosa <strong>the</strong>re (Ayres et al. 2003). In contrast, Willapa Bay,Washington has no native cordgrass and no potential forhybridization. Spartina alterniflora was introduced <strong>the</strong>remore than 100 years ago (Civille et al. 2005) and probablyarrived as a hitchhiker on <strong>the</strong> numerous trains from NewYork harbor that brought oysters for outplanting in WillapaBay at <strong>the</strong> end <strong>of</strong> <strong>the</strong> 19th and beginning <strong>of</strong> <strong>the</strong> 20thcenturies. The first hard historical evidence <strong>of</strong> <strong>the</strong> invasion isfrom a photo and publication in 1941. The large patch size<strong>of</strong> <strong>the</strong> plant implies that it had already been growing forseveral decades. The first aerial photos, from 1945, showseveral large colonies, fur<strong>the</strong>r evidence that introductionoccurred decades earlier. The multiple, widely separatedcolonies imply multiple introductions ra<strong>the</strong>r than spreadfrom a single focus. Coverage <strong>of</strong> S. alterniflora hasincreased at a rate very close to exponential between 1945and 2000, about 12% per year. This gives a doubling time <strong>of</strong>about six years. Approximately 1,670 ha (27%) <strong>of</strong> <strong>the</strong> 6,000ha <strong>of</strong> intertidal habitat <strong>of</strong> Willapa Bay had been colonized byS. alterniflora by 2000. Large-scale chemical control is nowgreatly reducing S. alterniflora in Willapa Bay (Strong andAyres 2009).In an echo <strong>of</strong> <strong>the</strong> slow spread <strong>of</strong> S. alterniflora in SanFrancisco Bay, we have found that very little viable seed isset at <strong>the</strong> leading edge <strong>of</strong> <strong>the</strong> invasion <strong>of</strong> Willapa Bay (Daviset al. 2004a). Seed settles at low densities onto open mudand recruits are widely separated from one ano<strong>the</strong>r. Decadeslater, when <strong>the</strong> circular clones coalesce to form continuousmeadows, seed set increases by an order <strong>of</strong> magnitude. Thisis a weak Allee effect. A strong Allee effect would result ifno seed is set by plants at low densities, indicating that S.alterniflora would have become extinct at <strong>the</strong> low densities<strong>of</strong> <strong>the</strong> initial invasion. The weak Allee effect has slowed <strong>the</strong>rate <strong>of</strong> invasion. Were <strong>the</strong>re no Allee effect, S. alterniflorawould have to be able to self-pollinate, and single plantswould set seed at <strong>the</strong> same rate as plants growing in highdensities. Without an Allee effect, <strong>the</strong> rate <strong>of</strong> increase incoverage <strong>of</strong> S. alterniflora would have been about 30% peryear, and <strong>the</strong> doubling time would have been as short as 2.5years instead <strong>of</strong> <strong>the</strong> actual six years (Taylor et al. 2004).Thus, without <strong>the</strong> Allee effect, <strong>the</strong> approximately three ha <strong>of</strong>S. alterniflora shown in <strong>the</strong> 1945 aerial photos would havegrown to completely cover <strong>the</strong> entire 19,000 ha <strong>of</strong> intertidalarea <strong>of</strong> Willapa Bay by about 1977.The dearth <strong>of</strong> viable seed set at <strong>the</strong> leading edge <strong>of</strong> <strong>the</strong>invasion <strong>of</strong> S. alterniflora in Willapa Bay and <strong>the</strong> weakAllee effect are a result <strong>of</strong> a lack <strong>of</strong> pollen produced by lowdensityplants (Davis et al. 2004b). We found ninefold morepollen on stigmas <strong>of</strong> high-density plants in old marshes thanon those <strong>of</strong> low-density plants at <strong>the</strong> leading edge <strong>of</strong> <strong>the</strong>invasion. Only in old marshes, where plants had growntoge<strong>the</strong>r to form dense meadows, was <strong>the</strong>re sufficient pollenon stigmas for much seed set. Experimental pollinationaugmentation <strong>of</strong> low-density plants, but not <strong>of</strong> high-densityplants, increased seed set. Experimental pollen exclusionfrom high-density plants, but not from low-density plants,decreased seed set.In summary, <strong>the</strong> cordgrass invasion <strong>of</strong> San FranciscoBay is by hybrids <strong>of</strong> S. alterniflora and S. foliosa, not by S.alterniflora alone. The hybrid has a truly phenomenal rate <strong>of</strong>spread, and this rate is accelerating. This is evidence that <strong>the</strong>hybrids are much more invasive than S. alterniflora alone.-62-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadThe first 40 or 50 years <strong>of</strong> spread in Willapa Bay producedless than 10 ha <strong>of</strong> S. alterniflora. Beginning with a verysmall number <strong>of</strong> hybrids in <strong>the</strong> 1970s or 1980s, <strong>the</strong> overallcoverage <strong>of</strong> <strong>the</strong> hybrid has grown to about 1,500 ha.At a rate <strong>of</strong> increase <strong>of</strong> about 12% per year, similar to S.alterniflora in Willapa Bay, San Francisco Bay would haveonly about 3.4 times <strong>the</strong> area <strong>of</strong> <strong>the</strong> initial introductionssome 30 years later. If <strong>the</strong> two introductions <strong>of</strong> S.alterniflora had amounted to about one ha <strong>of</strong> S. alterniflora,<strong>the</strong>re would be only 3.4 ha now. I speculate that withouthybridization, this conference would not have occurred.Transgressive traits confer greater fitness upon a subset<strong>of</strong> hybrids than ei<strong>the</strong>r parent exhibits. We entertain <strong>the</strong>hypo<strong>the</strong>sis that <strong>the</strong> key trait is increased self-compatibility <strong>of</strong>some hybrids. This would erase <strong>the</strong> weak Allee effect thatwe see in Willapa Bay as increased self-compatibility allowssingle plants to set abundant seed.The presence <strong>of</strong> hybrids means that control <strong>of</strong> nonnativecordgrass in San Francisco Bay will require researchand understanding beyond that required in Willapa Bay.Biochemical systematics are needed to detect hybrids at lowfrequency, in order that hybrids can be removed before <strong>the</strong>marsh is overrun by <strong>the</strong>m. Strategies concerning which andhow much hybrid cordgrass to eliminate for effective controlwill require understanding <strong>of</strong> <strong>the</strong> dynamic mechanisms <strong>of</strong>hybrid spread. Research on <strong>the</strong>se subjects is just beginning.REFERENCESAyres, D.R., D. Garcia-Rossi, H.G. Davis, and D.R. Strong. 1999.Extent and degree <strong>of</strong> hybridization between exotic (Spartinaalterniflora) and native (S. foliosa) cordgrass (Poaceae) inCalifornia, USA determined by Random Amplified PolymorphicDNA (RAPDs). Molecular Ecology 8:1179-1187.Ayres, D.R., D.R. Strong, and P. Baye. 2003. Spartina foliosa – acommon species on <strong>the</strong> road to rarity? Madroño 50: 209-213.Ayres, D.R., D.L. Smith, K. Zaremba, S. Klohr, and D.R. Strong.2004. Spread <strong>of</strong> exotic cordgrasses and hybrids (Spartina sp.) in<strong>the</strong> tidal marshes <strong>of</strong> San Francisco Bay, California, USA.Biological Invasions 6: 221–231.Civille, J.C., K. Sayce, S. Smith, and D.R. Strong. 2005.Reconstructing a century <strong>of</strong> Spartina invasion with historicalrecords and contemporary remote sensing. Ecoscience 12:330-338.Daehler, C.C. and D.R. Strong. 1997. Hybridization betweenintroduced smooth cordgrass (Spartina alterniflora; Poaceae)and native California cordgrass (S. foliosa) in San Francisco Bay,California, USA. American Journal <strong>of</strong> Botany 84:607-611.Davis, H.G., C.M. Taylor, J.C. Civille, and D.R. Strong. 2004a. AnAllee effect at <strong>the</strong> front <strong>of</strong> a plant invasion: Spartina in a Pacificestuary. Journal <strong>of</strong> Ecology 92:321-327.Davis, H.G, C.M. Taylor, J.G. Lambrinos, and D.R. Strong. 2004b.Pollen limitation causes an Allee effect in a wind-pollinatedinvasive grass (Spartina alterniflora). <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong>National Academy <strong>of</strong> Sciences USA 101: 13804-13807.Faber, P. 2000. Grass wars: good intentions gone awry. CaliforniaCoast & Ocean 16, #2. Summer 2000. 1-5.Strong, D. R. and D.R. Ayres. 2009. Spartina introductions andconsequences in salt marshes: arrive, survive, thrive, andsometimes hybridize. In B.R. Silliman, E.D. Grosholz, and M.D.Bertness, eds. Human impacts on salt marshes: a globalperspective. Berkeley, California; London: University <strong>of</strong>California Press.Taylor, C.M., Davis, H.G., Civille, J.C., Grevstad, F.S. and A.Hastings. 2004. Consequences <strong>of</strong> an Allee effect on <strong>the</strong> invasion<strong>of</strong> a Pacific estuary by Spartina alterniflora. Ecology. 85:3254-3266.-63-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadSPARTINA IN CHINA:INTRODUCTION,HISTORY,CURRENT STATUS, AND RECENT RESEARCHS. AN,H.QING,Y.XIAO,C.ZHOU,Z.WANG,Z.DENG,Y.ZHI AND L. CHENSchool <strong>of</strong> Life Science, Nanjing University, Nanjing 210093, China; anshq@nju.edu.cnSpartina spp., with <strong>the</strong>ir strong ability for survival, growth, and expansion, attracts ecologists’attention throughout <strong>the</strong> world. Four Spartina species: Spartina anglica, S. alterniflora, S. patensand S. cynosuroides were introduced into China in 1963, 1979, 1979 and 1998, respectively, for <strong>the</strong>purposes <strong>of</strong> agricultural and ecological engineering. So far <strong>the</strong> first three species still survive incoastal China with varying abundance. Spartina anglica is dying back, whereas S. alterniflora hasbeen rampantly invading. Spartina patens appears to be a potential invasive species in China. Weexamine <strong>the</strong> history <strong>of</strong> Spartina’s introductions into China, discuss <strong>the</strong> fate <strong>of</strong> <strong>the</strong> introduced speciesand <strong>the</strong>ir impacts on native ecosystems, and review <strong>the</strong> studies conducted in China over <strong>the</strong> past 40years (from 1963-2003), especially those carried out in <strong>the</strong> Institute <strong>of</strong> Spartina and TidelandDevelopment at Nanjing University.Keywords: Ecological engineering, genetic structure, invasive impacts, management, SpartinaINTRODUCTIONFour Spartina species have been introduced into Chinaas bio-engineers for agricultural and ecological engineeringsince 1963. They were used to accelerate <strong>the</strong> development<strong>of</strong> coastal tidelands for croplands, protect <strong>the</strong> dykes fromtyphoons and control erosion <strong>of</strong> tidelands from tidal waves(Chung and Zhuo 1979; Chung 1982, 1985). The specieshave had different fates in coastal areas <strong>of</strong> China resultingfrom both human activities and natural stresses. Seeds <strong>of</strong> S.patens and S. cynosuroides were introduced into China,toge<strong>the</strong>r with S. alterniflora. But, S. patens failed togerminate in <strong>the</strong> laboratory and S. cynosuroides failed tosurvive in <strong>the</strong> field. In 1998, seeds <strong>of</strong> S. patens wereintroduced into China again and successfully germinatedafter ion radiation treatment (Zhou 2003), <strong>the</strong>n tissue culturewas used to produce <strong>of</strong>fspring at a large scale. So far, about20 hectares (ha) <strong>of</strong> S. patens has been transplanted inTianjing, Jiangsu and Zhejiang Provinces (Zhou et al 2003).Spartina alterniflora has rapidly spread to o<strong>the</strong>r coastalareas in China, outcompeted native plants and become one<strong>of</strong> most harmful invading plants in China. In contrast, S.anglica is experiencing a dieback and may disappear from<strong>the</strong> coasts. Although S. anglica is a serious invasive speciesin o<strong>the</strong>r countries (Daehler and Strong 1996; Kriworken andHedge 2000; Baumel et al. 2001; Hacker et al. 2001), it isnow not an invasive species in China. More recently, S.patens has shown <strong>the</strong> potential to be an invasive species(Zhou et al. 2003). The four species have had differentdifferent fates including <strong>the</strong> failure <strong>of</strong> S. cynosuroidesintroduction and different invasive processes in coastalChina. Here we give an brief account <strong>of</strong> <strong>the</strong> experiencesfrom Chinese scientists and studies that have been conductedin China since <strong>the</strong> introduction <strong>of</strong> <strong>the</strong>se species to thiscountry. It is hoped that <strong>the</strong> information presented here canbe <strong>of</strong> some use to <strong>the</strong> scientists who are working on use,management, control and eradication <strong>of</strong> <strong>the</strong> species that areinvading both Atlantic and Pacific coastal areas worldwide.HISTORY OF SPARTINA INTRODUCTION IN CHINASpartina anglicaSpartina anglica was <strong>the</strong> first Spartina speciesintroduced into China from ano<strong>the</strong>r country. In 1963, 35plants from Essex in England and 100 plants from Hojer inDenmark were introduced to <strong>the</strong> ecological laboratory <strong>of</strong>Nanjing University by Pr<strong>of</strong>. Chung-Hsin Chung (Chong-XinZhong) with <strong>the</strong> help <strong>of</strong> <strong>the</strong> Chinese Committee <strong>of</strong> Scienceand Technology. The Spartina survivors — 21 from Englandand 50 from Denmark — were sent to SheyangExperimental Station (35 o N) on <strong>the</strong> Yellow Sea coast after asimple survivial check in <strong>the</strong> laboratory. The survivors <strong>of</strong> <strong>the</strong>English Spartina successfully produced 435,000 new rametsfrom July 1963 to April 1964, which were planted in <strong>the</strong>field in 1964; by 1966 <strong>the</strong> area <strong>of</strong> <strong>the</strong> plantation hadincreased to 32 ha. However, <strong>the</strong> Danish population soondied <strong>of</strong>f. Almost all <strong>of</strong> <strong>the</strong> S. anglica in China aredescendants <strong>of</strong> <strong>the</strong> original 21 English plants, excluding <strong>the</strong>plantations in Xiaoshan and Wenlin <strong>of</strong> Zhejiang Province. In1964, 507 seeds and 18 plants from Poole Harbor in Englandwere introduced to China, <strong>of</strong> which 440 seeds with potentialgermination capacity were treated and sowed in <strong>the</strong>laboratory; 157 <strong>of</strong> <strong>the</strong>m germinated, a germination rate <strong>of</strong>30.97%, and 44 seedlings survived. The 18 plants died <strong>of</strong>f in<strong>the</strong> lab. One year later, 30,601 individuals were obtainedfrom <strong>the</strong> 44 surviving seedlings (Fig. 1), and most <strong>of</strong> <strong>the</strong>mwere planted in Xiaoshan and Wenlin Experimental Stations-65-

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaalong <strong>the</strong> coast <strong>of</strong> <strong>the</strong> East Sea (29 o N). In <strong>the</strong> experiment,one rhizome segment produced 9.10 million individualsduring 29 months, and four ramets produced 7.60 millionindividuals during 33 months. Most <strong>of</strong> <strong>the</strong> 16.70 millionindividuals were planted in mid-1966 in two East Seastations, whose area was 73 ha. The total area <strong>of</strong> S. anglicaplantation was 78 ha by August 1966 in <strong>the</strong> two stations.But, from 1973 to 1980, most <strong>of</strong> <strong>the</strong> plantations in <strong>the</strong> twostations were reclaimed for croplands; thus almost all <strong>the</strong>individuals <strong>of</strong> S. anglica died <strong>of</strong>f <strong>the</strong>re.Four Spartina species: Spartina anglica, S. alterniflora,S. patens and S. cynosuroides have been introduced intoChina in 1963, 1979, 1979 and 1998, respectively, for <strong>the</strong>purposes <strong>of</strong> agricultural and ecological engineering.By 1978,all <strong>the</strong> coastal provinces <strong>of</strong> China planted S. anglica. Therewere 6,330 ha, 3,750 ha and 200 ha <strong>of</strong> S. anglica plantationsin Jiangsu, Zhejiang and Shandong Provinces, respectively.Meanwhile, 5 ha, 2/3 ha, 1/2 ha and 1/3 ha existed in Hebei,Liaoning, Tianjing and Guangdong Provinces, respectively.Less than 1/3 ha <strong>of</strong> <strong>the</strong> species was established in Shanghai,Fujian and Guangxi Provinces. In <strong>the</strong> field, S. anglica had astrong reproductive capacity by asexual propagation throughramets and rhizomes (Fig. 2). By 1980, all 83 counties <strong>of</strong>coastal China had S. anglica with a total area <strong>of</strong> 31,590 ha;by 1985 area had increased to 36,000 ha, although <strong>the</strong> sexualreproduction <strong>of</strong> <strong>the</strong> species was very low.Spartina alternifloraSimilarly, seeds and individuals <strong>of</strong> S. alterniflora werealso introduced to China by Pr<strong>of</strong>essor Chung at <strong>the</strong> end <strong>of</strong>1979 from North Carolina, Georgia and Florida, USA.Seedlings were successfully obtained from <strong>the</strong> seeds andnew ramets were reproduced from <strong>the</strong> exotic individuals in<strong>the</strong> laboratory and garden <strong>of</strong> Nanjing University in 1980,and <strong>the</strong>n were sent to Luoyuanwan Station (26 0 30’). TheTable 1. Growth <strong>of</strong> one-year Spartina alterniflora from three differentorigins. (Data from Xu & Zhuo 1985.)Culm height(cm)Leaf length(cm)Leaf width(cm)Ear length(cm)SourcesRange Max Range Max Range Max Range MaxBiomass(DW g/m 2)LeafcolorNorthCarolina 110-170 217 45-70 95.5 1.3-1.5 2.0 14-30 48 297.9 BlackgreenLightGeorgia 140-240 275 50-70 90.0 1.4-1.7 2.1 18-35 42 457.2greenFlorida 70-100 128 40-60 82.0 1.5-1.7 2.1 12-15 30 268.2 GreenTable 2. Height and biomass growth <strong>of</strong> Spartina alterniflora fromdifferent origins at <strong>the</strong> field plots (Xu & Zhuo 1985).Fig. 1 Asexual reproduction <strong>of</strong> exotic Spartina anglica from seedlings.(Data from Chung et al 1985.)IntroductiveSourcesCulmheight(cm)Above-groundbiomass(g DW/m 2)Below-groundbiomass(g DW/m 2)Total biomass(g DW/m 2)North Carolina 170 (35) 1289 (136) 1235 (160) 2524 (274)Georgia 240 (50) 2745 (502) 1366 (25) 4111 (486)Florida 110 (25) 972 (139) 928 (100) 1900 (83)ANOVA & T-testP

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and Spreadplants were planted in <strong>the</strong> field in <strong>the</strong> spring <strong>of</strong> 1981 into anarea <strong>of</strong> 1,000-1,300 m 2 . However, most <strong>of</strong> <strong>the</strong> Spartina saltmarsh <strong>of</strong> coastal China was developed from 0.5 kg <strong>of</strong> seedsfrom North Carolina.( Chung et al. 1985) Three morphs <strong>of</strong> S.alterniflora were identified by Xu & Zhou (1985) and Qin etal. (1985) based on <strong>the</strong>ir morphological data (culm height,leaf size, ear length and biomass), respectively, collectedfrom <strong>the</strong> garden <strong>of</strong> Nanjing University (Table 1) and fieldplots in Luoyuan Bay <strong>of</strong> East Sea (Table 2). Both <strong>the</strong>experiments showed that Georgia Spartina had <strong>the</strong> tallestand strongest culms, and <strong>the</strong> highest biomass. Meanwhile,Chen and Chung (1991) reported that <strong>the</strong>re were threegenetic types in <strong>the</strong>ir studies <strong>of</strong> isozymes <strong>of</strong> EST, MDH andPOD <strong>of</strong> <strong>the</strong> seedlings (Table 3). Since <strong>the</strong> people liked toplant higher and stronger individuals <strong>of</strong> <strong>the</strong> species, most <strong>of</strong><strong>the</strong> S. alterniflora plantations in China are now tall form(Georgia population) S. alterniflora. The tall form Spartinahad significant variations in survivorship, culm height andbiomass along <strong>the</strong> elevational gradients <strong>of</strong> <strong>the</strong> Chinese coast(Chen & Chung 1991).IMPORTANT HUMAN FORCES DRIVING SPARTINAINVASIONSIn 1962, <strong>the</strong> Chinese Government decided to introduceSpartina from its native habitats for agricultural goalsaccording to <strong>the</strong> scientists’ recommendations. In 1963, S.anglica was introduced from Europe by <strong>the</strong> Chinesegovernment. In 1969, <strong>the</strong> first workshop was held by <strong>the</strong>government <strong>of</strong> Zhejiang Province to accelerate development<strong>of</strong> S. anglica marshes. From 1973 to 1978, severalworkshops were held by <strong>the</strong> governments <strong>of</strong> Jiangsu andShandong provinces for <strong>the</strong> same purposes. Here we givesome <strong>of</strong> <strong>the</strong> events that are associated with Spartina inchronological order:• 1978: Spartina research group <strong>of</strong> Nanjing University received<strong>the</strong> National Award for Natural Science <strong>of</strong>China.• 1978: The Institute <strong>of</strong> Spartina and Tideland Developmentat Nanjing University was founded by China’sMinistry <strong>of</strong> Education, toge<strong>the</strong>r with China’s Committee<strong>of</strong> Science and Technology, China’s Ministry <strong>of</strong> Agricultureand China’s Ocean Bureau. The institute consisted<strong>of</strong> 17 members.• 1978-1979: Two workshops were held by China’sCommittee <strong>of</strong> Science and Technology to accelerate introduction<strong>of</strong> S. anglica.• 1979: Three o<strong>the</strong>r Spartina species were introduced byPr<strong>of</strong>essor Chung-Hsin Chung, funded by <strong>the</strong> Chinesegovernment.• 1985: The first Chinese Spartina research monograph“Research Advances in Spartina — Achievement <strong>of</strong> <strong>the</strong>Past 22 Years”, was published as a special issue <strong>of</strong> <strong>the</strong>Journal <strong>of</strong> Nanjing University; <strong>the</strong> monograph containedthree reviews, 26 reports and 24 short communications.• 1985-1990: Several honors were awarded to Pr<strong>of</strong>.Chung by different Chinese government departments forhis distinguished achievement in introduction <strong>of</strong>Spartina.• 1992: A monograph, Applied Studies on Spartina (eds.Qin, P. and C.H. Chung), was published by OceanPress; it contained 30 papers.• 1993: Pr<strong>of</strong>. Chung received an award from <strong>the</strong> Society<strong>of</strong> Wetland Scientists.• 1995: Pr<strong>of</strong>. Chung received a Distinguished FellowAward from Ohio State University.• 1996: The Spartina Research Group <strong>of</strong> Nanjing Universityreceived a Distinguished Contribution Award from<strong>the</strong> <strong>International</strong> Society <strong>of</strong> Ecological Engineering.• 1998: S. patens was again introduced into China as asalt-tolerant economic species.CURRENT STATUS OF SPARTINA INVASIONS IN CHINAIn 1985, S. anglica plantation, with human aid, reachedto an area <strong>of</strong> 36,000 ha, and was distributed in 83 coastalcounties in China. The species’ range had increased by 330times from 1966 to 1985. Most <strong>of</strong> <strong>the</strong> plants came from 21English individuals. After that time, however, <strong>the</strong> specieshas declined without human plantings. Only 50 ha <strong>of</strong> <strong>the</strong>plantations still existed in 2000 (Table 4). The species hasbeen also experiencing dieback because it cannot produceTable 4. Area changes and <strong>the</strong>ir causes <strong>of</strong> Spartina in China. (Partialdata from Chung et al. 1985.)Year Spartina anglicaArea Causes(ha)1966 110 Planting and naturalreproduction1973 2,000 Planting and naturalreproduction1978 10,295 Planting and naturalreproduction1980 31,590 Planting and naturalreproduction1985 36,000 Planting and naturalreproductionArea(ha)Spartina alternifloraCauses0.13 Planting260 Planting and naturalreproduction2000

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaany seeds and it has become shorter at all elevations.On <strong>the</strong> o<strong>the</strong>r hand, <strong>the</strong> range <strong>of</strong> S. alterniflora increasedby 10,000 times from 1980 to 1988. Alhough planting <strong>the</strong>species on a large scale was stopped in 1995, it stillincreased by 86 times from 1988 to 2000. The species is stillinvading coastal areas in China by natural dispersal and hasbecome a threat to <strong>the</strong> native ecosystems. From 1995 to2000, ca. 200,000 ha <strong>of</strong> S. alterniflora marshes and tidelandswere reclaimed, which did not stop <strong>the</strong> invasion <strong>of</strong> <strong>the</strong>species.Many native species, including plants, some endangeredbirds (Ma et al. 2004), and mollusks <strong>of</strong> economic value incoastal areas are threatened by S. alterniflora invasions. Itexcluded almost all <strong>the</strong> o<strong>the</strong>r native plants that wereoriginally dominant in wetlands, including Phragmitesaustralis, Typha spp., Scirpus spp., Suadae spp., and eveninvaded fishponds and young mangrove swamps (Qian andMa 1995). For example, in intertidal zones <strong>of</strong> <strong>the</strong> YangziRiver estuary, S. alterniflora invaded Scirpus mariquetercommunities. Through competition with <strong>the</strong> native species,S. alterniflora has greatly decreased native speciesabundance, and even excluded <strong>the</strong>m (Li et al. in thisvolume). Li et al. (this volume) also compared <strong>the</strong> structure<strong>of</strong> nematode communities among Spartina marsh and S.mariqueter and Phragmites marshes <strong>of</strong> Dongtan wetlands onChongming Island, and found significant differences introphic structure <strong>of</strong> nematode communities between <strong>the</strong>marshes.Spartina alterniflora is one <strong>of</strong> 16 notorious invasivepest plants in China as <strong>the</strong> species directly causes millions <strong>of</strong>dollars <strong>of</strong> economic loss per year (An et al, 2007). Spartinaalterniflora is still rapidly spreading, although <strong>the</strong> Chinesegovernment and scientists are doing <strong>the</strong>ir best to control oreradiate <strong>the</strong> species by physical, chemical, biological andintegrated methods (Lin 1997; Liu and Huang 2000).RESEARCH BEFORE 1995Since S. anglica arrived at Nanjing University <strong>of</strong> China,a series <strong>of</strong> studies (all <strong>the</strong> papers published before 1986 in<strong>the</strong> cited literature) had been carried out by <strong>the</strong> SpartinaResearch Group, which focused on S. anglica by 1985. Forexample:• 1963 to 1966: Asexual propagation from rhizome andramets; seed germination, tolerance <strong>of</strong> individuals tocold, hot, drought, water logging, silt sediment and saltstress; growth under different elevation; sexual andasexual reproduction from established individuals;transplanting and management techniques.• 1967 to 1973: Field transplanting; impacts <strong>of</strong> S. anglicagrowth on soil properties and planting techniques ininland riparian wetlands.• 1974 to 1978: Green manure for cropland, forage forsheep, goat and pig; height growth; fertilized flowersTable 5. Survivorship and growth <strong>of</strong> Spartina alterniflora at differentelevation levels. (Revised from Xu and Zhuo 1985.)Elevation(m)Inundationfrequency(days/month)Inundationtime(hours/day)Survivorship(%)Increase rate <strong>of</strong>individuals perclump (times)and seed production; biomass production, net biomassproduction; impacts <strong>of</strong> harvest on growth and reproductionand impacts <strong>of</strong> grazing by cows, sheep and goats.• 1979 to 1985: Silt sediment rate <strong>of</strong> Spartina plantation;uses in protection <strong>of</strong> dykes; prevention <strong>of</strong> tideland erosion;purification <strong>of</strong> polluted water; impact on beneathfauna; Gram negative bacteria composition; anatomy <strong>of</strong>roots and stems; germination <strong>of</strong> seeds produced inChina and impacts <strong>of</strong> puncture and air pressure on seedgermination; physiology; tissue culture; introduction toriparian wetlands in Yellow River <strong>of</strong> NorthwesternChina; biochemical composition and nutrient; impacts<strong>of</strong> fertilizer on growth and biomass production; foragefor fish; structures <strong>of</strong> salt glands; ultra-structures <strong>of</strong> leafcells; DNA contents; ecotype identification and methodsto increase <strong>the</strong> seed production.After 1985, most studies (in <strong>the</strong> cited literature between1986 and 1994) focused on biology and ecology <strong>of</strong> S.alterniflora, including survivorship and growth at differentelevations (Table 5); growth rate and growth form/type fromthree different origins; above- and below-ground biomassand biomass allocation <strong>of</strong> three growth-type populations;seed germination at different saline stresses; effect <strong>of</strong> N-fixed 4088 strain <strong>of</strong> bacteria on seed germination andindividual growth; transportation and distribution <strong>of</strong>phosphorus (P) in organs under salt stress; physiology andbiochemistry <strong>of</strong> leaf and root; effects <strong>of</strong> seed soaking onseed germination; enzyme activity, sugar contents, aminoacid and cold tolerance <strong>of</strong> <strong>the</strong> seedling under different saltstress; translocation <strong>of</strong> mineral elements; contents, allocationand uses <strong>of</strong> Selenium (Se) <strong>of</strong> plants; flavonoids andimmunity activity; isozyme differentiation; micromorphology<strong>of</strong> pollen grains; ultrastructure <strong>of</strong> mesophyllcells; anatomy <strong>of</strong> seeds; biomass dynamics, speciesstructures, energy storage and energy flow <strong>of</strong> ecosystem; <strong>the</strong>relationships between Spartina and clamworms; soil enzymeactivity in Spartina salt marsh; extraction method,toxicology, nutrient contents, functions and uses <strong>of</strong>biomineral beverages/tea.Height(cm)0.0 30 5.0 - 8.0 16.7 0 12.50.5 30 3.5 - 7.0 66.7 1.5 27.61.0 30 1.5 - 6.0 66.7 4.8 37.31.5 30 0.5 - 4.5 100.0 10.8 90.22.0 26 - 30 0.5 - 3.5 100.0 13.7 101.42.5 18 - 28 0.5 - 2.0 100.0 11.8 98.23.0 14 - 22 0.5 - 2.0 100.0 8.0 72.3-68-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadOverall, most <strong>of</strong> <strong>the</strong> studies on Spartina before 1995were involved in ecology after introduction, ecologicalengineering, multi-dimensional uses, physiology,biochemistry and <strong>the</strong> population biology <strong>of</strong> <strong>the</strong> species.RECENT STUDIES AND ADVANCESSince 1995, two research groups were reconstructedwith two pr<strong>of</strong>essors, six associate pr<strong>of</strong>essors, three assistantpr<strong>of</strong>essors and <strong>the</strong>ir students at Nanjing University.Meanwhile, a couple <strong>of</strong> Chinese scientists joined in <strong>the</strong>research from o<strong>the</strong>r institutions. Our studies on S.alterniflora focused on: multidimensional uses, invasiveconsequences and management, competition with differentspecies, wave reduction, ecological services and evaluation,energy flow and community dynamics, evaluation <strong>of</strong>ecological engineering, purification <strong>of</strong> wastewater, sedimentprocess, remote sensing, expending process and outbreak <strong>of</strong>populations, soil property <strong>of</strong> salt marsh, rare waterfowlconservation, cDNA library, flavonoids <strong>of</strong> biomass,decomposition dynamics <strong>of</strong> plants, genetic diversity andgenetic differentiation, greenhouse gas release <strong>of</strong> salt marsh.The work on S. anglica covered salt-induced transcript,microbial composition, mechanism <strong>of</strong> dieback and geneticstructures <strong>of</strong> declining populations. Fur<strong>the</strong>rmore, <strong>the</strong> studieson S. patens included tissue culture <strong>of</strong> seed, ionimplantations <strong>of</strong> plants, effect <strong>of</strong> ion beam on seedgermination and uses <strong>of</strong> <strong>the</strong> species. Most <strong>of</strong> <strong>the</strong> studies areinvolved in structures and functions <strong>of</strong> Spartina saltmarshes; genetic diversity and dieback <strong>of</strong> S. anglica; geneticvariation, outbreak, management and control <strong>of</strong> S.alterniflora, restoration <strong>of</strong> degraded coastal ecosystemsinvaded by S. alterniflora; and effects <strong>of</strong> S. alterniflorainvasion on release <strong>of</strong> carbon-, nitrogen- and sulfer-basedtrace gases. Some <strong>of</strong> <strong>the</strong> important results are summarizedbelow:• Rhizome and ramets <strong>of</strong> S. anglica have strong asexualreproductive capacity in both <strong>the</strong> laboratory and coastalecosystems, but <strong>the</strong> seeds have very low germinationrate. The species grows well in rapid sedimentation conditionsat 2.5 meters (m) to 4.2 m in elevation above sealevel if its leaves are not completely covered by <strong>the</strong>sediments, has broad temperature range from -25°C to42°C and can survive in open coasts with violent impactsfrom typhoons and tides. Spartina anglica cangrow at salinities <strong>of</strong> 0-6% (W/W), with <strong>the</strong> optimal salinity<strong>of</strong> 0.4-1.4%. Best growth occurs at 3.0±0.3 m inelevation with a tolerance range from 2.3 m to 4.0 m. Itdecreased <strong>the</strong> dissoluble ions <strong>of</strong> soil by 2.6% and in reclaimedtidelands <strong>the</strong> species decreased saline concentrationand pH by 38.2% and 2.4%, respectively.• Fertile flowers and seed production <strong>of</strong> S. anglica decreasedwith increasing density. On average, 24.6% <strong>of</strong>flowers can be fertilized and produce seeds in <strong>the</strong> secondyear after planting in <strong>the</strong> field when <strong>the</strong> individualdensity is low. But, in <strong>the</strong> fourth year, <strong>the</strong> crowdedramet populations did not flower. Our observationsshow that <strong>the</strong> species has lost its sexual propagationability. But, <strong>the</strong> trade-<strong>of</strong>f between density and seed productionneeds more experiments both in <strong>the</strong> field andunder controlled conditions.• During <strong>the</strong> first four years <strong>of</strong> invasion, individual density,leaf area index, biomass (both above- and belowgroundbiomass) and net assimilation production <strong>of</strong> S.anglica significantly (p

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaful method to eradicate Spartina, but more experimentsare needed to verify its effectiveness (Chen et al. 2007).• Spartina alterniflora has lower genetic diversity buthigher genetic differentiation capacity whereas S. anglicahas higher genetic diversity and lower differentiationcapacity (An et al. this volume). These genetic differencesmay have contributed to <strong>the</strong> different invasiondynamics, and different fates, for <strong>the</strong> two species <strong>of</strong>Spartina.ACKNOWLEDGMENTSThis study was financially supported by CommonwealFoundation <strong>of</strong> Forestry Administration <strong>of</strong> China (No.200804005) and by KSJYXRC Foundation <strong>of</strong> China’sMinistry <strong>of</strong> Education (No. 0208002002). We thank Dr.Chung-Hsin Chung and Mr. Rongzhong Zhuo for <strong>the</strong>irvaluable suggestions and comments on <strong>the</strong> manuscript.Special thanks for Dr. Bo Li for his careful editing <strong>of</strong> <strong>the</strong>manuscript.REFERENCESAbdul-Kareem, A.R., C. Sheng, M. Shi, et al. 2003. Isolation andcharacterization <strong>of</strong> a salt-induced transcript from halophyteSpartina anglica using DDRT-PCR. J. <strong>of</strong> Nankai Univ. 36(2):8-14.An, S., Y. Xiao, H. Qing, Z. Wang, C. Zhou, B. Li, S. Shi, D. Yu, Z.Deng, and L. Chen. 2010. Varying success <strong>of</strong> Spartina spp. invasionsin China: Genetic diversity or differentiation? In: Ayres,D.R., D.W. Kerr, S.D. Ericson and P.R. Ol<strong>of</strong>son, eds. <strong>Proceedings</strong><strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartina, 2004Nov 8-10, San Francisco, CA, USA. San Francisco Estuary InvasiveSpartina Project <strong>of</strong> <strong>the</strong> California State Coastal Conservancy:Oakland, CA (this volume).An, S.Q. and C.H. Chung. 1991. Spartina ecological engineering andbenefit analysis <strong>of</strong> disaster prevention. J. <strong>of</strong> Nanjing Univ., disasterissue, 367-373.An, S.Q., B.H. Gu, C.F. Zhou, Z.S. Wang, Z.F. Deng, Y.B. Zhi, H.L.Li, L. Chen, D.H. Yu, and Y.H. Liu. 2007. Spartina invasion inChina: implications for invasive species management and futureresearch. Weed Research 47: 183-191.Baumel, A., M.L. Ainouche, and J.E. Levasseur. 2001. Molecularinvestigations in populations <strong>of</strong> Spartina anglica C.E. Hubbard(Poaceae) invading coastal Brittany (France). Mol. Ecol. 10: 1689-1701.Chen, C.J. 1994. Effect <strong>of</strong> increasing deposition and defending seashoreby planting Spartina anglica on beach in Jiangsu Province.Bull. Mar. Sci., 13(2): 55-61.Chen, J.Q., and C.H. Chung. 1991. Differentiation <strong>of</strong> 3 ecotypes inresponses to salt stress in Spartina alterniflora. Acta Phytoecol. etGeobot. Sinica, 14(1): 33-39.Chen, L., Z.F. Deng, S.Q. An, C.J. Zhao, C.F. Zhou, and Y.B. Zhi.2007. Alternative irrigation restrains clonal growth and reproduction<strong>of</strong> Spartina alterniflora. Journal <strong>of</strong> Plant Ecology 31(4): 645-651.Chung, C.H. 1982. Low marshes, China. In: Lewis, R.R., ed. Creationand Restoration <strong>of</strong> Coastal Plant Communities. Boca Raton:CRC Press. pp. 131-145.Chung, C.H. 1983. Geographical distribution <strong>of</strong> Spartina anglicaHubbard in China. Bull. Mar. Sci. 33(3):753-758.Chung, C.H. 1985. The effects <strong>of</strong> introduced Spartina grass oncoastal morphology in China. Z. Geomorph. N.F. Suppl. Bd.57:169-174.Chung, C.H. 1985. Twenty-two Year Research Advances inSpartina. J. <strong>of</strong> Nanjing Univ., special issue. pp.358.Chung, C.H. 1989. Ecological engineering <strong>of</strong> coastlines with saltmarshplantations. In: Mitsch, W.J. and S.E. Jørgensen, eds. EcologicalEngineering. New York: John Wiley & Sons 255-289.Chung, C.H. 1990. Twenty-five years <strong>of</strong> introduced Spartina anglicain China. In: Gray, A.J. and P.E. Benham, eds. Spartina anglica —a Research Review. London: HMSO. pp. 72-76.Chung, C.H. 1993. Thirty years <strong>of</strong> ecological engineering withSpartina plantations in China. Ecol. Eng, 2:261-289.Chung, C.H., and R.Z. Zhuo. 1979. Spartina anglica. Grassland <strong>of</strong>China 1:45-48.Daehler, C.C. and D.R. Strong. 1996. Status, prediction and prevention<strong>of</strong> introduced cordgrass Spartina spp. invasions in Pacific estuaries,USA. Biol. Conserv., 78:51-58.Guan, Y.J. and S.Q. An. 2003. ecological benefits and biological invasions<strong>of</strong> Spartina in Chna. J. Nanjing For. Univ., special issue,95-101.Hacker, S.D., D. Heimer, C.E. Hellquist, T.G. Reeder, B. Reeves,T.J. Riordan, and M.N. Dethier. 2001. A marine plant (Spartinaanglica) invades widely varying habitats: potential mechanisms <strong>of</strong>invasion and control. Biol. Invasions, 3:211-217.Hu, Z.H. 1998. Anti-inflammatory activity <strong>of</strong> topically applied totalflavonoids <strong>of</strong> Spartina alterniflora Loisel in vivo. J. <strong>of</strong> Plant Res.and Environ. 7(2):6-11.Hu, Z.H. 1998. Effect <strong>of</strong> total flavonoids <strong>of</strong> Spartina alterniflora onserum lipids in vivo. Mar. Sci., 2:16-18.Jiang, F.X. and W.Z. Wang. 1992. The relationships betweenSpartina anglica and clamworms. Acta Ecol. Sinica 12(1):84-88.Jiang, F.X. and W.Z. Wang. 1987. Decomposition and nutrientchanges <strong>of</strong> dead Spartina anglica and S. alterniflora. ActaOceanologica Sinica 9(3):367-372.Jiang, H.X. and J.S. Huang. 1982. The preliminary observation on<strong>the</strong> digallic acid-treated transmission electron microscopic specimens<strong>of</strong> Spartina anglica C.E. Hubbard. J. Cell Biol., 4(2):15-16.Kriwoken, L.K. and P. Hedge. 2000. Exotic species and estuaries:managing Spartina anglica in Tasmania, Australia. Ocean &Coastal Manag., 43:573-584.Li, A.M. and X.C. Liu. 1994. Microbial composition <strong>of</strong> Spartina anglicasalt marshes. Tianjing Agri. Sci. 1: 27-28.Lin, Q.R. 1997. Damage and management <strong>of</strong> Spartina anglica and S.alterniflora. Fujian Geography, 1:16-20.Liu, J. and J.H. Huang. 2000. Management <strong>of</strong> Spartina alterniflora.Bull. Oceanogr., 10: 68-72.Liu, Y.X. 2001. Analysis <strong>of</strong> remote sensing images for vegetationsuccession on tidal salt marsh in Jiangsu. Rural Eco-Environ.17(3):39-41.Lü, Z.X., and C.H. Chung. 1981. A comparative study <strong>of</strong> seed germinationand free amino acid composition <strong>of</strong> Spartina anglica infresh and sea water conditions. J. Plant Physiol. 7:281-286.Lü, Z.X., and C.H. Chung. 1983. The influence <strong>of</strong> NaCl on freeamino acid composition and proline content <strong>of</strong> Spartina anglicaseedling. J. Plant Physiol., 8(4):393-398.Ma, Y.J. 2001. Studies on component <strong>of</strong> Spartina alterniflora Loisel.II: The research <strong>of</strong> volatile component with GC-MS. Chinese J. <strong>of</strong>Bioch. Pharma. 23(1):36-37.Ma, Z.J., B. Li, K. Jing, S.M. Tang, and J.K. Chen. 2004. Are artificialwetlands good alternatives to natural wetlands for waterbirds?-70-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadA case study on Chongming Island, China. Biodiversity and Conservation13: 333-350.Ou, H.C. 1982. A preliminary study <strong>of</strong> alkalinity and salinity toleranceby seedlings <strong>of</strong> Spartina anglica Hubbard. Plant Physiol.Bull. 1:15-17.Qian, Y.Q. and K.P. Ma. 1995. Bio-techniques and bio-safety. J. <strong>of</strong>Nat. Res. 10: 322-334.Qin, P., M.D. Jin, and M. Xie. 1985. The comparison <strong>of</strong> communitybiomass among <strong>the</strong> three ecotypes <strong>of</strong> Spartina alteniflora in Louyuanbay, Fujian. J. Nanjing Univ. Res. Adv. Spartina 226-236.Qin, P., and M. Xie. 1988. Distribution <strong>of</strong> N, P and K in Spartinaalterniflora salt marshes <strong>of</strong> Luoyuan Bay. Mar. Sci. 4: 63-67.Qin, P., and M. Xie. 1991. Primary production and formation <strong>of</strong> flavanoid<strong>of</strong> Spartina alterniflora. Acta Ecol. Sinica 11(4): 293-298.Qin, P., and C.H. Chung. 1992. Applied Studies on Spartina. Beijing:Marine Press.Qin, P., M. Xie, and Y.S. Jiang. 1998. Spartina green food ecologicalengineering. Ecol. Eng. 11:147-156.Qin, P., M. Xie, Y.S. Jiang, and C.H. Chung. 1997. Estimation <strong>of</strong> <strong>the</strong>ecological-economic benefits <strong>of</strong> two Spartina alterniflora plantationsin North Jiangsu, China. Ecol. Eng. 8:5-17.Shen, Y.M. 2002. Analysis <strong>of</strong> <strong>the</strong> expending process <strong>of</strong> <strong>the</strong> Spartinaalterniflora Loisel salt marsh on Jiangsu Province coast by remotesensing. J. Plant Res. Environ. 11(2):33-38.Sun, B.Y., and C.S. Zhu. 1989. Distribution and enzyme activity <strong>of</strong>soil microbes in Spartina alterniflora salt marshes. Acta Ecol. Sinica9:240-241.Wan, S.W., P. Qin, Y. Li, and X.P. Liu. 2001. Wetland creation forrare waterfowl conservation: A project designed according to <strong>the</strong>principles <strong>of</strong> ecological succession. Ecol. Eng., 18:115–120.Wang, C.Y. 1994. Seasonal changes <strong>of</strong> photosyn<strong>the</strong>sis rates and itseffects on primary production for Spartina. Rural Eco-Environ.10(3):14-17.Xu, D.Z. 2002. Experimental research on <strong>the</strong> purification <strong>of</strong> Spartinaanglica to household waste water. J. Guizhou Univ. 21(2):121-125.Xu, G.W., and R.Z. Zhuo. 1985. Preliminary studies <strong>of</strong> introducedSpartina alterniflora Loisel in China. J. <strong>of</strong> Nanjing Univ. (Speciesissue for research advances in Spartina—Achievements <strong>of</strong> past 22years) 212-225.Yu, W.H. and Y.Q. Cao. 1981. Preliminary studies on <strong>the</strong>rhizosphere bacteria <strong>of</strong> Spartina anglica. J. <strong>of</strong> Nanjing Univ.3:365-370.Zhao, Q.L., and Q. Zhao. 1997. Effects <strong>of</strong> Spartina anglica to <strong>the</strong> soilhabitats <strong>of</strong> Perinereis aibuhitensis. Chinese J. Ecol. 16(2):28-30.Zhang, P.Y. 1998. Construction <strong>of</strong> cDNA library from commoncordgrass under salt stress. Acta Prataculturae Sinica 7(4):67-73.Zhang, R.S. 2004. Formation <strong>of</strong> Spartina alterniflora salt marshes on<strong>the</strong> coast <strong>of</strong> Jiangsu Province, China. Ecol. Eng. 23: 95–105.Zhang, S.T., P. Qin, and S.W. Wan. 2000. Analysis on <strong>the</strong> resourceallocation structure <strong>of</strong> Spartina alterniflora ecological engineeringfrom <strong>the</strong> emergy benefits aspect. Acta Ecol. Sinica 20(6):1045-1049.Zhou, C.F. 2002. Effect <strong>of</strong> ions implantation on several physiologicaland biochemical indexes <strong>of</strong> Spartina patens. Plant Physiol. Bull.38(3):237-239.Zhou, C.F. 2003. Free radical responses in embryonic calli <strong>of</strong>Spartina patens under N+ beam irradiation. J. <strong>of</strong> Nanjing Univ.39(5):520-529.Zhou, C.F., P. Qin, and M. Xie. 2003. Vegetating coastal areas <strong>of</strong>east China: species selection, seedling cloning and transplantation.Ecol. Eng. 20: 275-286.Zhou, H.B., and H.X. Jiang. 1982. Morphology <strong>of</strong> <strong>the</strong> salt gland <strong>of</strong>Spartina anglica Hubbard. Acta Bot. Sinica 24(2):115-118.Zhou, Lu. 2003. Effect <strong>of</strong> brassinolide on callus growth and regenerationin Spartina patens (Poaceae). Plant Cell, Tissue and OrganCulture 73:87-89.Zhuang, S.H., and C.H. Chung. 1987. An investigation on ecotypedifferentiation <strong>of</strong> Spartina anglica Hubbard. Chinese J. Ecol.6(6):1-9.-71-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadSPREAD OF INVASIVE SPARTINA IN THE SAN FRANCISCO ESTUARYK. ZAREMBA 1,4 ,M.MCGOWAN 2 , AND D.R. AYRES 31 San Francisco Estuary Invasive Spartina Project, 2612-A 8 th St., Berkeley, CA 947102 Maristics, 1442A Walnut St., Suite 188, Berkeley, CA 94709, maristics@comcast.net3 U.C. Davis, Evolution and Ecology, Davis, CA 95616, drayres@ucdavis.edu4 Current address: 971 Village Dr. Bowen Island, BC, V0N 1G0 Canada; katyzaremba@yahoo.caWe mapped <strong>the</strong> location and extent <strong>of</strong> all non-native Spartina in <strong>the</strong> San Francisco estuary in 2000and 2001 and mapped a sample <strong>of</strong> 28 sites in 2003. We incorporated aerial photographs, groundsurveys, and genetic analyses into a GIS. These sites demonstrated an average increase <strong>of</strong> 317percent coverage <strong>of</strong> S. alterniflora x foliosa hybrids, radiating from points <strong>of</strong> <strong>the</strong> deliberateintroduction <strong>of</strong> S. alterniflora. Extrapolating to <strong>the</strong> entire estuary, this suggests a potential increasefrom ca 190 hectares(ha) (470 acres[ac]) recorded in 2001 to as much as 793 ha (1,960 ac) in 2003.Hybrids now occupy approximately four percent <strong>of</strong> <strong>the</strong> total area <strong>of</strong> marsh and mudflats in <strong>the</strong> bay.Spread was greater in tidal marshes or formerly diked baylands and mudflats than in creeks, sloughs,and urbanized marsh (riprap, boat ramps). Genetic testing found no new invasion sites. Manualcontrol methods applied in 2002-2003—digging or covering with geo-textile fabric—were effectiveat removing or killing small populations or single plants <strong>of</strong> Spartina species.Keywords: invasive Spartina, S. alterniflora, S. densiflora, S. patens, S. alterniflora x foliosahybrids, monitoringINTRODUCTIONThe San Francisco Bay Estuary contains <strong>the</strong> largest andmost ecologically important expanses <strong>of</strong> tidal mudflats andsalt marshes in <strong>the</strong> contiguous western United States with adiverse array <strong>of</strong> native plants and animals. Over <strong>the</strong> years,many non-native species <strong>of</strong> plants and animals have beenintroduced to <strong>the</strong> Estuary threatening to change <strong>the</strong> structure,function, and value <strong>of</strong> <strong>the</strong> Estuary’s tidal lands. In recentdecades four species <strong>of</strong> non-native Spartina have begun tospread rapidly in <strong>the</strong> Estuary. Though valuable in <strong>the</strong>irnative settings, <strong>the</strong>se introduced Spartina species are highlyaggressive in this new environment and frequently become<strong>the</strong> dominant plant in areas <strong>the</strong>y invade. (Callaway andJosselyn 1992; Cohen and Carleton 1995; Daehler andStrong 1996; Goals Project 1999; Ayres et al. 2003;California Coastal Conservancy 2003; Ayres et al. 2004).In 2000 <strong>the</strong> California Coastal Conservancy established<strong>the</strong> San Francisco Estuary Invasive Spartina Project (ISP) toprovide a regionally coordinated approach to controlling oreradicating non-native Spartina in San Francisco Bay. TheISP includes a monitoring program to map non-nativeSpartina and to assess <strong>the</strong> effectiveness <strong>of</strong> treatmentmethods. In 2000-2001 <strong>the</strong> ISP mapped <strong>the</strong> entire Estuaryusing <strong>the</strong> methods outlined in Collins et al. (2001). In 2003<strong>the</strong> ISP Monitoring Program mapped a subset <strong>of</strong> 28 sites,monitored sites treated in 2002 and 2003, used genetictesting to confirm identifications at known and suspectedinvasion sites, and compared methods for monitoringcordgrass in <strong>the</strong> Estuary.Five species <strong>of</strong> Spartina are currently found in <strong>the</strong> SanFrancisco Bay Estuary including <strong>the</strong> native, S. foliosa. Thefour non-native species currently found in <strong>the</strong> estuary are S.alterniflora, S. densiflora, S. anglica, and S. patens. Hybridsbetween Atlantic smooth cordgrass S. alterniflora, and <strong>the</strong>native Pacific cordgrass S. foliosa (hereafter termed“hybrids”) now threaten <strong>the</strong> ecological balance <strong>of</strong> <strong>the</strong>Estuary and are likely to cause <strong>the</strong> extinction <strong>of</strong> nativePacific cordgrass, choke tidal creeks, dominate newlyrestored tidal marshes, and displace thousands <strong>of</strong> acres <strong>of</strong>existing shorebird habitat (Ayres and Strong, this vol.;Stralberg et al. this vol.; Ayres et al. 2003; Ayres et al.2008). Invasive cordgrasses from <strong>the</strong> San Francisco Estuarycould spread to o<strong>the</strong>r California estuaries through seeddispersal on <strong>the</strong> tides.The 2000-2001 survey found 195 net hectares (ha) (483acres (ac)) <strong>of</strong> non-native Spartina distributed throughoutnearly 16,187 ha (40,000 ac) <strong>of</strong> tidal marsh and 11,736 ha(29,000 ac) <strong>of</strong> tidal flats (Ayres et al. 2004). Net area is <strong>the</strong>coverage if all non-native Spartina plants were contiguouswhile gross area would be all <strong>the</strong> marsh areas that have somenon-native Spartina plants. Of this total, 190 ha (470 ac)were hybrids, 5 ha (13 ac) were S. densiflora, 0.23 ha (0.58ac) were S. patens, and 0.04 ha (0.09 ac) were S. anglica.The hybrids have increased in area 100-fold since <strong>the</strong> 1970s,from just over one ha <strong>of</strong> planted S. alterniflora in 1978(Ayres et al. 2004). It is hypop<strong>the</strong>sized that <strong>the</strong> proliferation<strong>of</strong> hybrids is accelerating <strong>the</strong> rate at which areas are covereddue to <strong>the</strong> evolution <strong>of</strong> greater invasiveness (Ayres et al.2004; Ayres and Strong this vol.; Hall et al. 2006 and this-73-

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinavol.). In addition to <strong>the</strong> physical displacement <strong>of</strong> nativemarsh plants, <strong>the</strong> hybrid invasion threatens <strong>the</strong> geneticintegrity and continued existence <strong>of</strong> <strong>the</strong> native Spartinafoliosa (Ayres et al. 2003).METHODSOverall DescriptionA total <strong>of</strong> 31 estimates <strong>of</strong> area covered by three species<strong>of</strong> non-native Spartina were made at 28 separate samplingsites in 2003 (Table 1). The three species were S. patens, S.densiflora and hybrids. A sample <strong>of</strong> sites was selected toprovide coverage for <strong>the</strong> entire bay shoreline stratified by“region” as defined by <strong>the</strong> Wetlands Goals Project (GoalsProject 1999) (latitude) and “site type.” Field sites wereselected across <strong>the</strong> latitudinal extent <strong>of</strong> <strong>the</strong> non-nativeSpartina spp. invasion in <strong>the</strong> Estuary (Fig. 1). The regionswere North Bay (NB) (2 sites), Central Bay (CB) (21 sites),and South Bay (SB) (5 sites). The site types were I) tidal,micro-tidal, and formerly diked bayland, and back barriermarsh (9 sites); (II) fringing tidal marsh, mud flats andestuarine beaches (7 sites); (III) major tidal sloughs, creeksor flood control channels (5 sites); and (IV) urbanized rock,riprap, docks, boat ramps and marinas (7 sites). At least twomarshes <strong>of</strong> each site type were selected from each <strong>of</strong> <strong>the</strong>three regions <strong>of</strong> <strong>the</strong> Bay. Within each site type six to sevenmarshes were selected.The three non-native species differed in <strong>the</strong>irdistributions among <strong>the</strong> three bay regions. Spartina. patensoccurred at one location, Southampton, in <strong>the</strong> North Baywhere no o<strong>the</strong>r non-native species occurred. Spartinadensiflora occurred at seven locations: one in <strong>the</strong> North Bayand six in <strong>the</strong> Central Bay. At three <strong>of</strong> <strong>the</strong> Central Bay sitesS. densiflora and hybrids both occurred. Two <strong>of</strong> <strong>the</strong>se threesites are adjacent to each o<strong>the</strong>r, that is, Blackie’s Creek runsthrough Blackie’s Pasture. Hybrids occurred at 23 separatesites from <strong>the</strong> Central Bay and <strong>the</strong> South Bay. It was notpossible to sample equal numbers <strong>of</strong> sites within each typeand region due to <strong>the</strong> unequal frequencies <strong>of</strong> appropriatesites, <strong>the</strong> requirement to avoid clapper rails in somelocations and to sample particular sites at particular tideheights, and limited numbers <strong>of</strong> trained field staff andGlobal Positioning System (GPS) units. Never<strong>the</strong>less, asnoted below, <strong>the</strong> sampling in 2003 encompassed <strong>the</strong> fullknown regional extent <strong>of</strong> non-native Spartina speciesdistribution.MappingIn 2003 distribution and abundance <strong>of</strong> non-nativeSpartina were mapped at <strong>the</strong> 28 sample sites and comparedto <strong>the</strong> distribution and abundance at those locations in 2001.Overall methods followed Collins et al. (2001) with somemodifications.At each sampling site observers mapped <strong>the</strong> locationand areal extent <strong>of</strong> non-native Spartina using a GPS dataTable 1. Change in Area Covered Between 2001 and 2003 Using FieldMeasurements and Field Estimates*at 28 sites for three Spartina taxa.SitesTaxaArea in Area in2001 (ac) 2003 (ac)Area in2001(squaremeters)Area in2003(squaremeters)Changein Area2001 to2003Site Type IBunker Marsh* Hybrids 0.39 1.40 1580.11 5665.59 259%Citation Marsh* Hybrids 0.51 1.93 2059.74 7803.41 279%Cogswell Marsh(North Quadrant)*Hybrids 0.37 1.32 1503.16 5334.08 255%Piper Park West S. densiflora 0.02 0.08 99.92 318.40 219%SouthamptonMarshS. patens 0.31 0.05 1244.49 196.99 -84%Point Pinole S. densiflora 0.00 0.00 2.99 6.39 114%Palo Alto Baylands Hybrids 0.12 0.15 478.26 618.49 29%Corte MaderaMarsh Reserve 1Corte MaderaMarsh Reserve 2S. densiflora 0.011 0.014 44.95 60.75 35%Hybrids 0.01 0.05 26.46 214.92 712%Pickleweed Park S. densiflora 0.49 0.03 1964.07 118.73 -94%Site Type IIEmeryville West Hybrids 0.07 1.59 282.60 6430.15 2175%Coyote CreekMarshAlameda Island –N. Elsie Roemer*Hybrids 0.09 0.06 353.25 233.15 -34%Hybrids 0.29 0.98 1171.89 3948.94 237%Blackie’s Pasture 1 S. densiflora 0.021 0.020 84.12 82.66 -2%Blackie’s Pasture 2 Hybrids 0.01 0.08 41.37 330.90 700%Ideal Marsh* Hybrids 0.26 0.88 1070.72 3541.00 231%Richmond InnerHarbor – SteegeMarshHybrids 0.01 0.02 32.19 100.46 212%Bayshore Park Hybrids 0.05 0.25 168.19 1018.84 506%Site Type IIIBlackie’s Creek 1 Hybrids 0.03 0.02 108.33 93.97 -13%Blackie’s Creek 2 S. densiflora 0.00 0.00 12.06 8.46 -30%Colma Creek Hybrids 2.36 7.16 9551.11 28957.46 203%Corte MaderaCreekS. densiflora 1.19 2.70 4802.16 10948.30 128%San Leandro Creek Hybrids 2.16 3.83 8760.60 15481.79 77%San Mateo Creek Hybrids 0.20 0.78 825.41 3172.16 284%Site Type IVCoyote Pt Marina Hybrids 0.35 0.74 1408.81 3004.87 113%Oakland InnerHarborHybrids 2.74 5.40 11087.72 21685.73 96%Yosemite Slough Hybrids 0.02 0.12 75.47 476.02 531%India Basin Hybrids 0.10 0.01 408.20 41.47 -90%Pier 94 Hybrids 0.04 0.03 171.72 129.50 -25%Pier 98/Heron’sHead (treat tookplace in 2002)Loch LomondMarinaHybrids 0.002 0.009 8.83 37.40 324%Hybrids 0.005 0.016 19.63 64.98 231%Site and Marsh Types:Type I. Former Diked Bayland/Microtidal/Tidal/Back Barrier MarshType II. Fringing Tidal Marsh/Mudflats/Estuarine BeachesType III. Major Tidal Slough, Creek or Flood Control ChannelType IV. Urbanized rock, riprap, dock, ramp, marina.-74-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and Spreadentry unit (Trimble Geo-Explorer III).Plant data were entered into <strong>the</strong> unit aspoints, lines or polygons depending on<strong>the</strong> extent <strong>of</strong> <strong>the</strong> invasion. The GPSunits automatically collect data ondate, time, location, area and polygonperimeter and line length. Field staffmanually entered site name, speciesidentification, clone identification,clone diameter, line width, samplename, percent cover class, o<strong>the</strong>rapplicable data and comments.GPS data were downloaded usingPathfinder s<strong>of</strong>tware, exported toArcView, differentially corrected,and reviewed for errors. The data were<strong>the</strong>n exported to Excel and onceagain reviewed for data entry errors.Final data were summarized andmapped in ArcView and statisticallyanalyzed using SYSTAT.Three <strong>of</strong> <strong>the</strong> four non-native taxa<strong>of</strong> Spartina found in <strong>the</strong> San FranciscoEstuary—S. Alterniflora hybrids, S.densiflora, and S. patens—weremapped in 2003. The one locationwhere S. anglica was found in 2001was not included in <strong>the</strong> 2003monitoring program. Spartina specieswere identified in <strong>the</strong> field using plantmorphology (Zaremba 2001). Leafsamples were collected for geneticanalysis to confirm field identifications<strong>of</strong> S. alterniflora and hybrids.Aerial Photo InterpretationIn addition to field-collected data,aerial photo interpretation was used atfive <strong>of</strong> <strong>the</strong> 28 sites for large Spartinahybrid infestations, not single clones.Color infrared photos were taken at1:6000 feet scale at low tide during <strong>the</strong> peak <strong>of</strong> <strong>the</strong> growingseason between August-October when <strong>the</strong> plants were stillgreen to allow for accurate yearly comparison. Photos werescanned at 1200 dpi and orthorectified <strong>the</strong>n imported intoArcView 3.3 for review and analysis. Polygons weredigitized around <strong>the</strong> Spartina meadows and polygons weregiven a cover class (

##Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaGenetic AnalysesGenetic analyses to confirm speciesidentification were done at each <strong>of</strong> <strong>the</strong> 28sites and at additional sites requested byconcerned landowners, managers orstakeholders (Fig. 2).Additional genetic surveys weredone at a few sites along <strong>the</strong> coast north<strong>of</strong> <strong>the</strong> Golden Gate in Point ReyesNational Seashore and Bolinas Lagoon.In order to confirm species identification,leaves from ambiguous plants and atleast 3-5 plants per monitoring site werecollected according to standard methodsused previously (Ayres et al. 2004) andconforming to recommendations inCollins et al. (2001). Where needed,transects were run <strong>the</strong> length <strong>of</strong> marshessampling every 10 meters to determine(1) if <strong>the</strong>re were any “hidden” S.alterniflora hybrids, or (2) <strong>the</strong> percentinvasion <strong>of</strong> a particular marsh. Fromeach individual clone <strong>of</strong> unknownspecies identification and from each 10-meter sample along <strong>the</strong> transects, asingle plant leaf was collected for geneticanalysis using RAPD (RandomAmplified Polymorphic DNA) nuclearmarkers (Daehler and Strong 1997;Ayres et al. 1999).#Coyote Creek,Mill Valley -100 % nativeS. foliosaEfficacy MonitoringISP staff monitored <strong>the</strong> efficacy <strong>of</strong> 13different treatments at eight sites. Sometreatments were done in 2002 withassessment <strong>of</strong> efficacy in 2003.Treatments done in 2003 will be assessedin 2004. Treatment monitoring data werecollected consistent with methodsdescribed in Collins et al. (2001). Prior totreatment <strong>the</strong> extent <strong>of</strong> non-native Spartina was mapped todetermine <strong>the</strong> total area <strong>of</strong> non-native Spartina. For largetreatment sites mapped as polygons with a percent cover <strong>of</strong>non-native Spartina <strong>the</strong> sampling scheme described byCollins et al. (2001) was modified from a graphicalinformation system (GIS) based sampling approach to afield-based approach. A stratified random sub-samplingmethod was applied. Transects were run across <strong>the</strong> length <strong>of</strong><strong>the</strong> treatment area parallel to <strong>the</strong> shoreline at selectedelevations and <strong>the</strong>n random points were sampled along <strong>the</strong>transects. For smaller treatment sites where individual cloneswere mapped as points all <strong>the</strong> mapped plants up to aSan Francisco BayNon-native Spartina Genetic Survey Map 2003Beach Drive,San Rafael -2S.alternifloraclones found (12%)Loch LomondMarina -3S.alternifloraclonesPetaluma Marsh -Wiloki Creek,Donahue Sloughand San Antonio Creek -100% native S. foliosaCrissy Field -3S.alternifloraseedlings foundand pulled fall 2003#########Yosemite Slough -89% S. alt-hyb point samples84% S. alt-hyb transect samplesLegendSpartina genetic survey points 2003# S. alternifloralab tested# S.alterniflora-hybrid lab tested# S. foliosa lab tested# no resultsSpartina genetic survey lines 2003S. alt-hybS. foliosano results6 0 6 12 MilesFig. 2 - Baywide Map <strong>of</strong> Genetics Sampling SitesW#NS##Meeker Slough - 100% native S. foliosaEEast and West Stege Marsh -100% native S. foliosaOutbound Stege Marsh -2S.alterniflora-hybrid clonesfound and treated fall 2003maximum <strong>of</strong> 30 plants per site were monitored. If more than30 plants were mapped, 30 were randomly selected tomonitor pre-treatment and post-treatment to determine <strong>the</strong>treatment efficacy. Measurements <strong>of</strong> overall plant vigorincluding plant height, density per 0.25 meter, plant vigor(high/medium/low), tide wash (yes/no), plant species percentcover (native and non-native plant species) per 0.25 meter,percent flower Spartina per 0.25 meter, sediment type,high/medium/low marsh, and burn (yes/no) were collectedfrom each sampling location. The entire treatment area wasmapped again one year post-treatment and <strong>the</strong> sametreatment efficacy data were collected.##Plummer Creek Mitigation Site-100% native S. foliosa## ## ## # # # ##Cooley Landing/RavenswoodPreserve - 19% S. alterniflora-hybrids,seedlings 100% native S. foliosa-76-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadTable 2. Change in area covered by non-native Spartina species between2001 and 2003 averaged over all sites surveyed, comparing aerial photointerpretation measurements (APIM) with field measurements and fieldestimates.SpeciesS. AlternifloraHybridsAverage Site Change in Area from 2001-2003Field Measurement +Field Estimates(N=28)**APIM(N=5)*317.00% 213.05%S. densiflora 52.83% n/aS. patens -84.2% n/aAll Species 244.41% 213.1%* Five <strong>of</strong> <strong>the</strong> 28 sites had only aerial photo interpretation measurementsfor 2001. For <strong>the</strong>se five sites, a 2001 field estimate was calculated using aregression curve based on a correlation between <strong>the</strong> existing 2001 and2003 field measurements. This 2001 field estimate was used in <strong>the</strong> place<strong>of</strong> <strong>the</strong> 2001 field survey data to calculate <strong>the</strong> change in area from 2001-2003.** For five sites <strong>of</strong> <strong>the</strong> 28 sites, 2001 aerial photo interpretation measurementwas used in <strong>the</strong> place <strong>of</strong> <strong>the</strong> 2001 field survey data to calculate<strong>the</strong> change in area from 2001-2003.Data AnalysesData were subjected to a variety <strong>of</strong> quality assurancemethods (Zaremba et al. 2004). Cross-tabulation <strong>of</strong>categorical names was done to check for typographicalerrors and duplications. Summary statistics were <strong>the</strong>ncalculated for quantitative variables to check forunreasonable ranges and outliers. Descriptive statistics werecalculated and raw data transformed as needed to meetrequirements <strong>of</strong> parametric statistical tests.We tested <strong>the</strong> hypo<strong>the</strong>sis that <strong>the</strong>re was an increase inarea covered by non-native Spartina using a t-test betweenarea covered in 2001 and 2003. The hypo<strong>the</strong>sis that aparticular species, site type or bay region (latitude) had aninfluence on <strong>the</strong> change in area covered between 2001 and2003 was tested with an Analysis <strong>of</strong> Variance (ANOVA).We tested <strong>the</strong> hypo<strong>the</strong>sis that site latitude had a significanteffect on change in area covered between 2001 and 2003 byusing an Analysis <strong>of</strong> Covariance (ANCOVA). The accuracy<strong>of</strong> field identification <strong>of</strong> S. alterniflora, S. foliosa and <strong>the</strong>irhybrids was tested statistically with a Chi-square test <strong>of</strong>frequencies <strong>of</strong> correct field identification <strong>of</strong> S. alterniflora,S. alterniflora hybrids, and S. foliosa. This test used <strong>the</strong>genetic analysis as <strong>the</strong> true (or <strong>the</strong>oretically expected)frequencies and <strong>the</strong> field observations as <strong>the</strong> observedfrequencies.Based on data averaged over all species and all sitessampled, <strong>the</strong> non-native species and hybrids <strong>of</strong> Spartinaincreased 244% in area covered (paired sample t-test, p =0.003) between 2001 and 2003 (Table 2). Field surveys(field measurement plus estimates) showed an increase <strong>of</strong>non-native Spartina <strong>of</strong> 15% in <strong>the</strong> North Bay, 292% in <strong>the</strong>Central Bay and 177% in <strong>the</strong> South Bay from 2001 to 2003.These measurements showed that non-native Spartina (allspecies) increased from 2001-2003 by 172% at Type I sites(tidal, micro-tidal, and formerly diked bayland, and backbarrier marsh), by 504% at Type II sites (fringing tidalmarsh, mud flats and estuarine beaches), by 108% at TypeIII sites (major tidal sloughs, creeks or flood controlchannels), and by 169% at Type IV sites (urbanized rock,riprap, dock, ramp, marina).S. patens occurred at only one site in <strong>the</strong> North Bay, aSite Type I. It decreased in area covered by 84%.S. densiflora decreased in nearly as many sites (3) as itincreased (4). Statistical tests were precluded for S.densiflora because <strong>of</strong> low sample sizes but <strong>the</strong> averagepercent change in area covered was 53% with a range from-94% to +219%. Spartina densiflora was mapped at onlyone sample location in <strong>the</strong> North Bay (Pt. Pinole), where itincreased by 114%. Spartina densiflora increased by 43% in<strong>the</strong> Central Bay. Spartina densiflora increased in Type I andType III sites but apparently decreased slightly at Type IIsites. No S. densiflora were noted at Type IV sites.Hybrids make up most <strong>of</strong> <strong>the</strong> non-native coverage and<strong>the</strong>ir proportion among <strong>the</strong> non-natives increased in 2003compared to 2001 accounting for 83% <strong>of</strong> <strong>the</strong> non-nativecoverage in 2001 and increasing to 90% <strong>of</strong> non-nativecordgrasses in 2003. Hybrids increased in cover three-foldbetween 2001 and 2003 (paired sample t-test, p < 0.001)with a range from -90% to +2175% depending on site.Hybrids increased in area covered in 19 <strong>of</strong> 23 marshessampled, a statistically significant proportion <strong>of</strong> <strong>the</strong> sites(non-parametric sign test, p = 0.003).Hybrids in <strong>the</strong> South Bay were responsible for <strong>the</strong> 177%increase in non-native invasive Spartina between 2001-2003. Hybrids increased by 403% in <strong>the</strong> Central Bay.Hybrids were not found in any sample sites <strong>of</strong> <strong>the</strong> NorthBay. Hybrids had higher mean and total area covered in <strong>the</strong>Central Bay than in <strong>the</strong> South Bay in both 2001 and 2003.The slight trend <strong>of</strong> percent change in hybrid area coveredwith latitude (south-to-north trend) was not statisticallysignificant (p = 0.18).While hybrid acreage increased greatly from 2001 to2003 across all site types, site types I and II had higherpercent increases, on average, than site types III and IV(non-parametric sign test; p = 0.254). Analysis <strong>of</strong> <strong>the</strong> areacovered by hybrids (log-transformed square meters) in 2003found no statistical difference among site types (ANOVA,p = 0.377).Aerial Photos versus Field MeasurementsThe estimates <strong>of</strong> <strong>the</strong> area covered by non-nativeSpartina hybrids differed between <strong>the</strong> aerial photo-77-

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaTable 3. Change in Area Covered by Non-native Spartina between 2001 and 2003 for Five Sites Comparing Aerial Photo Interpretation Measurements(APIM), Field Estimate and Field Measurement Measurements.Site NameBunker MarshSiteTypeISpeciesArea in 2001(square meters)APIMFieldEstimate*Area in 2003(square meters)APIMFieldMeasurementAPIM 2001vsAPIM 2003Change in Area2001 to 2003Field Estimate2001 vs FieldMeasurement2003Difference BetweenMethods <strong>of</strong>Estimating AreasField Measurement2003 vsAPIM 2003Hybrids 5,100.86 1,580.11 8,501.62 5,665.59 67% 259% -33%Citation MarshCogswell Marsh (No. Quad.)Alameda Island - North ElsieRoemerIdeal MarshIIIIIIHybrids 244.83 2,059.735 2,448.35 7,803.41 900% 279% 219%Hybrids 700.12 1,503.16 700.12 5,334.08 0% 255% 662%Hybrids 2,882.37 1,171.89 3,610.10 3,948.94 25% 237% 9%Hybrids 2,229.01 1,070.72 3,863.62 3,541.00 73% 231% -8%Average Avg Avg 2,231.43 1,477.12 3,824.74 5,258.60 213% 252% 170%interpretation and field measurement methods. The aerialphoto interpretation measurements <strong>of</strong> area covered byhybrids were on average 170% less than <strong>the</strong> fieldmeasurements (Table 3). However, both methods recorded atleast a doubling <strong>of</strong> area invaded in two years. The aerialphoto interpretation measurements indicated that <strong>the</strong> hybridsincreased 322% at Site Type I and 49% at Site Type II.This discrepancy affects <strong>the</strong> estimates <strong>of</strong> total acreageinvolved but does not substantially change <strong>the</strong> estimates <strong>of</strong>rate <strong>of</strong> increase <strong>of</strong> <strong>the</strong> invasion between 2001 and 2003,which is approximately doubled or more, whe<strong>the</strong>r averagedby region or by site type. The relationship <strong>of</strong> aerial photomeasurements <strong>of</strong> area to field measurements <strong>of</strong> area wastested statistically using a t-test and linear regression (onlog-transformed data) <strong>of</strong> field estimates on photo estimatesfor five sites in 2003. No difference in mean area coveredwas found between methods (t-test, p = 0.616). However, noregression relationship was found (p = 0.797). This waslikely due to high variability among <strong>the</strong> small sample size <strong>of</strong>aerial photo measurements. Thus, aerial photo interpretationmeasurements should be used with caution to estimate cover.Accuracy <strong>of</strong> Field IdentificationOver <strong>the</strong> course <strong>of</strong> <strong>the</strong> 2003 monitoring season, 12landowners, managers or concerned stakeholders requestedsurveys for <strong>the</strong> presence <strong>of</strong> non-native Spartina (Fig. 2). Inaddition, genetic tests were performed to confirm speciesidentification by field staff at each <strong>of</strong> <strong>the</strong> inventorymonitoring sites (Table 1). A total <strong>of</strong> 68 plant samples wereidentified to species in <strong>the</strong> field using plant morphology andanalyzed using genetic tools (Daehler and Strong 1997;Ayres et al. 1999). Fifty-five <strong>of</strong> <strong>the</strong> plants were fieldidentified as S. alterniflora hybrids, and 13 as S. foliosa.Forty-nine <strong>of</strong> <strong>the</strong> plants were confirmed genetically to be S.alterniflora hybrids, and 18 were confirmed to be S. foliosa.There was not a statistically significant difference betweenfield classification and subsequent genetic (or true)identification at <strong>the</strong> 0.05 level (Chi Square). More native S.foliosa were called hybrid than vice versa. 89% <strong>of</strong> <strong>the</strong>hybrids and 100% <strong>of</strong> <strong>the</strong> S. foliosa were correctly identified.EfficacySites treated in 2002 were monitored in 2003 as well aslimited monitoring for <strong>the</strong> few sites that were treated in 2003(Table 4.) Manual Methods such as digging <strong>of</strong> S. densifloraat Piper Park or trampling and covering in Point ReyesNational Seashore (PRNS) were effective on a small scale.Return visits were required to make sure no new shoots hademerged.DISCUSSIONSpreadBased on <strong>the</strong> ISP’s 2003 survey results, <strong>the</strong> averageincrease in area (calculated by site) <strong>of</strong> all non-nativeSpartina species was 244% over a period <strong>of</strong> two years. ForS. alterniflora and hybrids, <strong>the</strong> increase was 317%. Applyingthis rate <strong>of</strong> increase, which represents an average across allsite types with differing and variable environmentalconditions and stages <strong>of</strong> invasion, to <strong>the</strong> 190 ha (470 ac)recorded in <strong>the</strong> 2001 survey suggests that hybrids could havespread to as high as 793 ha (1,960 ac). In contrast, S.alterniflora spread only 25% in 11 years in Willapa Bay,*No field measurement collected in 2001. 2001 field estimate was calculated using a regression curve based on a correlation between <strong>the</strong> existing 2001 and 2003 field measurements.-78-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadTable 4. Change in Area <strong>of</strong> Non-native Spartina at Pre- and Post-treatment Sites Monitored in 2003.Site namePiper Park(entire treatment areaincluding east stripmarsh)Piper Park(primary treatment areaw/out east strip marsh)PRNS-Drakes/LimantourEsteroTreatmentYear(s)2002 Post 0.03/ 123.02002 PostArea2001 Ac/m 2 2002Ac/m 2 2003Ac/m 20.03/122.172002 Post n/a0.02/97.120.02/85.920.06/233.53PRNS-Drakes Estero 2003 Pre n/a n/aBolinas Lagoon North 2002 Post n/a0.002/7.07% Change2002-2003% Change2001-20030.01/ 50.47 -48.0% -59.0%0.01/ 23.00 -73.2% -81.2%0.00005/0.200.005/19.630.00001/0.03-99.9% n/aTreatment & CommentDig Winter 2003;2002-3 treatment, volunteers did not finish<strong>the</strong> entire treatment area. Nor did <strong>the</strong>y digevery plant in <strong>the</strong> primary treatment areathus not 100% kill.Dig Winter 2003;2002-3 treatment, primary treatment area,however volunteers did not dig every plantthus not 100% kill.Trample & Cover Summer 2002; Oneclone at Creamery Bay had a patch thatgrew out from under <strong>the</strong> tarp.n/a n/a Trample & Cover Fall 2003.-99.6% n/aBolinas Lagoon South 2003 Pre n/a n/a 38.47/ 0.01 n/a n/a*Emeryville CrescentEmeryville Crescent –Mowed PortionEmeryville Crescent –Mowed & CoveredPortionRichmond Inner Harbor– Steege MarshAlameda Island – NorthElsie Roemer200220032002200320022003Pre &PostPre &PostPre &Post0.05/215.680.05/205.2775n/an/a0.00/ 3.53 n/a2003 Pre 0.01/ 32.19 n/a2003Pre &PostPier 94 2003 Pre0.53/2143.970.04/171.72n/a* Pre-treatment data only. No percent change results.n/an/a0.63/2547.520.38/1539.580.08/306.150.02/100.460.53/2125.050.08/318.94n/a 1081.2%n/a 650.0% Mow 2003.n/a 8566.7%n/an/a*n/a -0.88n/an/a*Dig Winter 2002; Return visits foundoccasional new sprouts.Trample & CoverSummer 2003.Mow & CoverSummer 2003; Small scale 2002 mowingtreatment had no effect thus mostly seeingspread.Mow & CoverSummer 2003.Trample & CoverFall 2003.Mow/Mow & Spray2002/Fall 2003; GIS based area calc 2003.Area calculations may be "imperfect".Dig, Trample & CoverSummer 2003.Washington from 800 ha (1977 ac) to 1,000 ha (2471 ac)between 1988 and 1999 (Daehler and Strong 1996). Theformation <strong>of</strong> hybrids between S. alterniflora and S. foliosamay have greatly increased <strong>the</strong> spread <strong>of</strong> hybrid Spartinarelative to S. alterniflora; e.g., hybrids expanded 150% inCogswell Marsh, Hayward, California between 1999 and2000, including some individual hybrid clones that increased300% (Zaremba, 2001). Ayres and Strong (2000) reported aremarkable 740% increase <strong>of</strong> S. alterniflora-hybrids from5% to 42% at San Lorenzo Marsh between 1997 and 2000. Ithas been proposed that <strong>the</strong> formation <strong>of</strong> hybrids expains <strong>the</strong>rapid rate <strong>of</strong> expansion in <strong>the</strong> San Francisco Estuary relativeto o<strong>the</strong>r estuaries where hybridization does not occur. Ayreset al. (2004) speculated that <strong>the</strong> recent rapid spread <strong>of</strong>hybrids in <strong>the</strong> Estuary may be a result <strong>of</strong> selection forhybrids with exponential clonal growth and seed production.Theoretical models found greater-than-exponential spreadrates can occur when clonal growth and seed production areunder selection (Hall et al. this vol.; Hall et al. 2006).Cover by non-native Spartina increased at all site typesfrom 2001 to 2003, however, site types differed in <strong>the</strong> rate atwhich cover increased. The large variations in spread ratesamong site types likely reflects <strong>the</strong> proximity <strong>of</strong> individualssites to seed source in addition to habitat suitability.Fringing tidal marshes, mudflats, and estuarine beaches(Site Type II) experienced <strong>the</strong> greatest increase in cover(504% increase). Tidal and microtidal marshes, formerlydiked baylands, and back barrier marshes (Site Type I ) andurbanized shorelines (Site Type IV) increased by about170%. Cover in tidal sloughs, creeks, and flood controlchannels (Site Type III) increased by 108%. The differencesamong site types may be related to successional processesPre-PostTreatment-79-

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinalinked to sedimentation consistent with S. alterniflora and S.townsendii invasions in New Zealand and S. anglica inEngland (Daehler and Strong 1996), and predictions byCallaway and Josselyn (1992).One interpretation <strong>of</strong> <strong>the</strong> spread rate differences amongsites is that invasion rates <strong>of</strong> susceptible habitat are initiallyhigh and <strong>the</strong>n slow until more susceptible habitat is createdby sediment accretion. Fringing marshes, mudflats andbeach habitat (Site Type II) experienced <strong>the</strong> highest rate <strong>of</strong>spread from 2001 to 2003, suggesting a new invasion frontwas being exploited by hybrids. Presumably, after <strong>the</strong> initialrapid colonization <strong>of</strong> suitable “empty-niche” habitat, <strong>the</strong> rate<strong>of</strong> spread would slow to match <strong>the</strong> rate at which <strong>the</strong> Spartinaaugments sedimentation and creates additional habitat toinvade.Established S. foliosa marshes and open mudflats <strong>of</strong> anewly opened restoration sites (Site Type I) are initiallyhighly susceptible to invasion by pollen and by seed,respectively. The formerly diked bayland Citation Marshexperienced a 900% increase in hybrid cover in three years.Then, <strong>the</strong> rate <strong>of</strong> cover increase slowed as suitable habitatwas filled in.Site Type IIIs may be initially slower to colonizebecause <strong>the</strong>y include deep channels and creeks, wheresediment must accrete before <strong>the</strong>y are at suitable elevationfor more extensive colonization. Once <strong>the</strong> channel beds havesufficiently accreted, colonization <strong>of</strong> remaining channelbanks and bottoms is rapid, with clonal colonies quicklycoalescing into meadows. Rate <strong>of</strong> spread in urbanizedshoreline (site type IV) may likewise be relatively low dueto poor or little sediment availability.Among <strong>the</strong> three non-native Spartina taxa, hybridsspread <strong>the</strong> most rapidly (317% in three years). Hybrids havehigh rates <strong>of</strong> vegetative spread, produce large quantities <strong>of</strong>pollen, have successful seed set, and readily backcross to <strong>the</strong>native S. foliosa. Spartina foliosa exposed to pollen <strong>of</strong>hybrids produces hybrid seeds, fur<strong>the</strong>r accelerating <strong>the</strong>hybrid invasion. The rate <strong>of</strong> colonization is particularly rapidin <strong>the</strong> early stages <strong>of</strong> invasion. For example, EmeryvilleWest showed an increase in cover <strong>of</strong> 2,175%, between 2001and 2003. All habitats in <strong>the</strong> Bay appear to be susceptible toinvasion by this S. alterniflora hybrid.S. densiflora, which has a cespitose growth form andinvests more reproductive effort in seed production thanvegetative spread, didn’t spread to <strong>the</strong> same degree as didhybrids: 52% averaged over all site types. Site Type Iexperienced <strong>the</strong> greatest increase in cover <strong>of</strong> S. densiflora(68%).S. patens apparently decreased in cover by 84%, but it isfound only at Southampton Marsh where <strong>the</strong>re is less thanan acre <strong>of</strong> total cover. The apparent decrease is likely due tomapping error not a true decrease in cover.Of <strong>the</strong> three Bay regions, <strong>the</strong> Central Bay, nearest to <strong>the</strong>original introduction sites in <strong>the</strong> Fremont and San Brunomarshes, had <strong>the</strong> largest increase in cover, 292%. Hybrids,<strong>the</strong> fastest spreading species (392%), dominate this region <strong>of</strong><strong>the</strong> Bay. Clearly, <strong>the</strong> hybrids are <strong>the</strong> dominant invasiveSpartina taxa and are well established, and both <strong>the</strong>iracreage and rate <strong>of</strong> spread is greatest in <strong>the</strong> heart <strong>of</strong> <strong>the</strong> Bay.Some new non-native Spartina populations were foundin <strong>the</strong> sites surveyed at <strong>the</strong> request <strong>of</strong> land owners,managers, and stakeholders. In <strong>the</strong> South Bay, a survey <strong>of</strong><strong>the</strong> Plummer Creek Mitigation Site found only nativeSpartina. However, 19% <strong>of</strong> seedlings from a Spartina spp.population at <strong>the</strong> Cooley Landing/Ravenswood PreserveRestoration Project were genetically tested and determinedto be hybrids. Both <strong>of</strong> <strong>the</strong>se sites are near extensive hybridpopulations, and considered to be at risk. Genetic surveyswere performed at a few small sites in <strong>the</strong> Central Bay(Crissy Field, Steege Marsh and Beach Drive) as part <strong>of</strong> amanagement regime to identify and remove newlyestablished non-natives.GISThree difficulties emerged during <strong>the</strong> 2003 monitoring:imprecision <strong>of</strong> measuring small clonal patches or small areasdue to <strong>the</strong> limits <strong>of</strong> precision <strong>of</strong> <strong>the</strong> GPS units, differencesbetween aerial photo interpretation estimates and groundtruthdata, and low power to detect small changes in areacoverage because <strong>of</strong> broad cover class intervals (Zaremba etal. 2004). Field mapping with three-meter- resolution GPSunits is imprecise for small areas such as new invasion sites,which necessarily contain small populations, and <strong>the</strong>physically smaller Spartina species, S. densiflora and S.patens. These possible sources <strong>of</strong> error could account for <strong>the</strong>apparent decrease in cover <strong>of</strong> S. patens at SouthhamptonMarsh or S. densiflora at Pickleweed Park or Blackie’sPasture and Creek.Remote sensing using aerial infra-red photography haspotential for synoptic mapping <strong>of</strong> <strong>the</strong> spread <strong>of</strong> invasiveSpartina compared with labor-intensive field mapping.However, among <strong>the</strong> five sites where we compared <strong>the</strong>semethods in 2003 <strong>the</strong>re were two sites where aerial estimateswere 8-33% less than field estimates and three sites whereaerial estimates were 9-661% greater than field estimates.Averaged over all five sites, <strong>the</strong> field measurements were170% greater than <strong>the</strong> aerial photo measurements. Aerialphoto interpretation methods may prove useful to monitor<strong>the</strong> spread <strong>of</strong> non-native Spartina in grossly invaded areas,but it is not currently precise enough to map without fieldtruthing. The examples <strong>of</strong> both higher and lower aerial photointerpretation cover estimates indicate that <strong>the</strong> currentmethod <strong>of</strong> digitizing polygons around <strong>the</strong> marsh or Spartinapatch with a cover class may be too coarse to estimate coverprecisely.-80-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadBroad cover-class boundaries also contribute toimprecision. For example, if a polygon area is within a coverclass, e.g. 10-30%, <strong>the</strong>n it is assigned <strong>the</strong> midpoint <strong>of</strong> <strong>the</strong>class, 15% in this example. In both 2001 and 2003 CogswellMarsh fell within cover class 10-30%. Ei<strong>the</strong>r <strong>the</strong>re was nochange in cover between 2001 and 2003 or <strong>the</strong> differencewas not large enough for 2003 to be assigned to <strong>the</strong> nextlarger class. Thus, a trend <strong>of</strong> expansion or contraction willbe detected only when it crosses a cover class boundary.However, <strong>the</strong> cover classes are useful for determining anapproximate infestation level and can be used to prioritizeremoval efforts.Accuracy <strong>of</strong> Field IdentificationGenetic tests were also performed to confirm speciesidentification by field staff at each <strong>of</strong> <strong>the</strong> InventoryMonitoring Sites. The genetic tests confirmed that field staffwere accurately identifying both S. alterniflora hybrids(89%) and S. foliosa (100%). The samples labeled as“unknown” or “unidentifiable” by field staff were found justas likely to be hybrids as S. foliosa (55%-45%).Trained field biologists can accurately identify <strong>the</strong>majority <strong>of</strong> non-native Spartina plants, but genetic testing <strong>of</strong>ambiguous hybrid specimens is necessary to reduce <strong>the</strong> risk<strong>of</strong> overlooking non-natives and <strong>the</strong>ir hybrids.EfficacyPost-treatment monitoring was done at five sites thatwere treated in 2002 and pre-treatment monitoring wasdone at seven sites. Manual removal (digging) <strong>of</strong> S.densiflora at Piper Park was effective; however, a number<strong>of</strong> divots remained visible in <strong>the</strong> marsh plain severalmonths after <strong>the</strong> clones were removed. It is uncertain howlong <strong>the</strong>se features will remain in <strong>the</strong> marsh and it isunclear what effect <strong>the</strong>y may have on <strong>the</strong> quality <strong>of</strong> <strong>the</strong>marsh habitat. We observed that volunteer removal effortsare excellent for public outreach, but it was frequentlynecessary for organizers and ISP staff to complete workbegun by volunteers.Clones in Bolinas Lagoon and Point Reyes that hadbeen dug or trampled and covered were reduced by 95%.However, <strong>the</strong> clone that was dug out in Bolinas Lagooncontinued to produce some shoots around <strong>the</strong> clone edgethat required pulling. A single clone at Pier 98, which hadbeen partially dug out and covered with geo-textile fabricby volunteers, had a number <strong>of</strong> remaining shoots thatrequired fur<strong>the</strong>r control. Clones must be entirely coveredfor <strong>the</strong> treatment to be effective. Roots and rhizomesshould be confirmed dead as was done at Point Reyesbefore uncovering treated clones. Preliminary posttreatmentmonitoring <strong>of</strong> <strong>the</strong> Emeryville treatment site thatwas mowed in 2002 indicated no noticeable treatmenteffect.CONCLUSIONS AND RECOMMENDATIONSData from 28 sites showed that non-native Spartinaspecies are spreading at a rapid rate in <strong>the</strong> San Francisco BayEstuary. Between 2000-2001 and 2003, <strong>the</strong> average percentincrease in area covered by all non-native Spartina speciesin <strong>the</strong> Estuary was 244%, and by hybrids, 317%. Some siteshad several hundred percent increases. While hybridsincreased in area at all site types, spread was greatest forfringing tidal marshes, mudflats, and estuarine beaches (SiteType II). There was a slight trend in increase <strong>of</strong> hybrid coverwith latitude (south to north trend). Aerial photo areameasurements were compared with field measurements atfive sites where both methods were used in 2003. There wasno difference in <strong>the</strong> mean coverage between methods, butgiven <strong>the</strong> small sample size and that no regressionrelationship was found, aerial photo interpretationmeasurements should be used with caution to estimate fieldmeasurements for a single site. The increasingly rapid rate <strong>of</strong>spread <strong>of</strong> Spartina, in particular hybrids, continues tothreaten existing habitat and species assemblages andpotentially threatens <strong>the</strong> success <strong>of</strong> ongoing and plannedrestoration projects within <strong>the</strong> San Francisco Estuary andouter coast marshes. Successful control will only beachieved with dedication <strong>of</strong> adequate resources, attention t<strong>of</strong>ollowing protocols completely, and follow-up monitoringand periodic re-treatment as needed.ACKNOWLEDGMENTSFunding for <strong>the</strong> Monitoring Program was provided by<strong>the</strong> California Coastal Conservancy and <strong>the</strong> CalFed BayDelta Program (Interagency agreement 4600001875).Thank you to Alex Lee who performed <strong>the</strong> geneticanalyses and to Aimee Good, Johanna Good, TrippMcCandlish, Erik Grijalva, numerous landowners,managers, and stakeholders for <strong>the</strong>ir able assistance with <strong>the</strong>fieldwork. We also thank Peggy Ol<strong>of</strong>son, Director <strong>of</strong> <strong>the</strong>Invasive Spartina Project, and Patricia Bossak and MaxeneSpellman <strong>of</strong> <strong>the</strong> California Coastal Conservancy for <strong>the</strong>iradministrative and staff assistance. Peggy Ol<strong>of</strong>son andOliver Burke deserve special thanks for <strong>the</strong>ir editorialassistance.REFERENCESAnttila, C.K, A.R. King, C. Ferris, D.R. Ayres, and D.R. Strong.2000. Reciprocal hybrid formation <strong>of</strong> Spartina in San FranciscoBay. Molecular Ecology 9:765-771.Anttila, C.K., C.C. Daehler, N.E. Rank, and D.R Strong. 1998.Greater male fitness <strong>of</strong> a rare invader (Spartina alterniflora,Poaceae) threatens a common native (S. foliosa) with hybridization.American Journal <strong>of</strong> Botany 85:1597-1601.ArcView 3.3 (2002). Environmental Systems Research Institute,Inc. (ESRI), 380 New York Street, Redlands, California.http://www.esri.com.Ayres, D.R., K. Zaremba, C.M. Sloop, D.R.Strong. 2008. Sexualreproduction <strong>of</strong> cordgrass hybrids (Spartina foliosa x al--81-

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaterniflora) invading tidal marshes in San Francisco BayDiversity and Distributions 9:187-195.Ayres, D.R., D.R. Strong, and P. Baye. 2003. Spartina foliosa—acommon species on <strong>the</strong> road to rarity? Madroño 50: 209-213.Ayres, D.R., D. Smith, K. Zaremba, S. Klohr, and D.R. Strong.2004. Spread <strong>of</strong> exotic cordgrasses and hybrids (Spartina sp.) in<strong>the</strong> tidal marshes <strong>of</strong> San Francisco Bay. Biological Invasions6:221-231.Ayres, D.R., D. Garcia-Rossi, H.G. Davis, and D.R. Strong. 1999.Extent and degree <strong>of</strong> hybridization between exotic (Spartina alterniflora)and native (S. foliosa) cordgrass (Poaceae) in California,USA determined by random amplified polymorphic DNA(RAPDs). Molecular Ecology 8:1179-1186.Ayres, D.R. and D.R. Strong. Evolution <strong>of</strong> Invasive Spartina Hybridsin San Francisco Bay. 2010. In: Ayres, D.R., D.W. Kerr,S.D. Ericson and P.R. Ol<strong>of</strong>son, eds. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong><strong>Conference</strong> on Invasive Spartina, 2004 Nov 8-10,San Francisco, CA, USA. San Francisco Estuary InvasiveSpartina Project <strong>of</strong> <strong>the</strong> California State Coastal Conservancy:Oakland, CA. (this volume).California State Coastal Conservancy. 2003. San Francisco EstuaryInvasive Spartina Project: Spartina Control Program. Volume 1:Final Programmatic Environmental Impact Statement/EnvironmentalImpact Report. September 2003. Preparedby and for <strong>the</strong> California State Coastal Conservancy and <strong>the</strong> U.SFish and Wildlife Service. 454 pp. www.Spartina.orgCallaway, J.C. and M. N. Josselyn. 1992. The introduction andspread <strong>of</strong> smooth cordgrass (Spartina alterniflora) in South SanFrancisco Bay. Estuaries 15: 218-226.Cohen, A.C. and J.T. Carleton. 1995. Nonindigenous aquatic speciesin a United States Estuary: a case study <strong>of</strong> biological invasions<strong>of</strong> <strong>the</strong> San Francisco Bay and Delta. U.S. Fish and WildlifeService and National Sea Grant Report No. PB96-166525.Available at http://nas.er.usgs.gov/publications/sfinvade.htm.Collins, J.N., Debra Smith, S. Klohr and K. Zaremba. 2001. Guidelinesto Monitor <strong>the</strong> Distribution, Abundance, and Treatment <strong>of</strong>Non-indigenous Species <strong>of</strong> Cordgrass in <strong>the</strong> San Francisco Estuary.Report to <strong>the</strong> California Coastal Conservancy.Daehler and Strong. 1996. Status, Prediction and Prevention <strong>of</strong>introduced cordgrass Spartina spp. invasions in Pacific Estuaries,USA. Biological Conservation 78: 51-58.Daehler, C.C. and D.R. Strong. 1997. Hybridization between introducedsmooth cordgrass (Spartina alterniflora, Poaceae) and nativeCalifornia cordgrass (S. foliosa) in San Francisco Bay, California,USA. American Journal <strong>of</strong> Botany 85: 607-611.Goals Project. 1999. Baylands Ecosystem Habitat Goals. A report<strong>of</strong> habitat recommendations prepared by <strong>the</strong> San Francisco BayArea Wetlands Ecosystem Goals Project. First Reprint. U.S. EnvironmentalProtection Agency, San Francisco, Calif./S.F. BayRegional Water Quality Control Board, Oakland, Calif.GPS Pathfinder Office v.3 (2003). Trimble Navigation LimitedMapping & GIS Systems, 645 North Mary Avenue, P.O. Box3642, Sunnyvale, California. www.trimble.com.Hall, R.J. A.M. Hastings and D.R. Ayres. 2006. Explaining <strong>the</strong>explosion : modelling hybrid invasions. <strong>Proceedings</strong> - Royal Society<strong>of</strong> London. Biological Sciences 273 :1385-1389.Hall, R.J., A.M. Hastings and D.R. Ayres. 2010. Modeling <strong>the</strong>spread <strong>of</strong> invasive Spartina hybrids in San Francisco Bay. 2010In: Ayres, D.R., D.W. Kerr, S.D. Ericson and P.R. Ol<strong>of</strong>son, eds.<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on InvasiveSpartina, 2004 Nov 8-10, San Francisco, CA, USA. San FranciscoEstuary Invasive Spartina Project <strong>of</strong> <strong>the</strong> California StateCoastal Conservancy: Oakland, CA. (this volume).Josselyn, M.,B. Larsson, and A. Fiorillo. 1993. An ecological comparison<strong>of</strong> an introduced marsh plant, Spartina alterniflora, inSan Francisco Bay. Technical Report <strong>of</strong> <strong>the</strong> Romberg TiburonCenter, Tiburon, CA.National Invasive Species Council. 2001. Meeting <strong>the</strong> InvasiveSpecies Challenge: Invasive Species Management Plan. 80pp.San Francisco Bay Conservation and Development Commission(BCDC). 2001. San Francisco Bay Ecology and Related Habitats.Staff Report September 28, 2001. http://www.bcdc.ca.gov.SPSS (2000). SYSTAT 10 Statistics. SPSS, Inc. Chicago, Illinois.www.systat.com.Strahlberg, D., V. Toniolo, G.W. Page and L.E. Stenzel. 2010.Potential impacts <strong>of</strong> Spartina spread on shorebird populations inSouth San Francisco Bay. In: Ayres, D.R., D.W. Kerr, S.D. Ericsonand P.R. Ol<strong>of</strong>son, eds. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong><strong>Conference</strong> on Invasive Spartina, 2004 Nov 8-10, SanFrancisco, CA, USA. San Francisco Estuary Invasive SpartinaProject <strong>of</strong> <strong>the</strong> California State Coastal Conservancy: Oakland,CA. (this volume).Wolf, D.E., N. Takebayashi, and L.H. Rieseberg. 2001. Predicting<strong>the</strong> risk <strong>of</strong> extinction through hybridization. Conservation Biology15:1039-1053.Zaremba K. 2001. Hybridization and Control <strong>of</strong> a Native-Non NativeSpartina Complex in San Francisco Bay. Master <strong>of</strong> Arts <strong>the</strong>sis,San Francisco State University, San Francisco, California.Zaremba K. 2001. San Francisco Estuary Invasive Spartina ProjectField Identification Guides. Prepared for <strong>the</strong> California CoastalConservancy.Zaremba, K, M. McGowan, D. Ayres. 2004. San Francisco EstuaryInvasive Spartina Project Monitoring Report for 2003.-82-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadREMOTE SENSING,LIDAR AND GIS INFORM LANDSCAPE AND POPULATION ECOLOGY,WILLAPA BAY,WASHINGTONJ.C. CIVILLE 1,4 ,S.D.SMITH 2 AND D.R. STRONG 31 Department <strong>of</strong> Evolution and Ecology, University <strong>of</strong> California, Davis, One Shields Ave., Davis, CA 95616-87552 True North GIS, 4857 Grand Fir Lane, Olympia, WA 985023 Department <strong>of</strong> Evolution and Ecology, University <strong>of</strong> California, Davis, One Shields Ave., Davis, CA 95616-87554 Current address: 2731 Be<strong>the</strong>l St. NE, Olympia, WA 98506; jciville@comcast.netKeywords: Spartina alterniflora, Willapa Bay, photogrammetry, mappingThe spread <strong>of</strong> Spartina alterniflora Loisel. (smoothcordgrass, hereafter referred to as Spartina) in PacificNorthwest estuaries presents a unique opportunity toexamine ecological interactions between an invasive clonalorganism and local abiotic factors. Willapa Bay is a large,shallow, tidal basin on <strong>the</strong> southwest coast <strong>of</strong> WashingtonState. The bay is approximately 360 square kilometers (km 2 )at mean high tide, with well over 190 km 2 <strong>of</strong> s<strong>of</strong>t mud andsand tideflats exposed at mean low tide (Civille 2005). Themixed diurnal tides <strong>of</strong> <strong>the</strong> nor<strong>the</strong>rn Pacific coast distinguishWashington estuaries from those on <strong>the</strong> Atlantic coast whereS. alterniflora is <strong>the</strong> predominant native salt marsh species.Spartina species are <strong>the</strong> principle components <strong>of</strong> Atlanticand Gulf coast estuaries, but <strong>the</strong> introduction <strong>of</strong> Spartina to<strong>the</strong> Willapa Bay estuary and its open mudflats has led to <strong>the</strong>rapid conversion <strong>of</strong> thousands <strong>of</strong> intertidal hectares intodense stands <strong>of</strong> upper tidal meadows (Sayce 1988; Daehlerand Strong 1996; Civille et al. 2005). The rapid expansion <strong>of</strong>this robust clonal grass across <strong>the</strong> open habitat <strong>of</strong> intertidalmudflats is a textbook example <strong>of</strong> unimpeded colonization,and presents challenges not only to management efforts, butto ecologists as well.The initial questions relevant to this research weregenerated by biologists and land managers in <strong>the</strong> state <strong>of</strong>Washington, who needed to understand Spartina expansionfor environmental impact analyses, adaptive managementplans, and to direct control efforts in <strong>the</strong> most efficientmanner (Sayce 1988; Aberle 1990; Aberle 1993; Civille1993). These questions included: where were plantsspreading fastest, where was intertidal habitat being lostmost rapidly, and whe<strong>the</strong>r stopping seed production wasmore effective than removing as many new plants aspossible (Moody and Mack 1988). Ecologists studyinginvasive population growth dynamics ask similar questions.What is regulating <strong>the</strong> growth <strong>of</strong> Spartina on Willapa Baymudflats? Are <strong>the</strong>se regulators density dependent orindependent? With <strong>the</strong> seemingly limitless expanse <strong>of</strong>habitat available for colonization, why is it that some areasare filling in more quickly than o<strong>the</strong>rs? Are <strong>the</strong>re factorsintrinsic to <strong>the</strong> biology and mating system <strong>of</strong> Spartina thatinfluence its spread, or are abiotic factors more important inshaping <strong>the</strong> colonization fronts? Does competition betweenSpartina plants limit <strong>the</strong> population, and will availability <strong>of</strong>resources affect <strong>the</strong>ir growth? Have evolutionary changestaken place in <strong>the</strong> Willapa Bay population that allowSpartina to compete more effectively? Many <strong>of</strong> <strong>the</strong>sequestions have been addressed in experimental and<strong>the</strong>oretical work by members <strong>of</strong> <strong>the</strong> Spartina BiocomplexityGroup at <strong>the</strong> University <strong>of</strong> California at Davis, (Davis et al.2004a; Davis et al. 2004b; Taylor 2004; Taylor and Hastings2004; Civille 2005; Davis 2005; Civille 2006) and werepresented at this conference by <strong>the</strong>ir primary authors (Daviset al.; Taylor et al.; this volume).Spartina was probably brought to <strong>the</strong> Willapa estuarythrough <strong>the</strong> transplanting <strong>of</strong> eastern oysters to bolster <strong>the</strong>flagging oyster industry <strong>of</strong> Willapa Bay in <strong>the</strong> 1890s andearly 1900s. The California gold rush had fueled <strong>the</strong>depletion <strong>of</strong> native oysters from <strong>the</strong> bay, and an effort wasmade by <strong>the</strong> United States Fisheries Bureau to replant <strong>the</strong>bay with Chesapeake stock. The completion <strong>of</strong> <strong>the</strong>transcontinental railway allowed <strong>the</strong> trip to be made in abouta week (Townsend 1896; Scheffer 1945; Civille et al. 2005),and eastern oysters continued to be brought via railcar toWillapa until around 1917. The climate was apparently toocool for eastern oysters, and although <strong>the</strong>y grew well, <strong>the</strong>ywere never able to reproduce successfully in Willapa atcommercially viable levels. Spartina, however, was able toestablish in small colonies that were close to beds used torear eastern oysters (Civille et al. 2005).The first written and photographic evidence <strong>of</strong> Spartinapresence in <strong>the</strong> estuary was ga<strong>the</strong>red at <strong>the</strong> Willapa NationalWildlife Refuge by T.H. Scheffer in 1940 (Scheffer 1945).A photograph taken by Scheffer shows several “greenislands,” and was donated to <strong>the</strong> California Academy <strong>of</strong>Science Herbarium to document his findings. If an averageheight <strong>of</strong> one meter (m) is assumed, <strong>the</strong> large plant in <strong>the</strong>photo is about 42 m in diameter, covering approximately1,367 square meters (m 2 ) <strong>of</strong> mudflat. The first aerialphotographs <strong>of</strong> <strong>the</strong> bay were taken by <strong>the</strong> Army Corps <strong>of</strong>Engineers in 1945, and Spartina plants <strong>of</strong> a similar size to-83-

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinathose in Scheffer’s photo can be seen in <strong>the</strong>se photos atwidely separated locations around <strong>the</strong> bay (Civille et al.2005).A time series analysis <strong>of</strong> Spartina spread across <strong>the</strong>Willapa tidal flats was conducted using 55 years <strong>of</strong> synopticaerial photography, spanning from 1945 to 2000. Themethodology used for <strong>the</strong> study was a combination <strong>of</strong>photogrammetry, image analysis and geographic informationsystem (GIS) techniques (Wilkie 1996). Over 2,000 photos(black and white, true color and color infra-red) weredigitized, examined for Spartina plants, <strong>the</strong>n geo- ororthorectified to fit <strong>the</strong> curvature <strong>of</strong> <strong>the</strong> earth and assignedgeographic coordinates. The resulting corrected images were<strong>the</strong>n classified into <strong>the</strong>matic representations <strong>of</strong> Spartinameadows and clones. The final year <strong>of</strong> <strong>the</strong> series, comprised<strong>of</strong> color infrared images acquired in 2000, was orthorectifiedto remove radial and scale distortions, <strong>the</strong>nclassified through hierarchical supervised and unsupervisednon-parametric methods into 330 <strong>the</strong>matic raster files, withless than one meter positional error and a 0.3 m minimummapping unit (Civille 2005). All o<strong>the</strong>r spatial data weregeorectified to this baseline layer. These tasks wereaccomplished with a large format scanner (28 x43 centimeter [cm] platform), Photoshop (Adobe), andImagine (ERDAS), a robust digital photogrammetrys<strong>of</strong>tware package. The resulting raster files were <strong>the</strong>nconverted into GIS vector coverages in ArcInfo (ESRI), andcombined with a bathymetry digital elevation model (DEM)to obtain z (elevation) values for each polygon.A vector-based coverage model was chosen for ouranalysis, ra<strong>the</strong>r than raster-based, because polygons areautomatically assigned a unique identifier, area andperimeter by ArcInfo s<strong>of</strong>tware (Flynn and Pitts 2000). Thisallowed us to extract demographic history for each plant(greater than 300,000 individuals in 1997 alone) that wedetected on <strong>the</strong> photos. These data were used toparameterize ma<strong>the</strong>matical population models (Taylor et al.2004), and establish <strong>the</strong> ages <strong>of</strong> plants used in genetic andreproductive fitness experiments (Davis et al. 2004).Regions, which are a higher order organizational object inArcInfo, were applied to <strong>the</strong> coverages, allowing us toanalyze clonal growth and meadow formation year-to-year.Each contributing polygon (clone) within a region (largemeadow), maintains its own unique identifier and also has aregion identifier, which are maintained through dataextraction and statistical analyses (Civille 2005).Spartina elevation data were analyzed with twoseparate bathymetry (elevation) layers; one that wascollected by <strong>the</strong> National Oceanic and AtmosphericAdministration (NOAA) with depth soundings in 1953(channels updated in 1986); and ano<strong>the</strong>r that was acquired in<strong>the</strong> spring <strong>of</strong> 2002. The new bathymetry data was collectedwith LiDAR (Light Detection And Ranging) techniques, andis <strong>the</strong> first <strong>of</strong> its kind to be ga<strong>the</strong>red for an entire estuarineintertidal zone with tidal control (Cracknell 1999; Irish andLillycrop 1999; Flowers 2002). Tidal exposure times werecalculated for <strong>the</strong> Willapa estuary during winter monthswhen plant aboveground growth has died back, leavingmudflat elevations as <strong>the</strong> primary targets reflected by <strong>the</strong>laser. Data collection was scheduled for predicted tidalelevations lower than 30 cm Mean Low Water (MLW),although <strong>the</strong> actual tidal elevations at <strong>the</strong> time <strong>of</strong> acquisitionvaried from this target (Civille 2005). The LiDAR data wascollected through a collaborative effort with <strong>the</strong> NOAACoastal Services Center in Charleston, South Carolina, anddata was acquired and compiled by photogrammetrists atSpencer B. Gross in Portland, Oregon.To determine whe<strong>the</strong>r individual Spartina plants weregrowing more rapidly in different abiotic conditions, weextracted areas from polygons that overlapped one ano<strong>the</strong>rin separate years. These were assumed to be <strong>the</strong> sameindividual increasing (or diminishing) in area through time.A stratified random selection <strong>of</strong> 408 <strong>of</strong> <strong>the</strong>se co-occurringclones was extracted from <strong>the</strong> 1994 and 1997 layers, <strong>the</strong>nstatistically analyzed according to a simple relative arealgrowth index [(Area1997-Area1994)/Area1994]. Thisindex relates <strong>the</strong> amount <strong>of</strong> areal growth to <strong>the</strong> initial size <strong>of</strong><strong>the</strong> plant. An ANOVA was performed on <strong>the</strong> 1994-1997data set and significant differences in relative growth rateswere observed between mud and sand substrates and tidalelevation <strong>of</strong> <strong>the</strong> clones (Fig. 1). Clones in s<strong>of</strong>t mudsubstrates with proximity to fresh water sources exhibitedsignificantly higher areal growth rates than those on sand.Clones at higher elevations in general are growing morequickly than those at lower elevations, although differenceswere not significant in all elevational classes. Spartinaplants are noticeably absent from sand substrates lower thanFig. 1: Differences in relative lateral growth rate <strong>of</strong> Spartina comparingtidal elevation and substrate type, 1994-1997.-84-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and Spreadone m above MLW, although present in large numbers at<strong>the</strong>se elevations on mud.Seedling recruitment is ano<strong>the</strong>r demographic propertythat is critical to <strong>the</strong> successful spread <strong>of</strong> S. alterniflora inWillapa Bay. New individuals observed in 1997 (polygonsless than three m in diameter without earlier counterparts)out-numbered those that were present in 1994 by an order <strong>of</strong>magnitude: over 46,000 single new plants versus 4,500 <strong>of</strong>similar size in 1994. By 2000, <strong>the</strong> number <strong>of</strong> newindividuals recruiting into <strong>the</strong> population rose by ano<strong>the</strong>rorder <strong>of</strong> magnitude to more than 155,550 new plants. Again,recruitment was greatest on those parts <strong>of</strong> <strong>the</strong> bay with s<strong>of</strong>tmud substrates and in proximity to fresh water sources(Civille 2005). The lateral growth and merger data areproviding important insights into abiotic factors that controlSpartina growth, and were used to populate and validate<strong>the</strong>oretical model development and exploration (Taylor et al.2004; Taylor and Hastings 2004).The high resolution spatial datasets have provideddemographic and ecological information at a level <strong>of</strong> detailand bay-wide scale that could never have been collected byindividual ecologists with transects and sampling hoopsacross such difficult and remote terrain. The photographicrecord that has been maintained and cataloged providesunique opportunities to look into <strong>the</strong> history <strong>of</strong> this invasion,and may provide insight into <strong>the</strong> mechanisms andcharacteristics <strong>of</strong> o<strong>the</strong>r invasions.ACKNOWLEDGMENTSThis research could not have been possible without <strong>the</strong>photography collected and archived by <strong>the</strong> Washington StateDepartment <strong>of</strong> Natural Resources (WADNR), <strong>the</strong>Washington State Department <strong>of</strong> Transportation and <strong>the</strong> U.S.Army Corps <strong>of</strong> Engineers, District 10. The mapping andphotogrammetry staff in <strong>the</strong>se agencies provided invaluableadvice, support and expertise as this project was developed.Earlier Spartina mapping efforts by Barb Aberle, LindaKunze, Elizabeth Lanzer, and Tom Mumford made thismore sophisticated approach possible. This work was fundedin part by <strong>the</strong> WADNR and by <strong>the</strong> National ScienceFoundation Biocomplexity Grant #DEB0083583 (P.I. AlanHastings).REFERENCESAberle, B.L. 1990. Chronology <strong>of</strong> Spartina control methods inWashington, California, and Oregon. Spartina Workshop Record.S. Harbell. University <strong>of</strong> Washington, Seattle, Washington,USA.Aberle, B.L. 1993. The Biology and Control <strong>of</strong> Introduced Spartina(cordgrass) Worldwide and Recommendations for its Control inWashington. Masters <strong>the</strong>sis. Evergreen State College, Olympia,Washington, USA.Civille, J.C. 1993. Integrated Management Plan for Spartina alterniflorain Willapa Bay, Washington. Washington State Department<strong>of</strong> Natural Resources, Olympia, Washington, USA.Civille, J.C. 2006. Invasion ecology and pattern assessed throughremote sensing: Spartina alterniflora in Willapa Bay, Washington.Doctoral dissertation. University <strong>of</strong> California, Davis, Davis,California, USA.Civille J.C., K. Sayce, S.D. Smith and D.R. Strong. 2005. Reconstructinga century <strong>of</strong> Spartina invasion with historical recordsand contemporary remote sensing. Ecoscience 12(3):330-338.Cracknell, A.P. 1999. Remote sensing techniques in estuaries andcoastal zones - an update. <strong>International</strong> Journal <strong>of</strong> Remote Sensing20(3):485-496.Daehler C.C. and C. Anttila. 1997. Impacts <strong>of</strong> introduced smoothcordgrass (Spartina alterniflora) on Pacific estuaries. Bulletin <strong>of</strong><strong>the</strong> Ecological Society <strong>of</strong> America 78(4 SUPPL):10.Daehler C. C. and D. R. Strong. 1996. Status, prediction and prevention<strong>of</strong> introduced cordgrass Spartina spp. invasions in Pacificestuaries, USA. Biological Conservation 78(1-2):51-58.Davis H.G., C.M. Taylor, J.C. Civille and D.R. Strong. 2004a. AnAllee effect at <strong>the</strong> front <strong>of</strong> a plant invasion: Spartina in a Pacificestuary. Journal <strong>of</strong> Ecology 92:3231-237.Davis H.G., Taylor C.M., Lambrinos J.G. and D.R. Strong. 2004b.Pollen limitation causes an Allee effect in a wind-pollinated invasivegrass (Spartina alterniflora). <strong>Proceedings</strong> <strong>of</strong> NationalAcademy <strong>of</strong> Sciences USA 101: 13804-13807.Davis, H.G. 2005. r-Selected Traits in an Invasive Population. EvolutionaryEcology 19(3):255-274.Flowers, R. 2002. Partnerships and new technologies: <strong>the</strong> JointAirborne Lidar Bathymetry Technical Center <strong>of</strong> Expertise. SeaTechnology 43(1):34-35.Flynn J. and T. Pitts. 2000. Inside ArcInfo. OnWord Press, Albany,New York, USA.Irish, J.L. and W.J. Lillycrop. 1999. Scanning laser mapping <strong>of</strong> <strong>the</strong>coastal zone: <strong>the</strong> SHOALS system. Journal <strong>of</strong> Photogrammetryand Remote Sensing 54(2):123-129.Lillesand, T.M. and R.W. Kiefer. 1999. Remote Sensing and ImageInterpretation. John Wiley & Sons, Inc., New York, New York,USA.Moody, M.E. and R.N. Mack. 1988. Controlling <strong>the</strong> Spread <strong>of</strong>Plant Invasions - <strong>the</strong> Importance <strong>of</strong> Nascent Foci. Journal <strong>of</strong> AppliedEcology 25(3):1009-1021.Sayce, K. 1988. Introduced cordgrass, Spartina alternifloraLoisel., in salt marshes and tidelands <strong>of</strong> Willapa Bay, Washington.,U.S. Fish and Wildlife Service, Ilwaco, Washington, USA.Scheffer, T.H. 1945. The introduction <strong>of</strong> Spartina alterniflora toWashington with oyster culture. Leaflets <strong>of</strong> Western BotanyIV:163-164.Taylor, C.M., H.G. Davis, J.C. Civille, F.S. Grevstad and A. Hastings.2004. Consequences <strong>of</strong> an Allee effect on <strong>the</strong> invasion <strong>of</strong> aPacific estuary by Spartina alterniflora. Ecology 85:3254-3266.Taylor, C.M. and A.M. Hastings. 2004. Finding optimal controlstrategies for invasive species: a density-structured model forSpartina alterniflora. Journal <strong>of</strong> Applied Ecology 41(6):1049-1057.Townsend, C.H. 1896. The transplanting <strong>of</strong> eastern oysters to WillapaBay, Washington with notes on <strong>the</strong> native oyster industry.U. S. Fisheries Commission, Washington, D.C., USA.Wilkie, D.S. and J.T. Finn. 1996. Remote Sensing Imagery forNatural Resources Monitoring. Columbia University Press, NewYork, New York, USA.-85-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadIMPLICATIONS OF VARIABLE RECRUITMENT FOR THE MANAGEMENT OF SPARTINAALTERNIFLORA IN WILLAPA BAY,WASHINGTONJ.G. LAMBRINOS 1,2 ,D.R.STRONG 3,5 ,J.C.CIVILLE 3,4 AND J. BANDO 1,61 Department <strong>of</strong> Environmental Science and Policy, University <strong>of</strong> California, Davis, CA 956162 Current address: Department <strong>of</strong> Horticulture, 4017 Ag and Life Sciences Building, Oregon State University,Corvallis, OR 97331; lambrinj@hort.oregonstate.edu3 Department <strong>of</strong> Evolution and Ecology, University <strong>of</strong> California, Davis, CA 956164 Current address: 2731 Be<strong>the</strong>l St. NE, Olympia, WA 98506; jciville@comcast.net5 drstrong@ucdavis.edu; 6 jun.bando@gmail.comThe spatial expansion <strong>of</strong> Spartina alterniflora populations in Willapa Bay, Washington has beendriven primarily by seedling recruitment. We measured this recruitment at <strong>the</strong> landscape scale usinga series <strong>of</strong> precision GPS guided airboat censuses. Recruitment is highly variable across both spaceand time. Over <strong>the</strong> four years <strong>of</strong> observation, yearly average estuary-wide recruitment varied from alow <strong>of</strong> 12 per hectare (ha) to a high <strong>of</strong> 500/ha. Variability across individual census plots was evengreater, ranging from 0 per ha to over 4000/ha. Several factors influence this variability. 1) Localrecruitment tracks variation in local seed production. 2) Tidal elevation and hydrological conditionsinfluence <strong>the</strong> spatial pattern <strong>of</strong> seed deposition and retention, and this is reflected in micro-spatialrecruitment patterns. 3) Substrate characteristics independent <strong>of</strong> elevation and hydrology significantlyinfluence seedling survivorship and growth. These results have broad implications for <strong>the</strong>management <strong>of</strong> invasive Spartina populations. Poor recruitment years afford windows <strong>of</strong> opportunityfor effective control, but could equally lead to complacency and an inadequate response duringrecruitment pulses. The strong relationship between local seed production and local recruitmentsuggests that control strategies that do not explicitly account for inter-regional dispersal can still besuccessful over <strong>the</strong> short-term, even if long-term management requires an explicit understanding <strong>of</strong>regional dispersal pathways. Finally, <strong>the</strong> strong differences in recruitment and growth patternscaused by substrate differences may allow <strong>the</strong> categorization <strong>of</strong> sites by recruitment risk.Keywords: heterogeneity, invasion, spread, wetlandINTRODUCTIONSpartina alterniflora has spread over 60 square kilometers(km 2 ) <strong>of</strong> <strong>the</strong> approximately 190 km 2 <strong>of</strong> intertidal habitat in WillapaBay, Washington (Civille et al., 2005; Civille et al. 2010).This large-scale invasion threatens to disrupt many importantecosystem services such as habitat for oysters and migratorybirds, nutrient cycling, and storm water run<strong>of</strong>f control (Sayce1988). A consortium <strong>of</strong> local residents, state agencies and <strong>the</strong>Federal government is currently attempting to manage <strong>the</strong> invasionwith <strong>the</strong> ultimate goal <strong>of</strong> eradicating S. alterniflora from<strong>the</strong> estuary.There have been few attempts to comprehensively managesuch a large plant invasion. Property ownership, funding allocations,and <strong>the</strong> idiosyncratic interests <strong>of</strong> stakeholders usuallyforce management to occur on an ad hoc and project-specificbasis. However, given <strong>the</strong> large extent <strong>of</strong> many invasions <strong>the</strong>reis a growing appreciation for <strong>the</strong> need to manage invasiveplants at explicitly landscape scales (With 2002).The management <strong>of</strong> large-scale invasions takes time.Some invasions will likely require ongoing management,while <strong>the</strong> eradication <strong>of</strong> o<strong>the</strong>rs will take years. In <strong>the</strong> case <strong>of</strong>S. alterniflora in Willapa Bay, coordinated control activitiesbegan over 10 years ago.The effective management <strong>of</strong> landscape-scale invasions,<strong>the</strong>refore, requires an understanding <strong>of</strong> how basic properties<strong>of</strong> invasions vary across space and time. Here we report thattwo basic components <strong>of</strong> S. alterniflora demography, seedproduction and seedling recruitment, vary considerablyacross different habitats in <strong>the</strong> bay and across years. Wediscuss <strong>the</strong> consequences <strong>of</strong> this variability for <strong>the</strong> management<strong>of</strong> S. alterniflora in Willapa Bay.METHODSSeedling RecruitmentTo test <strong>the</strong> degree <strong>of</strong> spatial and temporal variation inseedling recruitment we established 44 permanent plots ontidal mudflat that was free <strong>of</strong> S. alterniflora. Plots covered<strong>the</strong> full spatial extent <strong>of</strong> <strong>the</strong> estuary and spanned a range <strong>of</strong>substrate characteristics, distance to established S. alterniflora,and tidal elevation. Plots were 15 m wide andvaried in length depending on local site characteristics. Werecorded <strong>the</strong> location <strong>of</strong> each plot using GPS equipment(Trimble Pathfinder Pro XRS). Beginning in June 2001 we-87-

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaFig. 2. Recruitment <strong>of</strong> S. alterniflora seedlings across 44 permanentplots in Willapa Bay, Washington.Fig. 1. Seedling census using airboat and GPS guided navigation.used airboat surveys to visually census each plot for newseedling recruits (Fig. 1). Using <strong>the</strong> GPS as a navigationalguide we traversed <strong>the</strong> borders <strong>of</strong> each plot in <strong>the</strong> airboat. Ateam <strong>of</strong> three would count <strong>the</strong> total number <strong>of</strong> seedlings observedduring each airboat pass. We used average counts <strong>of</strong><strong>the</strong> three observers as <strong>the</strong> estimate <strong>of</strong> <strong>the</strong> number <strong>of</strong> seedlingsper plot. Seedling counts among observers never variedmore than 10%, and usually were within one or twoseedlings <strong>of</strong> each o<strong>the</strong>r. We conducted yearly censuses eachJune from 2001-2004.Seed ProductionTo test for variability in seed production, we established30 50 m x 50 m permanent plots in established S. alterniflorameadows across <strong>the</strong> full spatial extent <strong>of</strong> <strong>the</strong> invasion.Plots spanned a range <strong>of</strong> meadow ages. Beginning in<strong>the</strong> fall <strong>of</strong> 2002 we collected 15 inflorescences each from 10individual clones within <strong>the</strong> permanent plots. We also estimatedS. alterniflora cover and inflorescence density withineach plot. We placed <strong>the</strong> inflorescences in wet, cold (10 O C)storage to break dormancy. For each clone we calculated <strong>the</strong>germination rate per inflorescence. Using <strong>the</strong>se values and<strong>the</strong> estimates <strong>of</strong> inflorescence density at each site we calculatedan estimate <strong>of</strong> mean viable seed production per squaremeter (m 2 ) at each site.RESULTSRecruitment is episodicOver <strong>the</strong> course <strong>of</strong> <strong>the</strong> study, <strong>the</strong> yearly variation inseedling recruitment was dramatically episodic (Fig. 2).Yearly average estuary-wide recruitment varied from a low<strong>of</strong> 12/ha to a high <strong>of</strong> 500/ha. Site-specific yearly variationwas even greater. For instance, recruitment at <strong>the</strong> TarlattSlough site was 82 seedlings/ha in 2002 and 4,509 seedlings/hain 2003.Recruitment varies with substrateThe spatial variation in recruitment was equally variable.Within a given year, recruitment density across sitesranged from 0/ha to over 4,000/ha. One factor contributingto this variation was substrate type. Muddy sites had morethan five times greater recruitment than sandy or mixed sites(Fig. 3). Most potential sites for Spartina colonization withinWillapa Bay are sheltered from strong wave energy (maximumfetch less than 1 kilometer [km]). Most <strong>of</strong> <strong>the</strong> sandysites in this study are associated with regions <strong>of</strong> high tidalenergy such as channel edges. A few sites were located inrelatively high fetch areas (approximately 4 km) or on barsnear <strong>the</strong> mouths <strong>of</strong> rivers.Control influences local recruitmentMeadows that received control treatments in 2003 hadsignificantly reduced seed production relative to sites that-88-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadFig. 3. Recruitment <strong>of</strong> S. alterniflora seedlings in plots varying insubstrate texture. Values are means ± 1 S.E.Seedlings/ha25002000150010005000NoneControlFig. 4. The influence <strong>of</strong> control efforts in 2003 on local seedlingrecruitment in 2004. Control methods included a variety <strong>of</strong> mechanicalmethods and herbicide application. Values are means ± 1S.E.did not receive control. The local reduction in seed productioninfluenced local (sub-region) seed recruitment <strong>the</strong> followingyear (Fig. 4). Areas such as <strong>the</strong> Palix and Cedar riversthat did not receive control in 2003 had high seed productionthat fall and correspondingly high seedling recruitment<strong>the</strong> following year. In contrast, sites that received controlin 2003 such as Tarlatt Slough and Diamond Point experiencedlow rates <strong>of</strong> seed production that fall and had correspondinglylow recruitment <strong>the</strong> following year.DISCUSSIONCauses <strong>of</strong> variabilitySpartina alterniflora recruitment in Willapa Bay ishighly variable across years and across sub-regions <strong>of</strong> <strong>the</strong>bay. The yearly variation observed in this study corroborateslonger-term data from historical aerial photographssuggesting that <strong>the</strong>re have been strong episodic bursts <strong>of</strong> S.alterniflora colonization (Civille et al. 2005). A significantcomponent <strong>of</strong> this yearly variation in recruitment appears tobe directly related to levels <strong>of</strong> seed production <strong>the</strong> previousyear. For <strong>the</strong> limited two years that we are able to makecomparisons, <strong>the</strong> average estuary-wide recruitment tracksaverage estuary-wide seed production <strong>the</strong> previous year. Inaddition, <strong>the</strong>re is evidence that S. alterniflora seed set is significantlyinfluenced by pollen availability, and that pollenavailability is significantly reduced by wet and cool conditions(Davis et al. 2004, and this volume). Yearly averagebay-wide recruitment is positively correlated with <strong>the</strong> number<strong>of</strong> August degree days (<strong>the</strong> cumulative number <strong>of</strong> growingdegree days in August) and negatively correlated withtotal August precipitation in <strong>the</strong> previous year (Lambrinos,unpublished data).Substrate characteristics appear to play an importantrole in <strong>the</strong> site-to-site variability in recruitment. Spartinaalterniflora seedlings are far less likely to establish at sandysites than at muddy sites. The mechanisms generating thispattern are still unclear. Spartina seeds may be less frequentlydeposited at more erosive sandy sites. In river systemsseed and sediment deposition are closely related(Goodson et al. 2003). In addition, seedlings may be lesslikely to establish at sandy sites because <strong>of</strong> higher erosionand physical stress. Alternatively, seedlings could performpoorly in sandy sediments because <strong>of</strong> nutritional limitations.Increasing sediment sand corresponds with decreasing sedimentnitrogen (N) content at <strong>the</strong>se sites (Tyler et al. 2007).Evidence from transplant experiments and common gardenexperiments indicate that S. alterniflora seedlings growpoorly on sand relative to mud irrespective <strong>of</strong> <strong>the</strong> ambienthydrological conditions (Lambrinos and Bando 2008).Undoubtedly, ano<strong>the</strong>r factor in <strong>the</strong> site-to-site variabilityin recruitment is seed supply. Spartina alterniflora can potentiallydisperse long distances (Howard and Sytsma 2010).However, <strong>the</strong> fact that regions that received control in 2003had correspondingly reduced recruitment levels in 2004,while no reductions in recruitment occurred in regions thatdid not receive control suggests that a sizeable portion <strong>of</strong>recruitment is derived from local seed sources.Implications for managementThe strongly episodic and spatially variable nature <strong>of</strong> S.alterniflora recruitment in Willapa Bay has important implicationsfor management. Years <strong>of</strong> poor recruitment haveprobably slowed <strong>the</strong> rate <strong>of</strong> <strong>the</strong> invasion in Willapa Bay.Both vegetative growth and recruitment by seed contribute-89-

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinato <strong>the</strong> spatial expansion <strong>of</strong> <strong>the</strong> invasion. However, most <strong>of</strong><strong>the</strong> expansion in <strong>the</strong> bay has occurred through seedling recruitmentand subsequent growth <strong>of</strong> clones, not from <strong>the</strong>lateral growth <strong>of</strong> established meadows (Civille et al. 2005).With moderate and low annual budgets, <strong>the</strong> optimal controlstrategy in Willapa Bay is to target nascent populations(Taylor and Hastings 2004; Taylor et al., this volume).However, a “clone only” strategy may be incompatible witho<strong>the</strong>r management objectives such as restoring habitat formigratory birds. Periods <strong>of</strong> low recruitment may provideopportunities to target meadows with little reduction in controleffort against nascent clones. In contrast, periods <strong>of</strong>high recruitment would be poor times to allocate treatmentresources to established meadows.At <strong>the</strong> same time, periods <strong>of</strong> low recruitment may fostercomplacency. This is unlikely to occur with a comprehensiveand long-term management program such as <strong>the</strong> onecurrently being implemented in Willapa Bay. However, poorrecruitment years may hamper implementation <strong>of</strong> such comprehensivemanagement plans. For instance, if initial infestationsare followed by years <strong>of</strong> low recruitment <strong>the</strong>re maybe little impetus to expend resources on control because <strong>the</strong>spatial expansion <strong>of</strong> populations is limited. Even if controlis initiated at a particular site, <strong>the</strong> need to develop a regionalcontrol strategy may not be obvious. Untreated infestationsmay <strong>the</strong>n serve as significant seed sources during subsequentgood seed production years.The poor seedling recruitment at sandy sites comparedto muddy sites has allowed <strong>the</strong> S. alterniflora managementprogram in Willapa Bay to focus control efforts spatially.The sandy Long Beach Peninsula has been given a low priorityfor control, and control resources will only be spen<strong>the</strong>re during <strong>the</strong> final stages <strong>of</strong> <strong>the</strong> control program (K. Murphy,pers. comm.). The ability to prioritize estuaries in terms<strong>of</strong> susceptibility to S. alterniflora recruitment is an importantconsideration given <strong>the</strong> limited resources usually availableand <strong>the</strong> wide spatial extent <strong>of</strong> most estuaries.The ability to spatially allocate control resources is alsoaided by <strong>the</strong> fact that a significant proportion <strong>of</strong> local recruitmentis derived from local seed production. Controlefforts in one area are unlikely to be completely swamped byrecruitment coming from distant or hydrologically disjunctun-controlled areas. This suggests that S. alterniflora controlin different sub-regions <strong>of</strong> an estuary can be managedsomewhat independently (e.g., controlled sequentially) andstill be successful. Long-distance dispersal is certainly nottrivial, however, and long-term management will requireextensive monitoring and a better understanding <strong>of</strong> dispersalpathways.ACKNOWLEDGMENTSWe thank <strong>the</strong> kind and expert assistance <strong>of</strong> several peopleand agencies without whose assistance this work wouldnot have been possible: Brian Couch; Les Holcomb; KyleMurphy; Fritzie Grevstad; Kathleen Sayce; Brett Dumbauld;Dave Heimer; Charlie Stenvall; Todd Brownly; WendyBrown; The Washington Department <strong>of</strong> Fish and Wildlife;The Washington Department <strong>of</strong> Agriculture; The WashingtonDepartment <strong>of</strong> Natural Resources; The Willapa Bay NationalWildlife Refuge. Funding was provided by NSF Biocomplexitygrant DEB0083583.REFERENCESCiville, J.C., K. Sayce, S.D. Smith and D.R. Strong. 2005. Reconstructinga century <strong>of</strong> Spartina alterniflora invasion with historicalrecords and contemporary remote sensing. Ecoscience12(3):330-338.Civille, J.C., S.D. Smith, and D.R. Strong. 2005. Remote sensingLiDAR and GIS inform landscape and population ecology, WillapaBay, WA. In: Ayres, D.R., D.W. Kerr, S.D. Ericson andP.R. Ol<strong>of</strong>son, eds. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong>on Invasive Spartina, 2004 Nov 8-10, San Francisco,CA, USA. San Francisco Estuary Invasive Spartina Project <strong>of</strong> <strong>the</strong>California State Coastal Conservancy: Oakland, CA.Davis, H.G., C.M. Taylor, J.G. Lambrinos and D.R. Strong. 2004.Pollen limitation causes an Allee effect in a wind-pollinated invasivegrass (Spartina alterniflora). <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> NationalAcademy <strong>of</strong> Sciences USA 101:13804-13807.Goodson J.M., A.M. Gurnell, P.G. Angold, and I.P. Morrissey.2003. Evidence for hydrochory and <strong>the</strong> deposition <strong>of</strong> viableseeds within winter flow-deposited sediments: <strong>the</strong> River Dove,Derbyshire, UK. River Research & Applications 19:317-334.Howard, V., and M. Sytsma. 2010. Fragment propagules <strong>of</strong>Spartina alterniflora and potential eastern pacific dispersal. In:Ayres, D.R., D.W. Kerr, S.D. Ericson and P.R. Ol<strong>of</strong>son, eds.<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on InvasiveSpartina, 2004 Nov 8-10, San Francisco, CA, USA. San FranciscoEstuary Invasive Spartina Project <strong>of</strong> <strong>the</strong> California StateCoastal Conservancy: Oakland, CA.Lambrinos, J.G. and K.J. Bando. 2008. Habitat modification inhibitsconspecific seedling recruitment in populations <strong>of</strong> an invasiveecosystem engineer. Biological Invasions 10:729-741.Sayce, K. 1988. Introduced cordgrass, Spartina alternifloraLoisel., in saltmarshes and tidelands <strong>of</strong> Willapa Bay, Washington.Report prepared for <strong>the</strong> Willapa National Wildlife Refuge.USFWS #FWSI-87058 (TS).Taylor, C. M., and A. Hastings. 2004. Finding optimal controlstrategies for an invasive grass using a density-structured model.Journal <strong>of</strong> Applied Ecology. 41:1049–1057Tyler, A.C., J.G. Lambrinos, and E.D. Grosholz. 2007 Nitrogeninputs promote <strong>the</strong> spread <strong>of</strong> an invasive marsh grass. EcologicalApplications 17:1886–1898.With, K.A. 2002. The landscape ecology <strong>of</strong> invasive spread. ConservationBiology 16:1192-1203.-90-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadPOLLEN LIMITATION IN A WIND-POLLINATED INVASIVE GRASS, SPARTINAALTERNIFLORAH.G. DAVIS 1 ,C.M.TAYLOR 2 ,J.G.LAMBRINOS 3 ,J.C.CIVILLE 4,5 AND D.R. STRONG 1,41 Center for Population Biology, University <strong>of</strong> California, Davis, 2320 Storer Hall, One Shields Ave., Davis, CA 95616;hgdavis@sonic.net2 Department <strong>of</strong> Ecology and Evolutionary Biology, Tulane University, 6823 St. Charles Ave., New Orleans, LA 701183 Department <strong>of</strong> Environmental Science & Policy, University <strong>of</strong> California, Davis, One Shields Ave., Davis, CA 956164 Section <strong>of</strong> Evolution & Ecology, University <strong>of</strong> California, Davis, One Shields Ave., Davis, CA 956165 Current address: 2731 Be<strong>the</strong>l St. NE, Olympia, WA 98506; jciville@comcast.netLimits to, and <strong>the</strong> consequences <strong>of</strong>, pollen availability in wind-pollinated plants have been littlestudied. Reproductive failure or depression because <strong>of</strong> a lack <strong>of</strong> available mates could lead topopulation demographic consequences for many species. Of particular interest is how pollenlimitation affects <strong>the</strong> rate <strong>of</strong> spatial spread <strong>of</strong> invasive plants. We performed a manipulative pollenaddition and exclusion study to investigate <strong>the</strong> role <strong>of</strong> pollen limitation in an invasive perennialestuarine grass, Spartina alterniflora. We found pollen impoverishment at <strong>the</strong> low density leadingedge <strong>of</strong> a large invasion, causing an eight fold reduction in seed set among low density plants,though not among <strong>the</strong> high density plants. We found pollen loads on stigmas to be determined bypollen availability in <strong>the</strong> air. Fur<strong>the</strong>rmore, <strong>the</strong> amount <strong>of</strong> airborne pollen is dictated by <strong>the</strong> spatialpattern <strong>of</strong> plants, with much more pollen available over continuous meadows than in areas <strong>of</strong> lowplant density. The delay <strong>of</strong> appreciable numbers <strong>of</strong> seed persists for decades until vegetative growthcoalesces plants into continuous meadows, and this has slowed <strong>the</strong> rate <strong>of</strong> spread <strong>of</strong> <strong>the</strong> invasion.Keywords: Allee effect, invasive species, pollen limitation, Spartina alternifloraWhen Willapa Bay, Washington was first colonized bySpartina alterniflora in <strong>the</strong> 1890s, <strong>the</strong> very slow rate <strong>of</strong>spread gave little reason for concern. However, by <strong>the</strong>1980s, <strong>the</strong>re were vast areas <strong>of</strong> seedling recruitment andlarge meadow formation on mudflats previously unoccupiedby any o<strong>the</strong>r species <strong>of</strong> emergent plant. In 1995, <strong>the</strong> problemhad grown so daunting that <strong>the</strong> Washington State Legislaturedeclared <strong>the</strong> Spartina invasion an “environmentalemergency.” As with many invasive organisms, S.alterniflora maintained an initially low pr<strong>of</strong>ile due to a “lagtime” between colonization and rapid spread (e.g., Kowarik1995). In this paper, we examine <strong>the</strong> ultimate mechanisticcause responsible for <strong>the</strong> Spartina lag phase and <strong>the</strong>consequences <strong>of</strong> its cessation.Invasive S. alterniflora recruits onto open Pacificmudflats as seeds drift in on <strong>the</strong> tides yearly, leaving noseedbank (Woodhouse 1979). Seedlings <strong>the</strong>n growrhizomatously into widely spaced circular clones.Eventually, <strong>the</strong>se isolated clones grow toge<strong>the</strong>r to formcontinuous meadows. Individuals within a low-densitypopulation, such as isolated clones <strong>of</strong> S. alterniflora, maysuffer from depressed reproduction due to a lack <strong>of</strong> sexualpartners. This syndrome, known as an Allee effect (Allee1931), will cause populations to go extinct if <strong>the</strong> populationdensity drops below a threshold. A less well-knownmanifestation is <strong>the</strong> weak, as opposed to strong, Allee effect.Under <strong>the</strong> weak Allee effect, <strong>the</strong> per capita rate <strong>of</strong> growth(<strong>the</strong> growth rate for individuals within <strong>the</strong> population) isdepressed when population density is low, but neverbecomes negative as occurs with a strong effect. We surmisethat <strong>the</strong> Allee effect in Willapa Bay Spartina is weakbecause existing individuals are perennial; while <strong>the</strong>y maymake very few seeds when <strong>the</strong>y are isolated, <strong>the</strong>y surviveuntil population density becomes high with meadowformation and <strong>the</strong> Allee effect ends.To find whe<strong>the</strong>r density has an effect upon <strong>the</strong> number<strong>of</strong> seeds a plant can produce and <strong>the</strong> viability <strong>of</strong> those seeds,we collected five inflorescences from 20 individual S.alterniflora plants at five different sites within Willapa Bayin <strong>the</strong> year 2000 (Long Island, Peninsula, Palix River,Shoalwaters, Porter Point). At each site we collected half <strong>the</strong>inflorescences from low density clones (all individualsisolated within bare mud) and half from high densitymeadows. We assessed both seed set (# seeds / # florets /inflorescence) and germination percentage (# germinations /# florets / inflorescence). We found that considerably fewerisolated clones (37%) than meadow plants (92%) could setany seed at all. Fur<strong>the</strong>rmore, for <strong>the</strong> plants that could set atleast some seed, <strong>the</strong> meadow plants had a much higherproportion <strong>of</strong> seeds (0.30) than <strong>the</strong> isolated clones (0.08)(Davis et al. 2004a). Finally, <strong>the</strong> seeds <strong>of</strong> <strong>the</strong> isolated cloneswere much less viable with a proportion <strong>of</strong> 0.13 germinating-91-

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaas opposed to a proportion <strong>of</strong> 0.34 <strong>of</strong> <strong>the</strong> meadow plants’seeds germinating. There was no obvious effect <strong>of</strong> proximity<strong>of</strong> neighboring clones on isolated individuals. While somelow density areas were more crowded than o<strong>the</strong>rs, <strong>the</strong>re wasno obvious trend towards relatively higher or lower seed setas long as <strong>the</strong> clones were not touching one ano<strong>the</strong>r. Thissuggests that <strong>the</strong> clones need to merge toge<strong>the</strong>r before <strong>the</strong>step-wise increase in seed set is achieved by meadowindividuals. This work is reported in full in Davis et al.(2004a).We <strong>the</strong>n investigated whe<strong>the</strong>r pollen limitation could be<strong>the</strong> mechanism behind <strong>the</strong> Allee effect in Willapa BaySpartina. While this may appear to be an intuitivehypo<strong>the</strong>sis, pollen limitation is generally assumed not tooccur in wind-pollinated plants, such as grasses, conifers andoak trees. There are many studies <strong>of</strong> pollen limitation inanimal-pollinated plants (Burd 1994; Larson & Barrett2000), very few <strong>of</strong> wind-pollinated plants, and none beforethis one <strong>of</strong> wind-pollinated invasive plants. Pollen limitationin animal vectored plants can diminish reproduction andconstrain invasion success (Bar<strong>the</strong>ll et al. 2001; Parker1997). Using simulation and analytic models, we askedwhe<strong>the</strong>r <strong>the</strong> Allee effect in Willapa Bay Spartina could haveslowed <strong>the</strong> rate <strong>of</strong> invasion and found it did produce a lagtime (Taylor et al. <strong>the</strong>se proceedings; Taylor et al. 2004).We conducted an observational and manipulative studyto assess <strong>the</strong> role <strong>of</strong> pollen limitation in <strong>the</strong> ability <strong>of</strong>Willapa Bay S. alterniflora to set seed. To see if <strong>the</strong>re is adifference in pollen deposition rates on stigmas, we collectedstigmas from isolated clones and high density meadows overthree different sites (Cedar River n = 20; Bone River n = 40;Palix River n = 94) and screened one lobe (half) <strong>of</strong> stigmasfor pollen load. We found that meadow plants’ stigmascaptured nine times <strong>the</strong> pollen that <strong>the</strong> isolated clones’ did.Isolated plants on average had less than one pollen grain perstigma lobe while meadow plants had more than six.Although plants at individual sites differed in <strong>the</strong>ir pollenloads, in every case, <strong>the</strong> meadow plants always had more(Davis et al. 2004b). Interestingly, <strong>the</strong> Cedar Rivercollections had very little pollen with even meadow plantshaving on average less than one pollen grain per stigma lobe,probably because it rained on <strong>the</strong> day <strong>of</strong> collection and <strong>the</strong>previous two days suggesting pollen flow is inhibited on wetdays and possibly rainy years.To find how pollen deposition changes over awindward-to-leeward gradient and to see if <strong>the</strong> amount <strong>of</strong>pollen in <strong>the</strong> air is correlated with <strong>the</strong> amount <strong>of</strong> pollendeposited on stigmas, we set out pollen traps and collectedadjacent stigmas. We set 10 traps along each <strong>of</strong> six transects.Oriented along <strong>the</strong> direction <strong>of</strong> <strong>the</strong> prevailing wind: two inwindward isolated clones (I & II), one along <strong>the</strong> windwardmeadow edge (III), one in <strong>the</strong> windward meadow (IV), onein <strong>the</strong> leeward meadow (V) and one 100 m from <strong>the</strong> leewardside <strong>of</strong> <strong>the</strong> meadow across an unvegetated channel (VI)(Davis et al. 2004). We found extremely high correlationbetween <strong>the</strong> pollen loads on <strong>the</strong> traps and stigmas (r = 0.99)supporting <strong>the</strong> hypo<strong>the</strong>sis that <strong>the</strong> amount <strong>of</strong> pollen onstigmas is controlled by pollen abundance on <strong>the</strong> wind. On<strong>the</strong> pollen traps, we found very little pollen anywhere in <strong>the</strong>isolated clones or along <strong>the</strong> meadow edge. Pollen loadsincreased inside <strong>the</strong> windward end <strong>of</strong> <strong>the</strong> meadow andincreased yet more dramatically at <strong>the</strong> leeward end <strong>of</strong> <strong>the</strong>meadow. However, <strong>the</strong> amount <strong>of</strong> pollen droppedprecipitously across <strong>the</strong> unvegetated channel suggesting that<strong>the</strong> gap was too far for nearly all <strong>of</strong> <strong>the</strong> airborne pollen tocross (Davis et al. 2004).To conclusively determine whe<strong>the</strong>r isolated clones aremaking fewer seeds than <strong>the</strong> meadow plants because <strong>of</strong>pollen scarcity, we performed a manipulative experiment bysupplementing and excluding pollen received byinflorescences and leaving control inflorescences open toambient pollen availability. Twenty-four plants received alltreatments and a fur<strong>the</strong>r 20 plants received pollen additionand control treatments only. We found <strong>the</strong> pollen exclusiontreatment reduced seed set in <strong>the</strong> meadow plants by morethan sixfold, but caused no reduction in <strong>the</strong> isolated clones.The pollen addition treatment had no effect on <strong>the</strong> seed set<strong>of</strong> <strong>the</strong> meadow plants, but did raise that <strong>of</strong> <strong>the</strong> isolatedclones by more than threefold. We were not able to saturate<strong>the</strong> plants’ inflorescences with pollen, as an inflorescencehas receptive stigmas for about 10 days (~3 days per stigma)and we were able to supplement on only three consecutivedays. This is likely <strong>the</strong> reason why <strong>the</strong> isolated clones’pollen addition seed set was not as high as <strong>the</strong> meadowplants’ ambient control (Davis et al. 2004). These resultsindicate that isolated clones, colonists at low density, areextremely pollen-limited whereas seed set in <strong>the</strong> highdensity meadow is not limited by pollen. This work isreported in full in Davis et al. (2004b). We found that anAllee effect contributed to <strong>the</strong> lag time in spatial spread <strong>of</strong> S.alterniflora in Willapa Bay. Fur<strong>the</strong>rmore, this Allee effectappears to be <strong>the</strong> result <strong>of</strong> pollen limitation <strong>of</strong> <strong>the</strong> colonizingclones at <strong>the</strong> leading edge <strong>of</strong> <strong>the</strong> invasion. In <strong>the</strong> studiesoutlined here, and also in two years <strong>of</strong> unpublished data, weconsistently found that isolated individuals set very fewseeds. However, <strong>the</strong>se studies were done in <strong>the</strong> singlelocation <strong>of</strong> Willapa Bay, so similar results may or may notbe duplicated in o<strong>the</strong>r locations. We found a wide range inself-compatibility <strong>of</strong> Spartina alterniflora from <strong>the</strong> nativerange (Davis 2005), so <strong>the</strong> ability <strong>of</strong> isolated plants to use<strong>the</strong>ir own seeds in different locations may be largelyinfluenced by <strong>the</strong> source <strong>of</strong> <strong>the</strong> colonizing individuals. Itwould be advisable for managers <strong>of</strong> o<strong>the</strong>r invaded areas toconfirm <strong>the</strong> lack <strong>of</strong> seeds in isolated Spartina alterniflorabefore discounting <strong>the</strong>ir potential as significant seed sources.-92-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadLag times <strong>of</strong>fer a transient opportunity for management<strong>of</strong> invasive species. However, lag times can induce a falsesense <strong>of</strong> security about <strong>the</strong> threat from exotic plants so <strong>the</strong>y<strong>of</strong>ten go unnoticed and uncontrolled until <strong>the</strong> invader beginsto spread aggressively (Parker 2004). These studies raise <strong>the</strong>possibility <strong>of</strong> pollen limitation in o<strong>the</strong>r wind-pollinatedspecies. It is an open question whe<strong>the</strong>r this is a commonphenomenon, or whe<strong>the</strong>r it is more likely to be found amongpopulations that are actively declining or expanding.ACKNOWLEDGMENTSWe are grateful to <strong>the</strong> Washington State Departments <strong>of</strong>Natural Resources, Fish and Wildlife and Agriculture and<strong>the</strong>ir employees for <strong>the</strong> facilitation <strong>of</strong> this research. We alsothank <strong>the</strong> U.C. Davis “Spartina Lab,” all <strong>of</strong> <strong>the</strong> volunteers,editors, reviewers and <strong>the</strong> organizers <strong>of</strong> <strong>the</strong> <strong>Third</strong> AnnualInvasive Spartina <strong>Conference</strong>. This work was funded by <strong>the</strong>U.S. Environmental Protection Agency (EPA) Science toAchieve Results (STAR) program, <strong>the</strong> National ScienceFoundation (NSF) Biocomplexity Grant program (P.I. A.Hastings), a Washington State Sea Grant and <strong>the</strong> NSFIntegrative Graduate Education and Research Trainee(IGERT) program.REFERENCESAllee, W.C. 1931. Animal Aggregations, a Study in General Sociology.University <strong>of</strong> Chicago Press: Chicago.Bar<strong>the</strong>ll, J.F., J.M. Randall, R.W. Thorp & A.M. Wenner. 2001.Promotion <strong>of</strong> seed set in yellow star-thistle by honey bees: evidence<strong>of</strong> an invasive mutualism. Ecological Applications11:1870-1883.Burd, M. 1994. Bateman’s principle and plant reproduction: <strong>the</strong>role <strong>of</strong> pollen limitation in fruit and seed set. Botanical Review60, 83-139.Davis, H.G., C.M. Taylor, J.C. Civille & D.R. Strong. 2004a. AnAllee effect at <strong>the</strong> front <strong>of</strong> a plant invasion: Spartina in a Pacificestuary. Journal <strong>of</strong> Ecology 92:321-327.Davis, H.G., C.M. Taylor, J.G. Lambrinos & D.R. Strong. 2004b.Pollen limitation causes an Allee effect in a wind-pollinated invasivegrass (Spartina alterniflora). <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> NationalAcademy <strong>of</strong> Sciences USA 101:13804-13807.Davis, H.G. 2005. r-Selected traits in an invasive population. EvolutionaryEcology 19:255-274.Kowarik, I. 1995. Time lags in biological invasions with regard to<strong>the</strong> success and failure <strong>of</strong> alien species. In Plant Invasions: GeneralAspects and Special Problems, eds. Pysek, P., Prach, K., Rejmánek,M. & Wade, M. SPB Academic: Amsterdam, pp. 15-38.Larson, B.M.H. & S.C.H. Barrett. 2000. A comparative analysis <strong>of</strong>pollen limitation in flowering plants. Biological Journal <strong>of</strong> <strong>the</strong>Linnean Society 69:503-520.Parker, I.M. 1997. Pollinator limitation <strong>of</strong> Cytisus scoparius, aninvasive exotic shrub. Ecology 78:1457-70.Parker, I.M. 2004. Mating patterns and rates <strong>of</strong> biological invasion.<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> National Academy <strong>of</strong> Sciences USA101:13695-13696.Taylor, C.M., H.G. Davis, J.C. Civille, F.S.Grevstad and A. Hastings.2004. Consequences <strong>of</strong> an Allee effect in an invasive plant:Spartina alterniflora in Willapa Bay, Washington. Ecology85:3254-3266.Taylor, C.M., A. Hastings, H.G. Davis, J.C. Civille, and F.S.Grevstad. 2010. Modeling <strong>the</strong> spread and control <strong>of</strong> Spartina alterniflorain a Pacific Estuary. In: Ayres, D.R., D.W. Kerr, S.D.Ericson and P.R. Ol<strong>of</strong>son, eds. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong><strong>Conference</strong> on Invasive Spartina, 2004 Nov 8-10, SanFrancisco, CA, USA. San Francisco Estuary Invasive SpartinaProject <strong>of</strong> <strong>the</strong> California State Coastal Conservancy: Oakland,CA (this volume).Woodhouse, W.W. 1979. Building Saltmarshes along <strong>the</strong> Coasts <strong>of</strong><strong>the</strong> Continental United States. Special report no. 4. Army Corps<strong>of</strong> Engineers, Coastal Engineering Research Center, Belvoir,Virginia.-93-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadINVASIVE HYBRID CORDGRASS (SPARTINA ALTERNIFLORA X FOLIOSA)RECRUITMENTDYNAMICS IN OPEN MUDFLATS OF SAN FRANCISCO BAYC.M. SLOOP 1,2 ,D.R.AYRES 1 AND D.R. STRONG 11 Section <strong>of</strong> Evolution and Ecology, University <strong>of</strong> California, Davis, One Shields Avenue, Davis, CA 956162 Laguna de Santa Rosa Foundation, 900 Sanford Road, Santa Rosa, CA 95401; christina@lagunafoundation.orgHybrid Spartina are currently expanding <strong>the</strong>ir range in <strong>the</strong> San Francisco Bay (SFB) at a rateexceeding exponential growth. A subset <strong>of</strong> transgressive hybrid Spartina plants that positivelyexceed <strong>the</strong> fitness trait values <strong>of</strong> <strong>the</strong>ir parents are competitively and reproductively superior to bothparents and o<strong>the</strong>r hybrids and likely drive <strong>the</strong> invasion. In order to colonize <strong>the</strong> vast open SFBmudflats and found new populations hybrid cordgrass plants have to evolve self-compatibility andexhibit rapid vegetative and lateral growth. The mudflat tidal cycle covers or exposes plants for upto six hours, so new seedlings have to be robust and fast growing to survive and establish. A smallnumber <strong>of</strong> hybrid and native Spartina have colonized <strong>the</strong> open mudflats along <strong>the</strong> eastern shore <strong>of</strong>SFB. To discern mudflat seedling recruitment dynamics we investigated (1) <strong>the</strong> numbers andlocations <strong>of</strong> recruiting seedlings at three SFB sites in 2003 and 2004 via GPS/GIS, and (2) <strong>the</strong>genetic relationship <strong>of</strong> established adult plants and parentage <strong>of</strong> seedlings using microsatellitemarkers. Our results identify all sampled seedlings as hybrids, and show a dramatic increase inseedling recruitment numbers in 2004. Molecular investigations reveal surrounding, inter-relatedadult plants as <strong>the</strong> most likely sires for most seedlings, and also show evidence <strong>of</strong> increased selffertilizationin isolated plants. We found seedling recruitment to be spatially heterogeneous alongshorelines, with local pockets <strong>of</strong> recruitment at invaded sites and highly local aggregations <strong>of</strong>seedlings in proximity to <strong>the</strong>ir seed (and pollen) parents. These results give support to transgressivehybrid plants as <strong>the</strong> drivers <strong>of</strong> <strong>the</strong> invasion.Keywords: Hybrid Spartina, seedling recruitment, self-compatibilityINTRODUCTIONIn light <strong>of</strong> <strong>the</strong> accelerating spread <strong>of</strong> Spartina hybrids inSFB (Ayres et al. 2004) we propose that highly invasiveindividuals at <strong>the</strong> forefront <strong>of</strong> <strong>the</strong> invasion possess extremephenotypes for traits important in hybrid cordgrass survivaland spread. We posit <strong>the</strong>se traits are i) self-compatibility,which enables single individuals to found new populations;ii) height, which confers <strong>the</strong> ability to grow low on <strong>the</strong> openmud <strong>of</strong> <strong>the</strong> intertidal plane; iii) lateral spread, which anchorsplants into <strong>the</strong> shifting substrate and allows colonization <strong>of</strong>occupied or unoccupied neighboring patches; iv) tolerance tohigh (40 parts per trillion (ppt)) salinity which allows plantsto growth higher on <strong>the</strong> intertidal plane in <strong>the</strong> range <strong>of</strong>Salicornia virginica, and v) <strong>the</strong> timing and abundance <strong>of</strong>flowering, which is important in siring ability on earlyflowering native S. foliosa and in seed production (Ayres etal. 2008). These traits cause <strong>the</strong> selective superiority <strong>of</strong> thoseindividuals that possess all or most <strong>of</strong> <strong>the</strong>m with regard too<strong>the</strong>r hybrids and both parent species, and we propose that<strong>the</strong>se hybrid individuals will drive <strong>the</strong> invasion <strong>of</strong> SFB openmudflats. We have found that Spartina hybrids show highlyvariable self-compatibility, while both hybrid parent species(S. alterniflora and S. foliosa) are marginally self-compatible(H. Davis and D. Ayres, unpublished data) (Fig. 1). Hybridindividuals thus greatly differ in <strong>the</strong>ir ability to successfullyset self-fertilized seed, and certain hybrid plants aretransgressive, exceeding both parent species and o<strong>the</strong>rhybrids in <strong>the</strong>ir ability to self-fertilize (Fig. 1).RESULTS AND DISCUSSIONWe performed GPS surveys in 2003 and 2004 at threesites along <strong>the</strong> eastern shores <strong>of</strong> SFB (Elsie Roemer Marsh,Robert’s Landing, Hayward Shoreline) and found a dramaticincrease in seedling establishment from 2003-2004 (Fig. 2).At Hayward shoreline all genetically surveyed seedlingswere hybrids. Overlapping GPS points and <strong>the</strong> relativelylarger size <strong>of</strong> some individuals at Hayward suggest someseedling survival from 2003 to 2004 (Fig. 2). Microsatelliteparentage analysis (Sloop et al. 2005; Blum et al. 2004;Gerber et al. 2003) <strong>of</strong> a sub-sample <strong>of</strong> GPS-surveyedseedlings and surrounding plants suggests that seedlingsoriginate most likely from local seed sources, ra<strong>the</strong>r thanfrom seeds swept in by <strong>the</strong> tides (Sloop et al. 2009). Ourparentage data show a large likelihood <strong>of</strong> establishingseedlings as <strong>the</strong> progeny <strong>of</strong> <strong>the</strong> surrounding clones. In 200338% <strong>of</strong> all establishing seedlings were most likely selffertilizedat Hayward Shoreline. Microsatellite analysis alsorevealed that at Robert’s Landing <strong>the</strong> majority <strong>of</strong> 2003 seedscollected from isolated hybrid plants were self-fertilized.-95-

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaFig.1: Percent seed set vs. %germination <strong>of</strong> self-fertilizedhybrid seeds from CogswellMarsh, Hayward. Averagegermination/seed set rates for selffertilizedS. foliosa and S.alterniflora are 25%/12% and45%/25% respectively. HybridC3-1 represents a transgressiveindividual exceeding both parentsand o<strong>the</strong>r hybrids in germinationand seed set <strong>of</strong> self-fertilized seed.Fig. 2: Temporal variationin seedling recruitment atHayward shoreline.Distribution <strong>of</strong> seedlingsin Spring 2004 (stars)increased >500-fold fromSpring 2003 (triangles).Overlap <strong>of</strong> stars andtriangles suggests survival<strong>of</strong> 2003 seedlings.-96-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadSpartina hybrid seedling occurrence is highest atinvaded mudflat sites. While recruitment is mainly local,seed exportation on <strong>the</strong> tide will likely intensify aspopulations grow and expand. We predict that transgressivehybrid plants, with <strong>the</strong> greatest reproductive fitness willdrive <strong>the</strong> spread, colonizing new open mudflat sites allaround SFB at an ever-increasing rate. This will not onlyfur<strong>the</strong>r threaten <strong>the</strong> persistence <strong>of</strong> <strong>the</strong> native Californiacordgrass S. foliosa (Ayres et al. 2003), but will likelyirreversibly change <strong>the</strong> character <strong>of</strong> <strong>the</strong> SFB ecosystem.Open mudflats will turn into hybrid Spartina meadows,which over time, after <strong>the</strong> accretion <strong>of</strong> sediment, will changefrom intertidal to terrestrial systems.ACKNOWLEDGMENTSWe would like to thank Team Spartina at UC Davis, andacknowledge <strong>the</strong> financial support from <strong>the</strong> CaliforniaCoastal Conservancy (CalFed grant #99_110), CaliforniaSea Grant #27CN to DRS, and NSF Biocomplexity DEB0083583 to A. Hastings and D.R. Strong.REFERENCESAyres, D.R., D.R. Strong, and P. Baye. 2003. Spartina foliosa(Poaceae): A common species on <strong>the</strong> road to rarity? Madrono.50:209-213.Ayres, D.R., D.L. Smith, K. Zaremba, S. Klohr, and D.R. Strong.2004. Spread <strong>of</strong> Exotic Cordgrasses and Hybrids (Spartina sp.)in <strong>the</strong> Tidal Marshes <strong>of</strong> San Francisco Bay, California, USA.Biological Invasions 6:221-231.Ayres, D.R., K. Zaremba, C.M. Sloop, D.R. and Strong. 2008.Sexual reproduction <strong>of</strong> cordgrass hybrids (Spartina foliosa x alterniflora)invading tidal marshes in San Francisco Bay. Diversityand Distributions 14:187-195.Blum, M.J., C.M. Sloop, D.R. Ayres, and D.R. Strong. 2004. Characterization<strong>of</strong> microsatellite loci in Spartina species (Poaceae).Molecular Ecology Notes 4:39-42.Gerber, S., P. Chabrier, and A. Kremer. 2003. Famoz: a s<strong>of</strong>twarefor parentage analysis using dominant, codominant and uniparentallyinherited markers. Molecular Ecology Notes 3: 479-481.Sloop, C.M, H.G. McGray, M.J. Blum and D.R. Strong. 2005.Characterization <strong>of</strong> 24 additional microsatellite loci in Spartinaspecies (Poaceae). Conservation Genetics 6:1049–1052.Sloop, C.M., D.R. Ayres, D.R. Strong. 2009. The rapid evolution <strong>of</strong>self-fertility in Spartina hybrids (S. alterniflora x foliosa) invadingSan Francisco Bay, CA. Biological Invasions 11:1131-1144.-97-

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadTHE INFLUENCE OF INTERTIDAL ZONE AND NATIVE VEGETATION ON THE SURVIVALAND GROWTH OF SPARTINA ANGLICA IN NORTHERN PUGET SOUND, WA, USAC. E. HELLQUIST 1 AND R. A. BLACKSchool <strong>of</strong> Biological Sciences, Washington State University, Pullman WA, 991641 Current address: Department <strong>of</strong> Biological Sciences, State University <strong>of</strong> New York, Oswego, NY 13126;eric.hellquist@oswego.eduSpartina anglica colonizes mudflats, tidal channels, salt marshes, and gravel beaches in nor<strong>the</strong>rnPuget Sound, Washington, USA. We measured Spartina seedling survival and growth along anintertidal gradient from mudflat to salt marsh at Alice Bay, Skagit County, Washington. Spartinaseedlings were transplanted into plots with full and reduced competition in open mudflat, a lowmarsh intertidal zone dominated by Salicornia virginica, and a middle marsh zone dominated byDistichlis spicata. In <strong>the</strong> mudflat, where little native vegetation was present, two separate treatmentswere monitored: unmanipulated (in situ) seedlings and seedlings transplanted within <strong>the</strong> mudflat.The in situ seedlings had 100% survival whereas seedlings transplanted within <strong>the</strong> mudflat had 60%survival. The mean relative growth rate (RGR) <strong>of</strong> in situ seedlings in <strong>the</strong> mudflat was 16 milligramsper milligram per day (mg mg -1 day -1 ) compared to RGR <strong>of</strong> 8 mg mg -1 day -1 for <strong>the</strong> transplantedseedlings. In <strong>the</strong> Salicornia virginica zone, Spartina survival was 94% and RGR averaged 13 mgmg -1 day -1 with reduced competition. Spartina seedling survival was 90% and mean RGR was 11 mgmg -1 day -1 when grown in full competition with Salicornia. In <strong>the</strong> Distichlis zone, Spartina survivalwas 57% and <strong>the</strong> RGR was 5 mg mg -1 day -1 without competition. Spartina seedling survivorship waszero when grown in competition with Distichlis. In general, Spartina survival and RGR were highestat lower intertidal elevations and with reduced competition. The spread <strong>of</strong> S. anglica into upperintertidal levels appears limited by competition as well as physical conditions along <strong>the</strong> intertidalgradient.Keywords: competition, Distichlis spicata, Salicornia virginica, Spartina anglica, zonationINTRODUCTIONVegetation zonation in salt marshes is determined byboth competitive (Bertness 1991a; Pennings and Callaway1992) and facilitative (Bertness and Shumway 1993,Bertness and Hacker 1994; Callaway and Pennings 2000)biotic interactions as well as physical conditions includingsoil waterlogging, oxygen availability (Howes et al. 1981),sulfide toxicity (King et al. 1982) and porewater salinity(Callaway et al. 1990, Pennings and Callaway 1992). Innorthwestern Washington, salt marsh communities <strong>of</strong>tenform a fringe <strong>of</strong> vegetation along <strong>the</strong> upper edges <strong>of</strong>mudflats. These sites have high porewater salinities (>30grams per kilogram [g kg -1 ) and are typically isolated fromriverine freshwater. The lowest intertidal zones in <strong>the</strong>semarshes are typically colonized by mats <strong>of</strong> Salicorniavirginica with intermittent patches <strong>of</strong> Spergulariacanadensis, Triglochin maritimum, and Puccinelliamaritima. Frequently, individuals <strong>of</strong> <strong>the</strong>se species willcolonize <strong>the</strong> open mudflat several meters <strong>of</strong>f <strong>the</strong> seawardedge <strong>of</strong> contiguous salt marsh.The introduction <strong>of</strong> Spartina anglica into northwesternWashington has altered <strong>the</strong> zonation <strong>of</strong> salt marsh vegetationby creating a community <strong>of</strong> emergent vegetation below <strong>the</strong>Salicornia virginica zone on sediment that was previouslyopen mudflat. Throughout nor<strong>the</strong>rn Puget Sound, Spartinacan form dense stands in areas that were formerly open mudflats (e.g. Triangle Cove and Livingston Bay, IslandCounty). However, Spartina is only found sporadicallygrowing among native plants in higher intertidal zones (e.g.at Maylor Marsh and English Boom, Island County).Based on <strong>the</strong> tendency <strong>of</strong> Spartina to colonize openmudflats as opposed to closed mats <strong>of</strong> native vegetation, wehypo<strong>the</strong>sized that Spartina is an inferior competitor and thatsurvival and growth rates would be highest in <strong>the</strong> lowestintertidal zones and in experimental plots where competitionfrom native plants was limited.MATERIALS AND METHODSWe conducted a seedling transplant experiment in fourlocations at Alice Bay (48º 33’ 20” N, 122º 29’11” W;Skagit County). A series <strong>of</strong> 10 random plots were placedalong 30-meter (m) transects in <strong>the</strong> mudflat. We alsomonitored <strong>the</strong> growth <strong>of</strong> 10 naturally occurring,unmanipulated (in situ) seedlings in <strong>the</strong> mudflat to controlfor <strong>the</strong> effects <strong>of</strong> transplanting. Ten paired plots were-99-

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaRGR (mg mg -1 day -1 )20181614121086420accbFull competitionReduced competitionNo competitionbdIN SITU MUD SALIC SALIC DIST DISTFig. 1. Percent survival (mean±SE; n=10 plots) <strong>of</strong> Spartina seedlingsplanted within each intertidal zone. IN SITU: unmanipulated mudflatseedlings; MUD: transplanted mudflat seedlings; SALIC: Salicornia zone;DIST: Distichlis zone. Different letters indicate significant differences(p < 0.05) between treatment groups using Tukey’s multiple comparisonstest. Due to <strong>the</strong> lack <strong>of</strong> variation in <strong>the</strong> IN SITU and <strong>the</strong> Distichlis fullcompetition plots, <strong>the</strong>se treatments could not be compared using ANOVA.Instead, comparisons <strong>of</strong> <strong>the</strong>se treatments were made using 95% confidenceintervals.INCREASING RELATIVE INTERTIDAL HEIGHTFig. 2. Spartina seedling RGR (mean±SE; n=10 plots) across intertidalzones (abbreviations as in Figure 1). Different letters indicate significantdifferences (p < 0.05) between treatment groups using Tukey’s multiplecomparisons test. Due to a lack <strong>of</strong> variance, <strong>the</strong> Distichlis full competitiontreatment was omitted from <strong>the</strong> ANOVA and comparisons were made with95% confidence intervals.established in <strong>the</strong> Salicornia virginica-dominated salt marshzone immediately adjacent to <strong>the</strong> mudflat. Ten paired plotsalso were placed in <strong>the</strong> Distichlis spicata-dominated zonelocated immediately above <strong>the</strong> Salicornia zone. Plots wererandomly placed along 30 m transects in near monocultures<strong>of</strong> Salicornia and Distichlis. In <strong>the</strong> Salicornia and Distichliszones each plot consisted <strong>of</strong> a subplot with native vegetationintact (full competition) and native vegation removed(reduced competition). In June 2001, three seedlings weretransplanted into each plot or subplot within each zone.Seedlings were watered immediately after being transplantedand were tagged for identification. Reduced competitionplots were cleared <strong>of</strong> all native vegetation (ca. 0.25 mdiameter) and maintained throughout <strong>the</strong> study.In late September-early October 2002, each Spartinaplant was harvested from <strong>the</strong> experimental plots. Plants weretransported to <strong>the</strong> lab, washed repeatedly, dried for 72 hoursat 70ºC, and weighed. Biomass within each subplot wasaveraged for <strong>the</strong> three replicate seedlings in each subplot.Relative growth rates (RGR) <strong>of</strong> Spartina were calculated forall competition treatments at each intertidal level (Eqn. 1;Hunt 1978)RGR = lnW 2 - lnW 1 / T 2 -T 1 Eqn. 1where W 1 and W 2 are <strong>the</strong> plant weights at times 1 (T 1 ) and 2(T 2 ). Initial weight (W 1 ) was estimated using a regression <strong>of</strong>tiller number in relation to dried seedling biomass (Hellquist2005).Survival data were transformed using <strong>the</strong> arcsin squareroot transformation. Survival and RGR data were analyzedusing analysis <strong>of</strong> variance in <strong>the</strong> mixed model procedure <strong>of</strong>SAS 9.1 (SAS Institute 2000). Mixed models are appropriatewhen random factors (split plots) are present in anexperimental design (SAS Institute 2000). The Satterthwaiteadjustment was applied to compensate for heterogeneity <strong>of</strong>variances. Unless o<strong>the</strong>rwise noted (see Figs. 1 and 2),Tukey’s pairwise comparisons were made (p < 0.05) acrosstreatments.RESULTSAll seedlings survived over <strong>the</strong> course <strong>of</strong> <strong>the</strong> experimentwithin <strong>the</strong> unmanipulated, in situ mudflat treatment.Seedlings transplanted within <strong>the</strong> mudflat had a 40%decrease in survival (Fig.1). Percent survival was high (>90%) and not different between <strong>the</strong> full and reducedcompetition plots in <strong>the</strong> Salicornia zone. However, allseedlings died with full competition in <strong>the</strong> Distichlis zone.Fifty percent <strong>of</strong> <strong>the</strong> seedlings survived in reducedcompetition subplots within <strong>the</strong> Distichlis zone. Survival in<strong>the</strong> Salicornia zone was higher than in both Distichlistreatments (Fig. 1).The RGR <strong>of</strong> Spartina anglica generally decreasedmoving from <strong>the</strong> low Salicornia-dominated zone to <strong>the</strong>higher Distichlis-dominated zone (Fig. 2). Theunmanipulated, in situ plants had <strong>the</strong> greatest RGR. Themudflat in situ seedlings had a growth rate over two timesgreater than <strong>the</strong> seedlings transplanted within <strong>the</strong> mudflat.In <strong>the</strong> Salicornia zone <strong>the</strong>re were no differences in RGRbetween full competition and reduced competition subplots.Distichlis zone seedlings grew slower than within <strong>the</strong> in situand Salicornia treatments. The lowest RGRs were in <strong>the</strong>Distichlis zone and in <strong>the</strong> transplanted mudflat seedlings(Fig. 2).- 100 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadDISCUSSIONThe high RGR <strong>of</strong> seedlings on mudflats corresponds tosimilar findings by Dethier and Hacker (2005).Transplanting seedlings clearly reduced survivorship andRGR <strong>of</strong> Spartina seedlings on <strong>the</strong> mudflat where <strong>the</strong>re wasno competition from native vegetation (Fig. 1). Dislodging<strong>the</strong> roots during transplantation in <strong>the</strong> mudflat madeseedlings vulnerable to uprooting via tidal disturbance. Tidaluprooting can be an important source <strong>of</strong> S. anglica seedlingmortality in open mudflats (Groenendijk 1986). In this studyonly transplanted seedlings in <strong>the</strong> mudflat experiencedmortality that could be attributed to tidal disturbance.Seedling survival did not differ between competitiontreatments in <strong>the</strong> Salicornia zone, but was greatly reducedby full competition in <strong>the</strong> Distichlis zone. Reduced survivalin <strong>the</strong> Distichlis zone appeared to be due to competition withnative vegetation as well as rodent herbivory. The apparentasymmetric affect <strong>of</strong> Salicornia and Distichlis on Spartinaseedlings may be <strong>the</strong> result <strong>of</strong> interacting substrateconditions in <strong>the</strong> lower intertidal zones that are favorable toSpartina growth.Although Spartina survival in <strong>the</strong> Salicornia treatmentswas not different, Spartina tiller production and totalbiomass were reduced in <strong>the</strong> presence <strong>of</strong> Salicornia(Hellquist 2005). Decreased survival and RGR by Spartinaanglica when grown in competition with native vegetation(e.g. Distichlis) is consistent with previous studies <strong>of</strong>Spartina. In greenhouse and field studies, Puccinelliamaritima has been shown to have a competitive effect on S.anglica (Scholten and Rozema 1990; Thompson et al. 1993;Huckle et al. 2000). Competitive suppression by neighboringvegetation also has been documented for S. alterniflora in<strong>the</strong> nor<strong>the</strong>astern United States (Bertness 1991b) and S.maritima in Spain (Castellanos et al. 1994)Spartina anglica colonizes a variety <strong>of</strong> habitatsincluding mudflats, low and high salinity salt marshes, aswell as cobble and gravel beaches in Puget Sound (Hacker etal. 2001). At Alice Bay, a mudflat with abundant Spartinaadjoins a protected salt marsh that is truncated by a dike.Spartina is a sporadic colonizer among Salicornia, but wasabsent from <strong>the</strong> virtual monospecific stands <strong>of</strong> Distichlisspicata higher along <strong>the</strong> intertidal gradient. Our experimentshows that at Alice Bay, in <strong>the</strong> absence <strong>of</strong> competitors,Spartina is capable <strong>of</strong> surviving along <strong>the</strong> entire salt marshgradient. In high salinity salt marshes Spartina may berestricted to lower intertidal zones such as <strong>the</strong> Salicorniazone due to interspecific competition and a lack <strong>of</strong> availableopen substrate (Hacker et al. 2001). Although not testedhere, propagule pressure also can play a role in speciesinvasions (e.g. Von Holle and Simberl<strong>of</strong>f 2005) and mayinfluence <strong>the</strong> distribution <strong>of</strong> Spartina across intertidal zones.Although <strong>the</strong>se data and additional growth parameterdata (Hellquist 2005) indicate that biotic factors play a rolein <strong>the</strong> distribution <strong>of</strong> Spartina in high salinity marshes(defined as > 29 g kg -1 by Hacker et al. 2001) <strong>the</strong> importance<strong>of</strong> abiotic factors such as tidal uprooting (Groenendijk 1986)and soil physical characteristics also contribute to Spartinacolonization (Hacker et al. 2001; Dethier and Hacker 2005;Hellquist 2005). Physical factors have been shown to beimportant in controlling seed germination <strong>of</strong> S. anglica inPuget Sound (Dethier and Hacker 2005). Physical factorsmay play a more important role during seed germination andinitial establishment (Dethier and Hacker 2005), but bioticinteractions may become more important during seedlingmaturation (Hellquist 2005)An understanding <strong>of</strong> <strong>the</strong> physical and biotic factorsresponsible for plant zonation in salt marshes is necessary tobetter identify anthropogenic pressures that threaten saltmarshes and to establish effective restoration programs(Bertness and Pennings 2000). These data as well as Dethierand Hacker (2005) and Hellquist (2005), suggest thatremoval <strong>of</strong> Spartina in <strong>the</strong> mudflats will be an effective wayto slow <strong>the</strong> spread <strong>of</strong> Spartina by controlling fast-growingpopulations that can serve as seed sources. Successfulcontrol in high salinity marshes is probably aided by slowerinvasion rates mediated by biotic and physical factors(Hacker et al. 2001; Hellquist 2005). Knowledge <strong>of</strong> Spartinagrowth patterns, <strong>the</strong> influence <strong>of</strong> physical factors, and <strong>the</strong>importance <strong>of</strong> biotic interactions will provide a valuablecontext to prioritize sites for Spartina control (Dethier andHacker 2005; Hellquist 2005).ACKNOWLEDGMENTSWe are very grateful for <strong>the</strong> support <strong>of</strong> <strong>the</strong> GraduateStudent Research Fellowship (NA07OR0262) from <strong>the</strong>Estuarine Reserves Division, Office <strong>of</strong> Ocean and CoastalResource Management, National Ocean Service, NationalOceanic and Atmospheric Administration, Padilla BayNational Estuarine Research Reserve (NERR), Mt. Vernon,Washington to CEH. Terry Stevens, Doug Bulthuis, andSharon Riggs (Padilla Bay NERR) provided essentialsupport for this research. Additional support was providedby <strong>the</strong> United States Environmental Protection AgencyScience to Achieve Results (EPA STAR) Graduate StudentFellowship (U-91616801: 2003-05). The Betty W.Higinbotham Trust <strong>of</strong> <strong>the</strong> Washington State UniversitySchool <strong>of</strong> Biological Sciences provided critical researchfunding as well. Richard Alldredge and Paul Rabie providedstatistical advice. Denise Howe, Sven Nelson, Ben Rhodesand Justin Snider assisted with sample processing. Thecooperation <strong>of</strong> Kyle Murphy <strong>of</strong> <strong>the</strong> Washington StateDepartment <strong>of</strong> Agriculture and Bill Rogers <strong>of</strong> <strong>the</strong> SkagitCounty Noxious Weed Control Board also are appreciated.Thanks also to <strong>the</strong> reviewers for <strong>the</strong>ir suggestions to improve<strong>the</strong> manuscript.- 101 -

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaREFERENCESBertness, M.D. 1991a. Interspecific interactions among high marshperennials in a New England saltmarsh. Ecology 72:125-137.Bertness, M.D. 1991b. Zonation <strong>of</strong> Spartina patens and Spartinaalterniflora in a New England salt marsh. Ecology 72:138-148.Bertness, M.D., and S.D. Hacker. 1994. Physical stress and positiveassociations among marsh plants. Am Nat 144:363-372.Bertness, M.D., and S.C. Pennings. 2000. Spatial variation in processand pattern in salt marsh plant communities in eastern NorthAmerica. In: Weinstein, M.P. and D.A. Kreeger, eds. Conceptsand controversies in tidal marsh ecology, 39-57. Kluwer Academic,Dordrecht, <strong>the</strong> Ne<strong>the</strong>rlands.Bertness, M.D., and S.W. Shumway. 1993. Competition and facilitationin marsh plants. Am Nat 142:718-724.Callaway, R.M., S. Jones, W.R. Ferren, and A. Parikh. 1990. Ecology<strong>of</strong> a mediterranean-climate estuarine wetland at Carpinteria,California: plant distributions and soil salinity in <strong>the</strong> uppermarsh. Can J Bot 68:1139-1146.Callaway, R.M., and S.C. Pennings. 2000. Facilitation may buffercompetitive effects: indirect and diffuse interactions among saltmarsh plants. Am Nat 156:416-424.Castellanos, E.M., M.E. Figueroa, and A.J. Davy. 1994. Nucleationand facilitation in saltmarsh succession: interactions betweenSpartina maritima and Arthrocnemum perenne. J Ecol 82:239-248.Dethier, M.N. and S.D. Hacker. 2005. Physical factors vs. bioticresistance in controlling <strong>the</strong> invasion <strong>of</strong> an estuarine marsh grass.Ecol Appl 15: 1273-1283.Groenendijk, A.M. 1986. Establishment <strong>of</strong> a Spartina anglicapopulation on a tidal mudflat: a field experiment. J EnvironManage 22:1-12.Hacker, S.D., D. Heimer, C.E. Hellquist, T.G. Reeder, B. Reeves,T.J. Riordan, and M.N. Dethier. 2001. A marine plant (Spartinaanglica) invades widely varying habitats: potential mechanisms<strong>of</strong> invasion and control. Biol Invasions 3:211-217.Hellquist, C.E. 2005. Community and trophic consequences <strong>of</strong> <strong>the</strong>introduction <strong>of</strong> Spartina anglica C.E. Hubb. (Poaceae: EnglishCordgrass) in estuaries <strong>of</strong> Puget Sound, WA, USA. Ph.D. Dissertation.Washington State University, Pullman, WA.Howes, B.L., R.W. Howarth, J.M. Teal, and I. Valiela. 1981. Oxidation-reductionpotentials in a salt marsh: spatial patterns andinteractions with primary production. Limnol Oceanogr 26:350-360.Huckle, J.M., J.A. Potter, and R.H. Marrs. 2000. Influence <strong>of</strong> environmentalfactors on <strong>the</strong> growth and interactions between saltmarsh plants: effects <strong>of</strong> salinity, sediment and waterlogging. JEcol 88:492-505.Hunt, R. 1978. Plant growth analysis. Edward Arnold, London,UK.King, G.M., M.J. Klug, R.G. Wiegert, and A.G. Chalmers. 1982.Relation <strong>of</strong> soil water movement and sulfide concentration toSpartina alterniflora production in a Georgia salt marsh. Science218:61-63.Pennings, S., and R. Callaway. 1992. Salt marsh zonation: <strong>the</strong> relativeimportance <strong>of</strong> competition and physical factors. Ecology73:681-690.SAS Institute. 2000. SAS Online Doc Version 8. SAS Institute Inc.:Cary, NC. http://v8doc.sas.com/sashtml/Scholten, M., and J. Rozema. 1990. The competitive ability <strong>of</strong>Spartina anglica on Dutch salt marshes. In: Gray, A.J. andP.E.M. Benham, eds. Spartina anglica – a research review, 39-47. HMSO, London.Thompson, J.D., T. McNeilly, and A.J. Gray. 1993. The demography<strong>of</strong> clonal plants in relation to successional habitat change:<strong>the</strong> case <strong>of</strong> Spartina anglica. In: Miles, J. and D. W. H. Walton,eds. Primary succession on land. Blackwell Scientific Publications,Oxford.Von Holle, B., and D. Simberl<strong>of</strong>f. 2005. Ecological resistance tobiological invasion overwhelmed by propagule pressure. Ecology86:3212-3218.- 102 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadWILL SPARTINA ANGLICA INVADE NORTHWARDS WITH CHANGING CLIMATE?A.J. GRAY 1,2 AND R.J. MOGG 11 Centre for Ecology and Hydrology, CEH Dorset, Winfrith Technology Centre, Dorchester, Dorset, DT2 8ZD, UK2 email: ajg@ceh.ac.ukSpartina anglica’s successful invasion has depended on its ability to occupy mudflats atlower elevations than existing saltmarsh vegetation but has slowed, with dieback in <strong>the</strong>south and successional replacement in <strong>the</strong> north. The elevational niche <strong>of</strong> S. anglica in <strong>the</strong>UK was shown to extend below that <strong>of</strong> its main competitor Puccinellia maritima by 68centimeters (cms). The niche <strong>of</strong> <strong>the</strong> two species overlapped by 20 cms and within thiszone <strong>the</strong>ir distribution depends on <strong>the</strong> outcome <strong>of</strong> <strong>the</strong>ir competitive interaction. Thisinteraction was investigated in <strong>the</strong> light <strong>of</strong> projected climate changes and <strong>the</strong> fact that <strong>the</strong>two species utilize different photosyn<strong>the</strong>tic pathways. A competition experiment underelevated levels <strong>of</strong> temperature and carbon dioxide is described and its implications for <strong>the</strong>future development <strong>of</strong> Spartina-dominated marshes is discussed.Keywords: Spartina anglica, elevational niche, Puccinellia maritima, climate change,competition, C3 and C4 grassesTHE ORIGIN AND SPREAD OF S. ANGLICASpartina anglica CE Hubbard is arguably <strong>the</strong> bestknown example worldwide <strong>of</strong> an invasive allopolyploidspecies. It originated on <strong>the</strong> south coast <strong>of</strong> England sometime between 1870, when <strong>the</strong> sterile F1 hybrid S. townsendiiwas first noticed, and 1892 when <strong>the</strong> first fertile specimenwas collected. The allopolyploid was subsequently named asSpartina anglica and its spread both naturally and bydeliberate planting, has been extremely well documented —Gray et al. (1991) provide a summary. The parental species,S. maritima, a native <strong>of</strong> <strong>the</strong> old world, and S. alterniflora,accidentally introduced from North America intoSouthampton Water, probably in ships’ ballast, were incontact for a relatively short period (30-40 years?).Never<strong>the</strong>less <strong>the</strong> patterns <strong>of</strong> isoenzyme variation haveunambiguously established <strong>the</strong> hybrid origin <strong>of</strong> S. anglica(Raybould et al. 1991a), and analysis <strong>of</strong> <strong>the</strong> DNA sequence<strong>of</strong> <strong>the</strong> chloroplast leucine tRNA gene intron has shown thatS. alterniflora was <strong>the</strong> female parent in <strong>the</strong> original cross(cpDNA is maternally inherited in most grass species)(Ferris et al. 1997). This is not surprising in view <strong>of</strong> <strong>the</strong>relative frequency <strong>of</strong> <strong>the</strong> two species in <strong>the</strong> area <strong>of</strong>hybridization during <strong>the</strong> middle <strong>of</strong> <strong>the</strong> last century.Following hybridization and chromosome doubling(nei<strong>the</strong>r <strong>of</strong> which events have been repeated experimentally)<strong>the</strong> hybrids and <strong>the</strong>ir parents appear to have been quiterapidly isolated both genetically and ecologically. Spartinamaritima is found in <strong>the</strong> mid-level and high marsh zones <strong>of</strong>well established salt marshes and is today largely confined to<strong>the</strong> east coast <strong>of</strong> England in <strong>the</strong> counties <strong>of</strong> Essex andSuffolk (Raybould et al. 1991b). The nearest extantpopulation to <strong>the</strong> site <strong>of</strong> origin is on <strong>the</strong> Isle <strong>of</strong> Wight, about10 miles away. Spartina alterniflora has been reduced to asingle clonal population in Southampton Water and israpidly being lost as <strong>the</strong> lower zones <strong>of</strong> <strong>the</strong> salt marshes areeroded along this coast. (There are a small number <strong>of</strong> o<strong>the</strong>rclones <strong>of</strong> <strong>the</strong> species elsewhere but <strong>the</strong>se have beendeliberately introduced from known sources since <strong>the</strong> date <strong>of</strong><strong>the</strong> original hybridization.) Thus, somewhat ironically inview <strong>of</strong> <strong>the</strong> main <strong>the</strong>me <strong>of</strong> this conference, S. alterniflora in<strong>the</strong> UK is regarded as a rare and threatened species <strong>of</strong>conservation interest, and its presence was recentlyinstrumental in <strong>the</strong> refusal for permission to extend anddevelop <strong>the</strong> nearby port. The sterile hybrid S. townsendii isalso largely confined to <strong>the</strong> original hybridization site,although its distribution is uncertain because it wasintroduced to many estuaries along with <strong>the</strong> fertile form.By contrast S. anglica has colonized, or been planted in,almost all English and Welsh estuaries, is found in Scotland,and occurs in suitable habitats around <strong>the</strong> coast <strong>of</strong> Europefrom 48 to 57.5° N. It has famously been introduced to manyo<strong>the</strong>r parts <strong>of</strong> <strong>the</strong> world and in several areas has become aserious invader <strong>of</strong> native biotopes. As detailed below, S.anglica mostly occurs in <strong>the</strong> lowest zones <strong>of</strong> salt marshes,and at its peak occupied almost 25% <strong>of</strong> <strong>the</strong> total saltmarsharea in Britain (10,000 hectares (ha) <strong>of</strong> <strong>the</strong> 44,000 hasurveyed by Charman (1990)).The initial rapid genetic isolation, and current ecologicalisolation, <strong>of</strong> <strong>the</strong> species involved in <strong>the</strong> evolution <strong>of</strong> S.anglica makes a fascinating contrast to <strong>the</strong> situation in SanFrancisco Bay, where hybridization between <strong>the</strong> native plantS. foliosa, and its ecological replacement, are a feature <strong>of</strong> <strong>the</strong>invasion <strong>of</strong> <strong>the</strong> non-native S. alterniflora.- 103 -

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaTHE CURRENT STATUS OF S. ANGLICA IN EUROPEInvasions <strong>of</strong> European mudflats by S. anglica (hereafterrefered to simply as ‘Spartina’) has been characterised by adifferent sequence <strong>of</strong> events at different latitudes. In <strong>the</strong>south <strong>of</strong> England, and from <strong>the</strong> southwest Ne<strong>the</strong>rlandssouthwards, large monospecific stands developed rapidlyfollowing initial introduction and have in many places beenfollowed in <strong>the</strong> early years <strong>of</strong> <strong>the</strong> last century by ‘dieback,’<strong>the</strong> swards breaking up and retreating in area. By contrast innor<strong>the</strong>rn England north <strong>of</strong> about 54° N and in <strong>the</strong> Dutch andGerman Waddensea, invading Spartina is replacedsuccessionally by o<strong>the</strong>r species, most commonly <strong>the</strong> grassPuccinellia maritima (hereafter ‘Puccinellia’). For examplein Morecambe Bay a 100-ha marsh formerly dominated bySpartina is now a Puccinellia-dominated sward from whichSpartina has almost disappeared (Gray and Raybould 1997).Although confounded in some cases by latitudinal variationin sediment type (<strong>the</strong> nor<strong>the</strong>rn marshes tend to be moresandy) this contrast between nor<strong>the</strong>rn and sou<strong>the</strong>rn marshesmay be related to <strong>the</strong> effects <strong>of</strong> differences in climate anddaylength on <strong>the</strong> two species (see below).Although <strong>the</strong> Spartina invasion has not completelyhalted it has certainly slowed, and with dieback widespreadin <strong>the</strong> south and a slow rate <strong>of</strong> colonization in <strong>the</strong> north,<strong>the</strong>re is probably a net loss <strong>of</strong> area <strong>of</strong> Spartina marsh at <strong>the</strong>present day. Small isolated foci <strong>of</strong> recolonization may befound in <strong>the</strong> south where, in response to rising relative sealevels, sea defences have been removed to create newmudflats and salt marsh, a process termed ‘managedrealignment’. Active invasion by Spartina <strong>of</strong> such newlycreated intertidal mudflats can be observed at Tollesbury inEssex, a county which has lost more than 25% <strong>of</strong> its saltmarsh in <strong>the</strong> last thirty years. For <strong>the</strong> moment it seems <strong>the</strong>Spartina invasion is ‘on hold.’INTERACTIONS BETWEEN SPARTINA AND PUCCINELLIAThe most parsimonious explanation for Spartina’ssuccess as an invader is that it has been able to grow onintertidal mudflats at lower elevations than <strong>the</strong> existingperennial saltmarsh vegetation. This is undoubtedly due to arange <strong>of</strong> adaptive morphological and physiological features<strong>the</strong> net effect <strong>of</strong> which enable Spartina to withstand higherfrequencies <strong>of</strong> tidal submergence. The ‘elevational niche’ <strong>of</strong>Spartina in Britain has been measured by surveyingtransects across salt marshes and recording <strong>the</strong> highest andlowest levels <strong>of</strong> <strong>the</strong> sward (in meters above <strong>the</strong> standard UKdatum height) and relating this to tidal constants (Gray et al.1991, 1995). Comparison <strong>of</strong> this niche with that <strong>of</strong> o<strong>the</strong>rsaltmarsh species confirms that Spartina extends on average68 centimeters (cm) below Puccinellia, <strong>the</strong> species with <strong>the</strong>next lowest elevational limit. There is also a niche overlapbetween <strong>the</strong> two species by 20 cm. Within this overlap zone<strong>the</strong> distribution <strong>of</strong> <strong>the</strong> two species is likely to be principallydetermined by interspecific competition, and Scholten andRozema (1990) provide clear evidence <strong>of</strong> this using anelegant removal experiment. The outcome <strong>of</strong> interspecificcompetition was shown to be critically dependent onvariation in local marsh elevation.Measurement <strong>of</strong> Spartina’s elevational niche alsoindicated that its upper limit varies with latitude. Theregression equation describing this limit (Gray et al. 1991)was:Upper Limit = 4.74 + 0.483(R) + 0.068(F) – 0.099(L)where R = spring tide range (m), F = fetch in <strong>the</strong> direction <strong>of</strong><strong>the</strong> transect (kilometers) and L = latitude (decimal degreesN). This equation, which accounted for 90.2% <strong>of</strong> <strong>the</strong>variation in <strong>the</strong> species’ upper limit, shows that Spartinadoes not extend so far upshore in more nor<strong>the</strong>rly latitudes. Apossible explanation for this is that <strong>the</strong> competitiveinteraction between Spartina and Puccinellia increasinglyfavours <strong>the</strong> latter as one goes northwards, enabling it toinvade <strong>the</strong> Spartina zone at lower elevations.The increasing competitive advantage <strong>of</strong> Puccinellia athigher latitudes may be linked to <strong>the</strong> different effects <strong>of</strong>temperature on Spartina and Puccinellia, which are revealedby differences in <strong>the</strong>ir seasonal growth patterns. Fieldmeasurements on a salt marsh at 52° N showed thatPuccinellia shoot weight increased in March when airtemperatures rose above 5° C, with growth peaking in Juneand July, whereas Spartina did not begin to grow until May,when temperatures reached 9° C, and reached maximumgrowth in October (Dunn et al. 1981). This laterdevelopment and growth <strong>of</strong> Spartina can be related to <strong>the</strong>species’ utilization <strong>of</strong> <strong>the</strong> C4 photosyn<strong>the</strong>tic pathway. One<strong>of</strong> only eight C4 species in <strong>the</strong> UK flora, Spartina is onlypartially adapted to cooler climates (C4 photosyn<strong>the</strong>sis, inwhich <strong>the</strong> first product <strong>of</strong> carbon dioxide fixation isoxaloacetate instead <strong>of</strong> phosphoglycerate as in C3 species, ismost common in semi-arid tropical and subtropical regions)(Long 1983,1990).THE EFFECTS OF CLIMATE CHANGEThe studies outlined above suggest that <strong>the</strong> northwardinvasion <strong>of</strong> Spartina is being prevented, or slowed, by <strong>the</strong>species inability to grow at low temperatures, and that, at itsnor<strong>the</strong>rn limits it is replaced by Puccinellia (a C3 specieswith a circumpolar distribution from 70° N southwards).These two pioneer saltmarsh species bioengineer vast tracts<strong>of</strong> intertidal mudflats and must be regarded as keystonespecies in saltmarsh development. The obvious andintriguing question is: will increases in air temperaturespredicted under various climate change scenarios enableSpartina to invade northwards into marshes currentlydominated by Puccinellia? Higher temperatures will bothincrease growth rate and extend <strong>the</strong> period <strong>of</strong> growth during<strong>the</strong> early months <strong>of</strong> <strong>the</strong> year. Spartina is unlikely to beaffected by increased atmospheric carbon dioxide as, where- 104 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and Spreadwater supply is not limiting, photosyn<strong>the</strong>sis in C4 plants iscarbon dioxide saturated. However higher atmosphericcarbon dioxide is likely to increase photosyn<strong>the</strong>sis, andhence growth, in C3 plants such as Puccinellia due to higherconversion efficiency.The different effects <strong>of</strong> elevated temperatures andcarbon dioxide on <strong>the</strong> two species were modelled by Long(1990). The model predicts <strong>the</strong> primary production <strong>of</strong>Spartina and Puccinellia to be expected in <strong>the</strong> year 2050,assuming a 3 degree rise in temperature and a doubling <strong>of</strong>atmospheric carbon dioxide. It was validated by comparing<strong>the</strong> predictions <strong>of</strong> production under 1978 conditions wi<strong>the</strong>mpirical field data. The increase in annual net production <strong>of</strong>Spartina from 1.3 kilograms per square meter (kg/m 2 ) in1978 to 2.1 kg/m 2 in 2050 was largely attributable totemperature-driven increases in leaf area, enabling <strong>the</strong> pointwhere <strong>the</strong> leaf area index is sufficient to intercept 30% <strong>of</strong> <strong>the</strong>incoming solar radiation to be reached 50 days earlier. Theincrease in Puccinellia, from 1.4 kg/m 2 in 1978 to 2.5 kg/m 2in 2050, was largely attributed to higher conversionefficiency in a high carbon dioxide environment but <strong>the</strong>model indicated that Puccinellia would gain from highertemperatures in spring and autumn.Long’s model, as he acknowledges, excludes severalfactors which may change with climate and will impact onplant growth. These include salinity, nutrient availability andwater level. Also excluded is a consideration <strong>of</strong> <strong>the</strong>competitive interactions between <strong>the</strong> two species underchanging conditions.A COMPETITION EXPERIMENT BETWEEN THE TWO SPECIESIn order to gain insight into how changing climate mayaffect <strong>the</strong> distribution <strong>of</strong> Spartina and Puccinellia <strong>the</strong> twospecies were grown in a competition experiment undercontrolled conditions <strong>of</strong> carbon dioxide and temperature.The experiment and full results are described in Gray andMogg (2001) and only a summary is given below.Plants <strong>of</strong> <strong>the</strong> two species were originally sampled froma salt marsh in Morecambe Bay (at 54° 10’ N) and grown incommon conditions for 18 weeks before 160 tillers <strong>of</strong> eachspecies were removed, size-matched and transplanted intopots containing standard soils. The experimental design wasa replacement series (de Wit 1960), each series comprisingfive pots <strong>of</strong> four tillers (4 Spartina, 3 Spartina + 1Puccinellia, 2 Spartina + 2 Puccinellia, 1 Spartina + 3Puccinellia, and 4 Puccinellia). Eight series were set up inlarge pots and eight in smaller pots to give two density levelsand, following a three-week establishment period, <strong>the</strong> potswere transferred to <strong>the</strong> experimental conditions. Thesecomprised eight hemispherical glasshouses (<strong>the</strong>‘solardomes’) in which atmosphere and temperature arecontrolled to a high degree <strong>of</strong> precision. The eight domesallowed two replicates <strong>of</strong> each treatment and <strong>the</strong> fourtreatments were: ambient temperature + ambient CO 2 ;ambient temperature + a CO 2 concentration <strong>of</strong> 340 parts permillion; temperature elevated by 3° C and trackedcontinuously above ambient + ambient CO 2 ; and elevatedtemperature + elevated CO 2 . The pots were maintained withnon-limiting water supplies, fertilized, and harvested after 11months growth when each plant was separated into aboveand below ground material before weighing. The results aresummarized in Table 1.The main conclusions to be drawn from <strong>the</strong> experimentare (1) most yield variables were significantly affected bytreatment, ei<strong>the</strong>r by temperature or carbon dioxide level or<strong>the</strong> interaction between <strong>the</strong>se treatments and density(Spartina height and inflorescence number beingexceptions), (2) <strong>the</strong> main competitive interaction effectswere due to <strong>the</strong> competitive superiority <strong>of</strong> Puccinellia,which had a significant negative effect on Spartina’s heightand above ground weight and displayed strong intraspecificcompetitive effects on tiller number and biomass, (3)Spartina responded to higher temperature as predicted, butalso, at low plant density, to carbon dioxide concentration.In both cases <strong>the</strong> response was mainly an increase in belowgroundgrowth including rhizomes, (4) Puccinelliaresponded to higher atmospheric carbon dioxide and highertemperature mainly by increasing above-ground growth, (5)Spartina had an unexplained poor performance in <strong>the</strong> ++treatment, an effect seen at low density and <strong>the</strong>reforeunlikely to be explained by competition with Puccinellia,Table 1. Individual plant means for six yield variables at final harvest.Trait Species Amb +T +CO 2 ++ Sig mn effectsSpa 4.10 6.17 6.00 3.48 D*, CT*Till no.Pucc 31.46 17.69 25.76 25.98 P**, CTD*Fl. TillHt(cm)S wt(g)BG wt(g)Biom(g)Spa 0.13 0.11 0.28 0.01 nsPucc 1.79 0.42 1.34 1.53 D***, DP*Spa 15.37 16.69 19.98 17.88 P*Pucc 44.47 44.08 48.24 54.98 D***, CD*Spa 0.45 0.70 0.89 0.37 D*, P*Pucc 1.77 1.28 1.61 2.70 D***, CDT**Spa 0.61 0.97 1.38 0.60 CT*, CTD**Pucc 0.42 0.24 0.36 0.37 CTP**Spa 1.04 1.67 1.98 0.97 CT**, CTD**Pucc 2.19 1.53 1.97 3.08 D***, CTP*The treatments (see text) were: Amb = ambient temperature and CO 2 , +T =elevated temperature and ambient CO 2 , +CO 2 = ambient temperature andelevated CO 2 , ++ = elevated temperature and elevated CO 2 . The variableswere: Till no. = number <strong>of</strong> tillers, Fl. Till = number <strong>of</strong> flowering tillers,Ht(cm) = plant height, S wt(g) = shoot weight, BG wt(g) = weight <strong>of</strong>below-ground material, Biom(g) = total biomass. The significance levelsare from a generalised ANOVA <strong>of</strong> log transformed values in which D =effect <strong>of</strong> density (pot size) (1df), C = effect <strong>of</strong> elevated CO 2 (1df), T =effect <strong>of</strong> elevated temperature (1df), and P = effect <strong>of</strong> <strong>the</strong> proportion <strong>of</strong>Puccinellia (which is a measure <strong>of</strong> competitive interaction effects) (3df).Only main interaction effects are given. Significance * p

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaand (6) <strong>the</strong> pervasive effects <strong>of</strong> plant density on mostvariables underlines <strong>the</strong> importance <strong>of</strong> varying density insuch experimental designs (Gibson et al. 1999).A comparison <strong>of</strong> plant growth in ambient conditionswith that in <strong>the</strong> ++ treatment provides broad agreement withLong’s (1990) model predictions. The mean biomass <strong>of</strong>individual Puccinellia plants increased by about 100% inpure stands, a similar order <strong>of</strong> magnitude to <strong>the</strong> 80%increase in cumulative net primary production in <strong>the</strong> model.Increases in Spartina yield <strong>of</strong> 72% and 95% in elevatedtemperature and carbon dioxide respectively comparefavourably with <strong>the</strong> model prediction <strong>of</strong> 62%, but <strong>the</strong> smallincrease in ++ conditions (+5%) was unexpected. Theresponse <strong>of</strong> Puccinellia is broadly in line with that found ino<strong>the</strong>r C3 species (Bazzaz 1990) as is Spartina’s response toelevated temperature. However <strong>the</strong> increased yield <strong>of</strong>Spartina in higher carbon dioxide is at variance with o<strong>the</strong>rwork on C4 grasses, including Spartina patens (Curtis et al.1989).DISCUSSION AND SPECULATIONAs mentioned earlier, plant performance will be affectedby several o<strong>the</strong>r factors related to projected climate change,some <strong>of</strong> which are difficult to predict, and it is generallyrisky to make predictions on <strong>the</strong> basis <strong>of</strong> photosyn<strong>the</strong>ticpathway or CO 2 response alone (Dukes and Mooney 1999).Never<strong>the</strong>less it is interesting to speculate on <strong>the</strong> implications<strong>of</strong> <strong>the</strong> above experiment for <strong>the</strong> future changes in Spartinadistribution. Indeed <strong>the</strong>re are some aspects <strong>of</strong> <strong>the</strong> interactionbetween Spartina and Puccinellia which encourage us tobelieve that our predictions have a better than averagechance <strong>of</strong> being somewhere near <strong>the</strong> mark – a relativelysimple two-species system, predominately vegetativecompetition related to elevation as a resource, a strongseasonal element, wide dispersal <strong>of</strong> propagules, locallyshared resource levels and so on (see Gray and Mogg 2001for a fuller discussion).Locally, competition will be influenced by <strong>the</strong> balancebetween increasing temperatures and carbon dioxideconcentration. The average global warming <strong>of</strong> 1 to 3.5° Cover <strong>the</strong> next century predicted as a result <strong>of</strong> increasedgreenhouse gases is likely to vary spatially and to be higherin nor<strong>the</strong>rn latitudes in winter (Houghton 1996). In generalhowever we may expect Spartina to extend its rangenorthwards <strong>of</strong> 57° N as temperatures and carbon dioxiderise. Seed set in nor<strong>the</strong>rn populations, which is currentlylimited by <strong>the</strong> length <strong>of</strong> <strong>the</strong> growing season (for Spartinathis is <strong>the</strong> number <strong>of</strong> days above 9° C) may also increase andadd to <strong>the</strong> plant’s capacity to colonize new mudfats. Theincreased biomass below ground should enable <strong>the</strong> speciesto survive in <strong>the</strong> lower parts <strong>of</strong> its elevational niche.However, it seems likely that <strong>the</strong> competitively superiorPuccinellia will replace Spartina at appropriate elevations. Itis even possible that <strong>the</strong> high performance <strong>of</strong> Puccinelliaunder elevated temperature and carbon dioxide indicates thatit will replace Spartina at lower elevations in nor<strong>the</strong>rnlatitudes. The future management <strong>of</strong> nor<strong>the</strong>rn marshes,especially <strong>the</strong> extent to which <strong>the</strong>y are grazed (a processwhich favours Puccinellia over Spartina) will influence <strong>the</strong>interaction between <strong>the</strong> two species.In conclusion, our experiment suggests that changingclimatic conditions could kick-start Spartina’s stalledinvasion, enabling it to colonize mudflats and marshesnorthwards <strong>of</strong> its present distribution. Much will depend onchanges in o<strong>the</strong>r ecosystem processes and on o<strong>the</strong>r features<strong>of</strong> <strong>the</strong> changing climate. Higher rainfall, through its effect onsalinity, and changes in storm frequency and wind directionare likely to be important factors. A key factor will be <strong>the</strong>impact <strong>of</strong> rising relative sea levels and <strong>the</strong>ir influence onlocal sediment availability and accretion.ACKNOWLEDGMENTSWe are grateful to Les Jones and <strong>the</strong> staff at CEHBangor who manned <strong>the</strong> solardomes and to Ralph Clarke forstatistical advice. AJG wishes to acknowledge <strong>the</strong> input over<strong>the</strong> years <strong>of</strong> <strong>the</strong> many students who have contributed to <strong>the</strong>Spartina work, in particular Paul Adam, Colin Ferris, AlanRaybould , John Thompson and Liz Warman.REFERENCESBazzaz, F.A. 1990. The response <strong>of</strong> natural ecosystems to <strong>the</strong> risingglobal CO 2 levels. Annual Reviews <strong>of</strong> Ecology and Systematics.21: 167-196.Charman, K. 1990. The current status <strong>of</strong> Spartina anglica in GreatBritain. In: Gray, A.J. and P.E.B. Benham, eds. Spartina anglica– a Research Review, 11-14. HMSO, London.Curtis, P.S., B.G. Drake, and D.F. Whigham 1989. Nitrogen andcarbon dynamics in C3 and C4 estuarine marsh plants grownunder elevated CO 2 in situ. Oecologia. 78:297-301.De Wit, C.T. 1960. On competition. Verslagen van LandbouwkundigeOnderzoekinge., 66:1-82.Dukes, J.S. and H.A. Mooney. 1999. Does global change increase<strong>the</strong> success <strong>of</strong> biological invaders? TREE., 14:135-139.Dunn, R., S.P. Long, and S.M. Thomas. 1981. The effects <strong>of</strong>temperature on <strong>the</strong> growth and photosyn<strong>the</strong>sis <strong>of</strong> <strong>the</strong> C4 grassSpartina townsendii. In: Grace J., E.D. Ford, and P.G. Jarvis,eds. Plants and <strong>the</strong>ir atmospheric environment, 301-311. Blackwell,Oxford.Ferris, C., R.A. King, and A.J. Gray. 1997. Molecular evidence for<strong>the</strong> maternal parentage in <strong>the</strong> hybrid origin <strong>of</strong> Spartina anglicaCE Hubbard. Molecular Ecology. 6:185-187.Gibson, D.J., H. Connolly, D.C. Hartnett, and J.D. Weidenhamer.1999. Designs for greenhouse studies <strong>of</strong> interactions betweenplants. Journal <strong>of</strong> Ecology. 87:1-16.Gray, A.J. and R.J. Mogg. 2001. Climate impacts on pioneersaltmarsh plants. Climate Research. 18: 105-112.Gray, A.J. and A.F. Raybould. 1997. The history and evolution <strong>of</strong>Spartina anglica in <strong>the</strong> British Isles. <strong>Proceedings</strong> 2nd <strong>International</strong>Spartina <strong>Conference</strong>, 13-16. Washington State University,Olympia.Gray, A.J., D.F. Marshall, and A.F. Raybould. 1991. A century <strong>of</strong>evolution in Spartina anglica. Advances in Ecological Research.21:1-61.- 106 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadGray, A.J., A.E. Warman, R.T. Clarke, and P.J Johnson. 1995. Theniche <strong>of</strong> Spartina anglica on a changing coastline. Coastal ZoneTopics. 1:29-34.Houghton, J.T., ed. 1996. Climate change 1995: <strong>the</strong> science <strong>of</strong>climate change: contribution <strong>of</strong> Working Group 1 to <strong>the</strong> SecondAssessment Report <strong>of</strong> <strong>the</strong> Intergovernmental Panel on ClimateChange. Cambridge University Press, Cambridge.Long S.P. 1983. C4 photosyn<strong>the</strong>sis at low temperatures. Plant Celland Environment. 6:345-363.Long S.P. 1990 The primary production <strong>of</strong> Puccinellia maritimaand Spartina anglica : a simple predictive model <strong>of</strong> response toclimate change. In: Beukema, J.J., W.J. Wolff, and J.J.W.M.Brouns, eds. Expected effects <strong>of</strong> climatic change on marinecoastal ecosystems, 33-39. Kluwer, Dordrecht.Raybould, A.F., A.J. Gray, M.J. Lawrence, and D.F. Marshall.1991a. The evolution <strong>of</strong> Spartina anglica CE Hubbard: originand genetic variability. Biological Journal <strong>of</strong> <strong>the</strong> LinneanSociety. 43:111-126.Raybould, A.F., A.J. Gray, M.J. Lawrence, and D.F. Marshall.1991b. The evolution <strong>of</strong> Spartina anglica CE Hubbard: variationand status <strong>of</strong> <strong>the</strong> parental species in Britain. Biological Journal<strong>of</strong> <strong>the</strong> Linnean Society. 44:369-380.Scholten, M.C.J. and J. Rozema. 1990. The competitive ability <strong>of</strong>Spartina anglica on Dutch saltmarshes. In: Gray, A.J. and P.E.B.Benham, eds. Spartina anglica – A Research Review, pp. 39-47.HMSO, London.- 107 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadCOMPETITION AMONG MARSH MACROPHYTES BY MEANS OF VERTICALGEOMORPHOLOGICAL DISPLACEMENTJ.T. MORRISBelle W. Baruch Institute for Marine & Coastal Sciences, University <strong>of</strong> South Carolina, Columbia, SC 29208This paper describes a <strong>the</strong>ory <strong>of</strong> biogeomorphology that addresses how intertidal macrophytes canmodify landscape elevation. Competitive interactions among marsh plant species are mediated by<strong>the</strong> influence <strong>of</strong> vegetation on sediment accretion and <strong>the</strong>ir modification <strong>of</strong> <strong>the</strong> relative elevation <strong>of</strong><strong>the</strong> marsh surface. A model described here demonstrates <strong>the</strong> effects <strong>of</strong> feedback between physicalprocesses like sediment accretion and biological processes such as those that determine specieszonation patterns. Changes in geomorphology, primary productivity and <strong>the</strong> spatial distribution <strong>of</strong>plant species are explained by competitive interactions and by interactions among <strong>the</strong> tides, biomassdensity, and sediment accretion that move marsh elevation towards an equilibrium with mean sealevel (MSL). This equilibrium is affected positively (relative elevation <strong>of</strong> <strong>the</strong> marsh surface increases)by <strong>the</strong> biomass density <strong>of</strong> emergent, salt marsh macrophytes and negatively by <strong>the</strong> rate <strong>of</strong> sea-levelrise. It was demonstrated that a dominant, invading species is able to modify its environment toexclude competitively inferior species. However, <strong>the</strong> outcome depends on a number <strong>of</strong> variablesincluding <strong>the</strong> rate <strong>of</strong> sea-level rise and <strong>the</strong> fundamental biomass distributions <strong>of</strong> species across <strong>the</strong>intertidal gradient. The model predicts that a marsh can move toward alternative states, dependingon <strong>the</strong> rate <strong>of</strong> sea-level rise and species biomass distributions within <strong>the</strong> tidal frame.Keywords: sedimentation, marshes, Spartina, sea level, models, geomorphology, competitionINTRODUCTIONPhysical ecosystem engineering is a process commonto organisms that possess <strong>the</strong> ability to physically modify<strong>the</strong>ir habitats (Jones et al. 1997). This trait is common tointertidal marsh macrophytes that have <strong>the</strong> potential toraise <strong>the</strong> relative elevation <strong>of</strong> <strong>the</strong>ir habitat and modify <strong>the</strong>geomorphology <strong>of</strong> <strong>the</strong> coastal landscape, potentially to <strong>the</strong>benefit and detriment <strong>of</strong> o<strong>the</strong>r species (Zedler & Kercher2004). Historically, coastal wetlands have maintained anelevation in equilibrium with mean sea level by <strong>the</strong> accumulation<strong>of</strong> mineral sediment or organic matter (Redfield1972; Stevenson et al. 1986). Commonly, stable intertidalsalt marshes occupy a broad, flat expanse <strong>of</strong> landscape <strong>of</strong>tenreferred to as <strong>the</strong> marsh platform at an elevation within<strong>the</strong> intertidal zone that approximates that <strong>of</strong> local meanhigh water (MHW) (Krone 1985). Marsh species typicallysegregate along gently sloping gradients across <strong>the</strong> marshplatform (Hacker & Bertness 1999; Silvestri et al. 20005).The elevations <strong>of</strong> <strong>the</strong> marsh platform relative to sea leveldetermine inundation frequency, duration and, consequently,wetland productivity and species distributions.Recent work in a North Inlet, SC marsh has shown that<strong>the</strong> relative elevation <strong>of</strong> <strong>the</strong> sediment surface is a criticallyimportant variable that ultimately controls <strong>the</strong> productivity<strong>of</strong> <strong>the</strong> salt marsh plant community (Morris et al. 2002).Productivity has a positive feedback on <strong>the</strong> rate <strong>of</strong> accretion<strong>of</strong> <strong>the</strong> marsh surface. This feedback is key to predicting <strong>the</strong>responses <strong>of</strong> coastal wetlands to rising sea level, includingchanges to <strong>the</strong> geometry <strong>of</strong> <strong>the</strong> land margin and to <strong>the</strong> totalarea <strong>of</strong> wetland habitat. It is also fundamental to understanding<strong>the</strong> spread <strong>of</strong> invasive marsh macrophytes, such asSpartina hybrids in San Francisco Bay, that have <strong>the</strong> abilityto physically alter <strong>the</strong>ir environment in a manner that benefits<strong>the</strong> invading species (Cuddington & Hastings 2004).The objective <strong>of</strong> <strong>the</strong> study reported here is to explore <strong>the</strong>behavior <strong>of</strong> a model that explicitly treats <strong>the</strong> modification <strong>of</strong>habitat elevation by a hypo<strong>the</strong>tical group <strong>of</strong> marsh macrophytescompeting for habitat space within <strong>the</strong> intertidal zone.The model accounts for <strong>the</strong> species-specific effect <strong>of</strong> marshvegetation on mineral sediment accretion, and for feedbacksamong sea-level rise, relative elevation, species replacement,and primary production. From this work it should be possibleto generalize about <strong>the</strong> types <strong>of</strong> adaptations that enable macrophytespecies to exploit a particular range <strong>of</strong> habitat within<strong>the</strong> intertidal zone and that endow some species with superiorcompetitive abilities. The concept <strong>of</strong> geomorphologicaldisplacement is described whereby one species displacesano<strong>the</strong>r by modifying <strong>the</strong> relative elevation <strong>of</strong> its habitat.MODEL DESCRIPTIONThe model described here is based on fieldwork thatwas carried out in a salt marsh at North Inlet, South Carolina(Morris et al. 2002). The model was initially developedand calibrated for a single species, Spartina alterniflora,which forms a monoculture over <strong>the</strong> majority <strong>of</strong> this marshwithin a narrow range between 0.22 and 0.481 m relative- 109 -

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinato NAVD88 (Morris et al. 2005). At North Inlet, <strong>the</strong> meanhigh water level was 0.618 m (2001 through May, 2003)with a mean tidal range <strong>of</strong> 1.39 m. Details about <strong>the</strong> NorthInlet marsh can be found in a variety <strong>of</strong> sources (Eiser &Kjerfve 1986; Dame et al. 2000; Morris 2000; Morris et al.2005). Below, <strong>the</strong> model is generalized for two or morespecies and its predictions for <strong>the</strong> case <strong>of</strong> two hypo<strong>the</strong>ticalspecies are discussed.The Elevation <strong>of</strong> <strong>the</strong> Marsh SurfaceThe elevation <strong>of</strong> <strong>the</strong> marsh platform equilibrates at aposition where erosion and deposition are equal. It is clearthat sediment deposition must approach zero at elevationsnear that <strong>of</strong> <strong>the</strong> highest high tide and should increase as<strong>the</strong> concentration <strong>of</strong> suspended sediment and hydroperiodincrease. This logic is based upon a consideration<strong>of</strong> abiotic processes alone and predicts that intertidalmarshes approach an equilibrium elevation that approximatesMHW (Krone 1985). A rise in relative sea levelwill increase flooding <strong>of</strong> <strong>the</strong> marsh and inundation time<strong>the</strong>reby increasing <strong>the</strong> opportunity for sediment depositionand re-establishing <strong>the</strong> elevation <strong>of</strong> <strong>the</strong> marsh relativeto <strong>the</strong> new MSL (Pethick 1981, Krone 1985, French 1993).Sedimentation rate is <strong>the</strong> product <strong>of</strong> settling velocity andtime <strong>of</strong> inundation. Since depth below mean higher highwater (MHHW) and time <strong>of</strong> inundation are roughly proportional,<strong>the</strong> net rate <strong>of</strong> change in <strong>the</strong> elevation <strong>of</strong> <strong>the</strong>marsh surface (dY/dt), is proportional to <strong>the</strong> depth (D) <strong>of</strong><strong>the</strong> marsh surface: dY/dt D. Strictly speaking, sedimentationrate should be proportional to depth, whereaso<strong>the</strong>r processes that affect elevation, like compactionand decomposition, may or may not be depth-dependent.However, compaction can be taken to be a constant that isaccounted for in <strong>the</strong> local rate <strong>of</strong> sea-level rise.Feedbacks between <strong>the</strong> marsh vegetation and <strong>the</strong> sedimentsare also important. The rate <strong>of</strong> change <strong>of</strong> elevation <strong>of</strong><strong>the</strong> marsh platform is a positive function <strong>of</strong> <strong>the</strong> standingdensity <strong>of</strong> plant biomass (B). For simplicity it is assumed thatthis relationship is linear: dY/dt q + kB, where parametersq and k are proportional to <strong>the</strong> settling velocity and <strong>the</strong> efficiency<strong>of</strong> <strong>the</strong> vegetation as a sediment trap, respectively. Theproduct <strong>of</strong> kB represents <strong>the</strong> positive effect that abovegroundbiomass density has on <strong>the</strong> trapping <strong>of</strong> suspended sediments(e.g. Leonard & Lu<strong>the</strong>r, 1995; Christiansen et al 2000). Thevalues <strong>of</strong> q and k are likely to vary locally and regionallyas a function <strong>of</strong> sediment availability and tidal range (e.g.Stevenson et al. 1986). Note that k also may vary by species.This implies that <strong>the</strong> efficiency <strong>of</strong> sediment trapping may varyamong species (e.g. Rooth and Stevenson 2000). Combining<strong>the</strong>se concepts, <strong>the</strong> absolute change in elevation <strong>of</strong> <strong>the</strong> marshsurface, for depths (D) within <strong>the</strong> intertidal zone, can beapproximated by:dY/dt = (q + kB)D, for D>0 (1)The kB term implicitly accounts for <strong>the</strong> contribution <strong>of</strong>organic matter accretion when <strong>the</strong> model is calibrated to <strong>the</strong>total accretion rate. It assumes that organic matter accretionis proportional to <strong>the</strong> standing biomass density <strong>of</strong> <strong>the</strong> vegetation.We have recent unpublished results that show that thisassumption overly simplistic, but <strong>the</strong> qualitative behavior<strong>of</strong> <strong>the</strong> model with respect to total sediment accretion andsea-level rise is not changed by breaking out organic matteraccretion as a separate term.The Vertical-Distribution <strong>of</strong> Standing BiomassBiomass density, B (g/m 2 ), is variable and changes witha number <strong>of</strong> environmental conditions including <strong>the</strong> relativeelevation <strong>of</strong> <strong>the</strong> marsh platform. For any intertidal species,<strong>the</strong>re exist upper and lower limits <strong>of</strong> relative elevation. For<strong>the</strong> hybrid in San Francisco Bay as for S. alterniflora on <strong>the</strong>Atlantic coast, <strong>the</strong> lower limit is probably set by <strong>the</strong> hypoxiaresulting from tidal flooding, while <strong>the</strong> upper elevation isdetermined by salt stress, desiccation, and competitivepressure from o<strong>the</strong>r species. The dominant representativesfrom different plant communities along a topographicgradient will have different biomass distributions (a, band c values in Eq. 2) that may or may not overlap, andinterspecific competition or facilitative interactions maymodify <strong>the</strong> shapes <strong>of</strong> <strong>the</strong> curves where overlap occurs (eg.Bertness 1991; Emery et al. 2001; Bertness & Ewanchuk2002; Pennings et al. 2005). These distributions can bedescribed by a family <strong>of</strong> curves:B i= a iD + b iD 2 + c i(2)where a, b, and c are coefficients that determine <strong>the</strong> upperand lower depth limits, and magnitude <strong>of</strong> B and where <strong>the</strong>subscript i refers to a specific dominant species or communitytype. Depth, D (cm), is positive for depths lessthan MHHW. These curves can be viewed as dimensions<strong>of</strong> a species’ fundamental (in <strong>the</strong> absence <strong>of</strong> competitors)or realized (in <strong>the</strong> presence <strong>of</strong> competitors) niche, sensuHutchinson (1957). The values <strong>of</strong> <strong>the</strong> coefficients a, b, and cwill also differ regionally as a function <strong>of</strong> tide range, salinity,or climate. In examples discussed below, <strong>the</strong> values <strong>of</strong>a, b, and c where chosen to represent hypo<strong>the</strong>tical specieswith biomass distributions that span different ranges alongan intertidal gradient.MHHW is used as <strong>the</strong> zero datum (D 0) for conveniencesince it is approximates <strong>the</strong> elevation where biomassdensity and sedimentation rate approach zero. However,because MHHW is based upon an arithmetic average <strong>of</strong><strong>the</strong> higher high water height <strong>of</strong> each tidal day observedover <strong>the</strong> National Tidal Datum Epoch, D 0may differ fromMHHW regionally, depending on factors such as windtides. The departure <strong>of</strong> D 0from MHHW is probably greatestin microtidal estuaries where wind tides <strong>of</strong>ten dominateastronomical tides. Moreover, MHHW will vary due tolow frequency variability in mean sea level and with <strong>the</strong>18.6-yr lunar nodal cycle (Stumpf and Haines, 1998). Thelunar nodal cycle changes <strong>the</strong> tidal amplitude by about 5 cmand is simulated below by allowing D 0to vary by 5 cm at afrequency <strong>of</strong> 1/18.6 yr.- 110 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadFeedbacks Among Biomass Distribution, Relative Elevationand Sea-Level RiseSubstituting for B in Eq. 1 yields <strong>the</strong> following equationthat describes <strong>the</strong> absolute change in elevation <strong>of</strong> <strong>the</strong> marshsurface as a function <strong>of</strong> depth:dY/dt = [q + k i(a iD + b iD 2 + c i)]D, for D>0 (3)Equation 3 was solved numerically for a hypo<strong>the</strong>ticalmarsh landscape with one or two species differing in <strong>the</strong>irbiomass distributions. The time-zero marsh elevation andrates <strong>of</strong> sea-level rise were set arbitrarily and model simulations<strong>of</strong> 120-yr duration were generated to allow time forconvergence on a dynamic steady state. A rate <strong>of</strong> sea-levelrise <strong>of</strong> ei<strong>the</strong>r 0.2 or 0.8 cm/yr was specified, and a sinusoidalfunction was used to simulate changes in tidal amplitude <strong>of</strong>± 2.5 cm over <strong>the</strong> 18.6-yr lunar nodal cycle, centered abouta mean tidal amplitude <strong>of</strong> 0.6 m. The coefficient values forq and k were set at 0.00018 yr -1 and 1.5 × 10 -5 cm m 2 g -1 yr -1(0.15 cm 3 g -1 yr -1 ). These parameter values were taken froma calibration <strong>of</strong> <strong>the</strong> model to North Inlet data (Morris etal. 2002). In addition, species distributions were varied inorder to explore effects on equilibrium elevation and speciespersistence. Species distributions (Eq. 2) were describedby specifying a maximum biomass and optimum elevation,operationally defined here as <strong>the</strong> elevation that results in <strong>the</strong>greatest biomass density, and species range, defined as <strong>the</strong>upper and lower depth limits.RESULTS AND DISCUSSIONThe Single Species ExampleWhen <strong>the</strong> model was solved for a single species witha distribution between 15 and 30 cm above mean sea level(Fig. 1A), and a rate <strong>of</strong> sea-level rise <strong>of</strong> 0.2 cm/yr, <strong>the</strong> marshsurface elevation equilibrated at about 26 cm above mean sealevel and a depth <strong>of</strong> 34 ± 2.9 cm (mean ± amplitude) belowMHHW (Fig 1B). This elevation is above <strong>the</strong> optimum elevationspecified in Fig. 1A, which is a condition for stability(Morris et al. 2002). The constraint on productivity imposedby high pore-water salinities that develop at super-optimalelevations is an important factor in maintaining relativeelevation, because a rise in relative sea level brings about anincrease in flooding, decreases pore water salinity (Morris1995), and increases biomass density (Morris 2000). Theincrease in biomass density will enhance sediment depositionby increasing sediment trapping efficiency (Gleason etal. 1979, Leonard & Lu<strong>the</strong>r 1995, Yang 1998).MHHW varies over <strong>the</strong> course <strong>of</strong> <strong>the</strong> 18.6-yr lunar nodalcycle by ± 2.5 cm. Consequently, <strong>the</strong> depth <strong>of</strong> marsh surfacebelow MHHW varies at <strong>the</strong> same frequency (Fig. 1B), resultingin a cycle <strong>of</strong> standing biomass with a range <strong>of</strong> 296 g/m 2(Fig. 1C). This is consistent with empirical measurementsfrom North Inlet, SC where we have seen changes in S.alterniflora productivity <strong>of</strong> 248 g m -2 yr -1 over <strong>the</strong> lunar nodalcycle. It should be noted that MHHW varies independently<strong>of</strong> MSL owing to long-period astronomical forcing, e.g. <strong>the</strong>lunar nodal cycle, but MHHW will also vary directly withchanges in MSL. Consequently, MHHW is a better tidaldatum for purposes <strong>of</strong> defining <strong>the</strong> growth ranges <strong>of</strong> intertidalspecies than is MSL.Result <strong>of</strong> Adding a Competitor with Wider Distribution andHigh BiomassWhen a competing species was added to <strong>the</strong> model witha distribution like that <strong>of</strong> species 2 in Fig. 2A, species 1was eliminated within about 10 yr (Fig. 2C). Examples arecommon <strong>of</strong> this type <strong>of</strong> interaction, such as <strong>the</strong> invasion <strong>of</strong>Fig. 1 Model results with a single species (S1), distributed as above (A), and a rate <strong>of</strong> sea-level rise <strong>of</strong> 0.2 cm/yr, as shown in B. Panel C shows <strong>the</strong>predicted biomass trajectory as (B) <strong>the</strong> marsh surface equilibrates at a relative surface elevation <strong>of</strong> about 25 cm above mean sea level (MSL) and a depthbelow MHHW <strong>of</strong> about 35 cm.- 111 -

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaFig. 2 Model results with two species with biomass distributions as above (A) and a rate <strong>of</strong> sea-level rise <strong>of</strong> 0.2 cm/yr. Shown in panel C is <strong>the</strong> predicted biomasstrajectory <strong>of</strong> both species as (B) <strong>the</strong> marsh equilibrates at a higher relative surface elevation <strong>of</strong> about 45 cm above mean sea level (MSL) and a depth belowMHHW <strong>of</strong> about 15 cm.Fig.3. Model results with two species with biomass distributions as in <strong>the</strong> previousexample (Fig. 2A) when <strong>the</strong> rate <strong>of</strong> sea-level rise was increased to 0.8 cm/yr; parameter values were o<strong>the</strong>rwise identical to those in <strong>the</strong> previous example(Fig. 2). Shown in panel B is <strong>the</strong> predicted biomass trajectory <strong>of</strong> both species as(A) <strong>the</strong> marsh surface equilibrates at a relative surface elevation <strong>of</strong> about 25 cmabove mean sea level (MSL) and a depth <strong>of</strong> about 35 cm below MHHW.Phragmites australis into brackish Spartina patens marsh(Windham 1999). Note that competition occurs in this modelindirectly by virtue <strong>of</strong> a change in relative elevation, favoringspecies 2. Species responses are defined exclusively by <strong>the</strong>irhabitat distributions (Eq. 2) and not by direct interference.For example, species 2 displaced species 1 when species 2was given a habitat range wider than species 1, a higheroptimum elevation, and a greater biomass density at itsoptimum elevation. In this example, <strong>the</strong> marsh equilibrated ata higher surface elevation, about 45 cm (Fig. 2B), than in <strong>the</strong>preceding case. The equilibrium surface elevation (Fig. 2B)was significantly higher than <strong>the</strong> optimum elevation forspecies 2 (Fig. 2A), which is a consequence <strong>of</strong> <strong>the</strong> relativelyhigh biomass density <strong>of</strong> species 2. Thus, as <strong>the</strong> amplitude<strong>of</strong> a species’ biomass distribution increases, <strong>the</strong> equilibriumelevation <strong>of</strong> <strong>the</strong> marsh surface will increase.This prediction is important for predicting and understandingchanges in salt marshes following <strong>the</strong> introduction<strong>of</strong> an alien species. An invader with a greater biomass and awider habitat range can raise <strong>the</strong> elevation <strong>of</strong> <strong>the</strong> marsh above<strong>the</strong> range <strong>of</strong> <strong>the</strong> native species, although <strong>the</strong>re are severalqualifications that must be noted. Firstly, <strong>the</strong> model discussedhere is zero- dimensional or plot-scale. In two dimensions, <strong>the</strong>geomorphology <strong>of</strong> <strong>the</strong> marsh landscape will adjust to changesin its relative elevation, and it is possible that <strong>the</strong> resultingtopographic gradients will support a dynamically stable zonation<strong>of</strong> species, i.e. marshes transgress and species migrate andsegregate across topographic gradients. Secondly, <strong>the</strong>re areconditions, such as <strong>the</strong> rate <strong>of</strong> sea-level rise, explored below,that lead to <strong>the</strong> persistence <strong>of</strong> a weaker species.Facultative Behavior and <strong>the</strong> Rate <strong>of</strong> Sea-Level RiseA facultative interaction between species is possible byvirtue <strong>of</strong> <strong>the</strong> additive effects on sediment accretion <strong>of</strong> overlappingspecies distributions. The additive effect on accretionrate can maintain <strong>the</strong> elevation <strong>of</strong> <strong>the</strong> marsh surface withinboth species’ ranges at a high rate <strong>of</strong> sea-level rise. This wasdemonstrated by raising <strong>the</strong> rate <strong>of</strong> sea-level rise from 0.2to 0.8 cm/yr (Fig. 3). The increase in rate <strong>of</strong> sea-level riseresulted in stabilization <strong>of</strong> <strong>the</strong> relative surface elevation atabout 28 cm (Fig. 3A) at an elevation that is greater than <strong>the</strong>optimum that had been specified for species 1 (22.5 cm) andsuboptimal for species 2 (32.5 cm, Fig. 2A). At this elevationboth species were able to coexist (Fig. 3B).- 112 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadFig.4. Model results with one species with a biomass distribution as above (A) and a rate <strong>of</strong> sea-level rise <strong>of</strong> 0.8 cm/yr. Shown in panel C is <strong>the</strong> predictedbiomass trajectory <strong>of</strong> species 2 as (B) <strong>the</strong> marsh surface equilibrates at an elevation below MSL and outside <strong>of</strong> species 2’s range. Parameter values wereo<strong>the</strong>rwise identical to those in <strong>the</strong> previous examples (Fig. 2 and 3).The biomass <strong>of</strong> species 2 varied with depth at <strong>the</strong> frequency<strong>of</strong> <strong>the</strong> lunar nodal cycle, but 180º out-<strong>of</strong>-phase, whereas<strong>the</strong> lunar nodal cycle and biomass <strong>of</strong> species 1 were in phase(Fig. 3B). The biomass <strong>of</strong> species 2 decreased with a rise inMHHW, i.e. as equilibrium depth increased, because <strong>the</strong>marsh surface elevation, 28 cm, was suboptimal for growth.Conversely, <strong>the</strong> biomass <strong>of</strong> species 1 increased with a rise inMHHW because surface elevation was super-optimal for itsgrowth. Thus, species can coexist <strong>the</strong>oretically when <strong>the</strong>reis a periodic change in flood duration and when one speciesresponds to <strong>the</strong> change positively and <strong>the</strong> o<strong>the</strong>r negatively.This is one <strong>of</strong> several conditions that promote a facultativeinteraction between species.Facultative interactions among marsh macrophytesinvolving amelioration <strong>of</strong> soil salinity have been described(Bertness & Ewanchuk 2002), and simulation results shownhere suggest that biogeomorphological interactions couldalso be facultative. When two species with overlapping distributions(Fig. 2A) were present, and <strong>the</strong> rate <strong>of</strong> sea-levelrise was raised to 0.8 cm/yr, <strong>the</strong> resulting interaction couldbe described as facultative (Fig. 3), because nei<strong>the</strong>r speciespersists in <strong>the</strong> absence <strong>of</strong> <strong>the</strong> o<strong>the</strong>r (Fig. 4). When species 1was removed from <strong>the</strong> simulation, <strong>the</strong> equilibrium elevationquickly dropped below MSL and below <strong>the</strong> lower limit forspecies 2. Species 1 suffered <strong>the</strong> same fate when species 2was removed. However, at a low rate <strong>of</strong> sea-level rise, 0.2cm/yr, species 1 did not survive (Fig. 2). Thus, <strong>the</strong> outcome<strong>of</strong> competition and <strong>the</strong> emergence <strong>of</strong> facultative behaviordepend on <strong>the</strong> rate <strong>of</strong> sea-level rise.The Rate <strong>of</strong> Sea-Level Rise and Alternative Stable StatesIt can also be demonstrated that <strong>the</strong> marsh will movetoward alternative stable states or habitat preemption by onespecies or ano<strong>the</strong>r, depending on <strong>the</strong> rate <strong>of</strong> sea-level riseand <strong>the</strong> species’ biomass distributions. When <strong>the</strong> rate <strong>of</strong> sealevelrise was 0.8 cm/yr and <strong>the</strong> distribution <strong>of</strong> species 1 wasmodified by raising its maximum biomass to equal that <strong>of</strong>species 2 (Fig. 5A), only species 1 persisted throughout <strong>the</strong>simulation, while species 2 was intermittent (Fig. 5C). Themarsh surface equilibrated at an elevation <strong>of</strong> about 23 cm(Fig. 5B), which was at <strong>the</strong> lower limit <strong>of</strong> species 2 and above<strong>the</strong> optimum elevation for species 1 (Fig. 5A). Keeping <strong>the</strong>species distributions as in Fig. 5A and lowering <strong>the</strong> rate <strong>of</strong>sea-level rise to 0.2 cm/yr resulted in a different outcome. Atthis lower rate <strong>of</strong> sea-level rise, only species 2 persisted (Fig.6C). The marsh surface equilibrated at an elevation <strong>of</strong> about46 cm (Fig. 6A), which is above <strong>the</strong> upper limit <strong>of</strong> species 1and above <strong>the</strong> optimum <strong>of</strong> species 2 (Fig. 5A).CONCLUSIONSThis paper describes a <strong>the</strong>ory <strong>of</strong> biogeomorphology thataddresses how intertidal macrophytes can modify landscapeelevation and affect <strong>the</strong> outcomes <strong>of</strong> species interactionsby means <strong>of</strong> vertical geomorphological displacement.Competition by geomorphogical displacement is indirectand is a function <strong>of</strong> <strong>the</strong> fundamental and realized niches thatdescribe species biomass distributions in <strong>the</strong> tidal frame. Theoutcomes <strong>of</strong> competitive interactions are a great deal morecomplex than described here when <strong>the</strong>re are direct interferencesor facultative interactions <strong>of</strong> <strong>the</strong> sort described byBertness and Ewanchuk (2002). These types <strong>of</strong> interactionscan modify a species’ realized distributions dynamically andwould result in truly complex behavior.Marsh primary production and standing biomass aresensitive to hydroperiod and have positive effects on sedimentaccretion. In that part <strong>of</strong> <strong>the</strong> tidal frame that is higher- 113 -

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaFigure 5. Model results with two species with biomass distributions as above (A) and a rate <strong>of</strong> sea-level rise <strong>of</strong> 0.8 cm/yr (B). Shown in panel C is <strong>the</strong>predicted biomass trajectory <strong>of</strong> both species (species 2 approaches extinction) as <strong>the</strong> marsh surface equilibrates at a relative elevation <strong>of</strong> about 20 cm (B)above mean sea level (MSL).than a species ‘optimum’ elevation, rising sea level increasesprimary production, which stimulates sedimentation andmaintains equilibrium with MSL. Optimum elevation isin quotes because, while that is <strong>the</strong> elevation that favorsmaximum growth, it borders on instability. Several speciesor communities <strong>of</strong> plants may coexist if <strong>the</strong>y partition <strong>the</strong>habitat space, and if a topographic gradient <strong>of</strong> sufficient slopeexists. Alternatively, coexistence is unlikely following aninvasion by a species with greater niche breadth and higherproductivity as in Fig. 2A S2. Locally, sediment accretionFig. 6. Model results with two species with biomass distributions as in <strong>the</strong>previous example (Fig. 5A). Shown in (B) is <strong>the</strong> predicted biomass trajectory<strong>of</strong> both species (s1 became extinct) as <strong>the</strong> marsh surface equilibrates ata relative elevation <strong>of</strong> about 45 cm (A) above mean sea level (MSL). Therate <strong>of</strong> sea-level rise was decreased to 0.2 cm/yr (A); parameter values wereo<strong>the</strong>rwise identical to those in <strong>the</strong> previous example (Fig. 5).may drive species replacements or succession. Because <strong>of</strong> <strong>the</strong>sensitivity <strong>of</strong> species’ productivities to hydroperiod, anomaliesin mean sea level and astronomically forced changesin MHHW can alter <strong>the</strong> dynamics <strong>of</strong> competing species ondecadal or shorter time scales.Species differ in <strong>the</strong>ir tolerance to flooding, hypoxia, desiccation,and salt stress (Pennings & Callaway 1992, Kuhn &Zedler 1997). Consequently, <strong>the</strong> fundamental distributions<strong>of</strong> plant species within <strong>the</strong> intertidal zone are determined byphysical factors that set upper and lower depth limits, andoptimum depths. The realized distributions may or may notresemble closely <strong>the</strong> fundamental distributions, dependingon <strong>the</strong> strength <strong>of</strong> competitive or facilitative interactionsamong species (Ungar 1998, Bertness 1991). For a givenrate <strong>of</strong> sea-level rise, <strong>the</strong> relative shapes <strong>of</strong> <strong>the</strong>se distributionsultimately determine <strong>the</strong> equilibrium elevation, speciesreplacements and persistence. One species may replaceano<strong>the</strong>r by modifying <strong>the</strong> elevation <strong>of</strong> its habitat to <strong>the</strong> detriment<strong>of</strong> competitors. This is characteristic <strong>of</strong> species withrelatively high biomass densities such as <strong>the</strong> Spartina hybridswarms that have become established in San Francisco Bay(Daehler & Strong 1997, Ayres et al. 2004).These insights are important to our fundamental understating<strong>of</strong> <strong>the</strong> ecology <strong>of</strong> marsh ecosystems and should alsobe relevant to <strong>the</strong> management community responsible forforecasting and controlling <strong>the</strong> course <strong>of</strong> species invasions.On a fundamental level, <strong>the</strong> work demonstrates how processesthat operate at different temporal scales can interactto modify ecosystem structure and function. For example,<strong>the</strong> outcome <strong>of</strong> interspecific competition among marsh macrophytes,a biological process that operates on relative shorttime scales, can be affected by <strong>the</strong> rate <strong>of</strong> sea-level rise, aprocess that occurs on very long time scales.- 114 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadACKNOWLEDGMENTSThis research has been supported in part by a grantfrom <strong>the</strong> US Environmental Protection Agency’s Scienceto Achieve Results (STAR) Estuarine and Great Lakes(EaGLe) program through funding to <strong>the</strong> ACE-INC, US EPAAgreement R82-867701, but does not necessarily reflect <strong>the</strong>views <strong>of</strong> <strong>the</strong> Agency and no <strong>of</strong>ficial endorsement should beinferred. This research was also supported in part by <strong>the</strong>NSF LTREB and LTER programs. I thank <strong>the</strong> anonymousreviewers whose thoughtful criticisms were quite helpful.REFERENCESAyres D.R., D.L. Smith, K. Zaremba, S. Klohr, and D.R. Strong.2004. Spread <strong>of</strong> exotic cordgrasses and hybrids (Spartina sp.)in <strong>the</strong> tidal marshes <strong>of</strong> San Francisco Bay, California, USA.Biological Invasions 6:221-231.Bertness, M.D. 1991. Zonation <strong>of</strong> Spartina patens and Spartinaalterniflora in a New England salt marsh. Ecology 72:138-148.Bertness, M.D. and P.J. Ewanchuk. 2002. Latitudinal and climatedrivenvariation in <strong>the</strong> strength and nature <strong>of</strong> biological interactionsin New England salt marshes. Oecologia 132:392-401.Christiansen, T., P.L. Wiberg., and T.G. Milligan. 2000. Flow andsediment transport on a tidal salt marsh surface. Estuarine,Coastal and Shelf Science 50:315-331.Cuddington, K., and A. Hastings. 2004. Invasive engineers.Ecological Modelling 178:335-347.Daehler, C.C., and D.R. Strong. 1997. Hybridization betweenintroduced smooth cordgrass (Spartina alterniflora; Poaceae)and native California cordgrass (S. foliosa) in San Francisco Bay,California, USA. American Journal <strong>of</strong> Botany 84:607-611.Dame, R., M. Alber, D. Allen, M. Mallin, C. Montague, A. Lewitus,A. Chalmers, R. Gardner, C. Gilman, B. Kjerfve, J. Pinckney, andN. Smith. 2000. Estuaries <strong>of</strong> <strong>the</strong> south Atlantic coast <strong>of</strong> NorthAmerica: Their geographical signatures. Estuaries 23:793-819.Eiser, W.C., and B. Kjerfve. 1986. Marsh topography and hypsometriccharacteristics <strong>of</strong> a South Carolina salt marsh basin.Estuarine, Coastal and Shelf Science 23:595-605.Emergy, N.C., P.J. Ewanchuk, and M.D. Bertness. 2001. Competitionand salt-marsh plant zonation: Stress tolerators may be dominantcompetitors. Ecology 82:2471-2483.French, J. R. 1993. Numerical simulation <strong>of</strong> vertical marsh growthand adjustment to acceleerated sea-level rise, north Norfolk, U.K. Earth Surface Processes and Landforms 18: 63-81.Gleason, M.L., D.A. Elmer, N.C. Pien,and J.S. Fisher. 1979. Effects<strong>of</strong> stem density upon sediment retention by salt marsh cord grass,Spartina alterniflora Loisel. Estuaries 2:271-273.Hacker, S.D., and M.D. Bertness. 1999. Experimental evidence forfactors maintaining plant species diversity in a New England saltmarsh. Ecology 80:2064-2073.Hutchinson, G.E. 1957. Concluding Remarks. Cold Spring HarborSymposia on Quantitative Biology. 22: 415-42Jones, C.G., J.H. Lawton, and M. Shachk. 1997. Positive andnegative effects <strong>of</strong> organisms as physical ecosystem engineers.Ecology 78:1946-1957.Krone, R.B., 1985. Simulation <strong>of</strong> marsh growth under rising sealevels. In: Waldrop, W.R. (Ed.), Hydraulics and Hydrology in <strong>the</strong>Small Computer Age, Hydraulics Div. ASCE, Washington, D.C.,pp. 316-323.Kuhn, N.L. and J.B. Zedler. 1997. Differential effects <strong>of</strong> salinityand soil saturation on native and exotic plants <strong>of</strong> a coastal saltmarsh. Estuaries 20: 391-403.Leonard, L.A. and M.E. Lu<strong>the</strong>r. 1995. Flow dynamics in tidal marshcanopies. Limnology and Oceanography 40:11474-11484.Morris, J.T. 1995. The salt and water balance <strong>of</strong> intertidal sediments:results from North Inlet, South Carolina. Estuaries 18:556-567.Morris, J.T. 2000. Effects <strong>of</strong> sea level anomalies on estuarineprocesses. In: Hobbie, J., ed., Estuarine Science: A Syn<strong>the</strong>ticApproach to Research and Practice. Island Press, pp.107-127.Morris, J.T., D. Porter, M. Neet, P.A. Noble, L. Schmidt, L.A.Lapine and J. Jensen. 2005. Integrating LIDAR elevation data,multi-spectral imagery, and neural network modeling for marshcharacterization. <strong>International</strong> Journal <strong>of</strong> Remote Sensing.26:5221-5234Morris, J.T., P.V. Sundareshwar, C.T. Nietch, B. Kjerfve and D.R.Cahoon. 2002. Responses <strong>of</strong> coastal wetlands to rising sea level.Ecology 83: 2869-2877.Pennings, S.C., M.B. Grant and M.D. Bertness. 2005. Plantzonation in low-latitude salt marshes: disentangling <strong>the</strong> roles<strong>of</strong> flooding, salinity and competition. Journal <strong>of</strong> Ecology93:159-167.Pennings, S.C., and R.M. Callaway. 1992. Salt-marsh plant zonation-The relative importance <strong>of</strong> competition and physical factors.Ecology 73:681-690.Pethick, J.S. 1981. Long-term accretion rates on tidal marshes.Journal <strong>of</strong> Sedimentary Petrology 61:571-577.Redfield, A.C. 1972. Development <strong>of</strong> a New England salt marsh.Ecological Monographs 42:201-237.Rooth, J.E. and J.C. Stevenson. 2000. Sediment deposition patternsin Phragmites australis communities: Implications for coastalareas threatened by rising sea-level. Wetlands Ecology andManagement 8:173-183.Silvestri, S., A. Defina and M. Marani. 2005. Tidal regime, salinityand salt marsh plant zonation. Estuarine Coastal and ShelfScience 62:119-130.Stevenson, J.C., L.G. Ward and M.S. Kearney. 1986. Vertical accretionin marshes with varying rates <strong>of</strong> sea level rise. In: Wolfe,D.A., ed., Estuarine Variability. Academic Press, New York, pp.241-259.Stumpf, R.P. and J.W. Haines. 1998. Variations in tidal level in <strong>the</strong>Gulf <strong>of</strong> Mexico and implications for tidal wetlands. Estuarine,Coastal and Shelf Science 46:165–173.Unger, I.A. 1998. Are biotic factors significant in influencing <strong>the</strong>distribution <strong>of</strong> halophytes in saline habitats? Botanical Review64: 176-199.Windham, L. 1999. Microscale spatial distribution <strong>of</strong> Phragmitesaustralis (common reed) invasion into Spartina patens (salthay)-dominated communities in brackish tidal marsh. BiologicalInvasions 1:137-148.Yang, S.L., 1998. The role <strong>of</strong> Scirpus marsh in attenuation <strong>of</strong> hydrodynamicsand retention <strong>of</strong> fine sediment in <strong>the</strong> Yangtze estuary.Estuarine, Coastal and Shelf Science 47:227-233.Zedler, J.B., S. Kercher. 2004. Causes and consequences <strong>of</strong> invasiveplants in wetlands: Opportunities, opportunists, and outcomes.Critical Reviews in Plant Sciences 23:431-452.- 115 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadMODELING THE SPREAD OF INVASIVE SPARTINA HYBRIDS IN SAN FRANCISCO BAYR.J. HALL,A.M.HASTINGS AND D.R. AYRESUniversity <strong>of</strong> California, Davis, One Shields Ave., Davis, CA 95616; rjhall@ucdavis.eduThe emergence <strong>of</strong> highly fit hybrids between native and introduced species is an increasinglywidespread problem that can impact entire ecosystems. In San Francisco Bay, a swarm <strong>of</strong> hybridcordgrass (Spartina foliosa x alterniflora) is covering vast areas <strong>of</strong> intertidal mudflat, threatening <strong>the</strong>native S. foliosa with extinction. Here we outline a modeling approach for assessing <strong>the</strong> relativeimportance <strong>of</strong> elevated hybrid fitness traits in explaining <strong>the</strong> rapidity <strong>of</strong> this invasion. Wedemonstrate that elevated growth rate, seedling survival and pollen production <strong>of</strong> hybrids relative to<strong>the</strong> native species interact to predict <strong>the</strong> observed faster-than-exponential spread <strong>of</strong> cordgrassthrough <strong>the</strong> Bay.Keywords: invasion, model, hybridization, SpartinaINTRODUCTIONAn exotic introduced into a new range is <strong>of</strong>tenphenomenally successful; this may be due to exploitation <strong>of</strong>a previously unfilled niche, a release from natural enemies,or increased competitive ability relative to <strong>the</strong> native biota(Williamson 1996). However, in many ecosystems, <strong>the</strong>emergence <strong>of</strong> hybrids between native and exotic species is<strong>the</strong> greatest threat to local biodiversity, particularly if <strong>the</strong>hybrids have superior fecundity or suvivorship comparedwith <strong>the</strong> native (Wolf et al. 2001). An example <strong>of</strong> this is <strong>the</strong>invasion <strong>of</strong> hybrid cordgrass Spartina foliosa x alterniflorain San Francisco Bay. Since <strong>the</strong> introduction <strong>of</strong> S.alterniflora in <strong>the</strong> 1970s, hybrids have emerged whosegrowth rates, fecundity and tolerance to environmentalconditions exceed that <strong>of</strong> both parental lineages (Ayres et al.2003). The hybrids impact on many organisms in <strong>the</strong> Bayby covering open mudflat crucial to feeding shorebirds,and changing <strong>the</strong> tidal pr<strong>of</strong>ile through sedimentaccumulated in <strong>the</strong>ir root mass. Contrary to classicalecological <strong>the</strong>ory, which suggests that a species withunlimited resources can grow at most exponentially (e.g.,Turchin 2003), <strong>the</strong> area covered by hybrid cordgrass in <strong>the</strong>Bay has increased super-exponentially (i.e. <strong>the</strong> growth rateper unit area is increasing through time; Fig. 1). If thisinvasion continues unchecked, it seems likely that <strong>the</strong>native S. foliosa is doomed to extinction throughintrogression (Ayres et al. 2003).In order to design effective control strategies, it is vitalto determine <strong>the</strong> key hybrid traits responsible for drivingthis invasion. Increased vegetative growth rates, elevatedpollen production and seedling survival <strong>of</strong> hybrid Spartinarelative to <strong>the</strong> native have all been proposed as contributingto <strong>the</strong> rate and extent <strong>of</strong> this invasion (Ayres et al. 2004).Ma<strong>the</strong>matical models are a useful tool for investigating <strong>the</strong>relative importance <strong>of</strong> <strong>the</strong>se mechanisms. In this paper weoutline <strong>the</strong> modeling approach used to describe <strong>the</strong>dynamics <strong>of</strong> this hybridization. We use <strong>the</strong> model to testhow increased vegetative growth rate, seedling survival andpollen production <strong>of</strong> hybrids affect <strong>the</strong> rate <strong>of</strong> populationexpansion in areas <strong>of</strong> high and low recruitment. We find thatdepending on <strong>the</strong> level <strong>of</strong> local recruitment, increasedvegetative growth rate or seedling survival <strong>of</strong> hybrids canresult in super-exponential population growth. Elevatedhybrid pollen production alone results in an increase in <strong>the</strong>frequency <strong>of</strong> hybrids in <strong>the</strong> population, but does not predictsuper-exponential growth unless coupled with elevatedhybrid growth rate or seedling survival.MODEL AND METHODSHere we outline <strong>the</strong> key modeling assumptions andsketch <strong>the</strong> derivation <strong>of</strong> <strong>the</strong> model. The derivation <strong>of</strong> <strong>the</strong> fullquantitative genetic, integro-difference equation model isFig. 1: super-exponential (bold line) vs exponential (dashed line) populationgrowth. After a sufficiently long time, <strong>the</strong> area occupied by aninvasive growing super-exponentially differs from an exponentiallygrowing population by orders <strong>of</strong> magnitude.- 117 -

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaFig. 2 (a) The distribution <strong>of</strong> gamete genotypes as a function <strong>of</strong> parental genotypes. The broad distribution <strong>of</strong> hybrid genotypes reflects <strong>the</strong> higher geneticdiversity <strong>of</strong> <strong>the</strong> hybrids compared with <strong>the</strong> pure lineages. (b) Functional form <strong>of</strong> <strong>the</strong> genotype-dependent (bold line) and genotype-independent (dashed line)trait values (growth rate, seedling survival or pollen production).outlined in Hall et al. 2006. In order to elucidate <strong>the</strong> role <strong>of</strong>genetic recombination in driving this invasion, we makeseveral simplifying assumptions about <strong>the</strong> populationdynamics. Specifically, we assume spatial andenvironmental homogeneity, and focus on <strong>the</strong> spread <strong>of</strong>plants into open mudflat (i.e., we ignore density-dependenteffects such as crowding). In <strong>the</strong> spirit <strong>of</strong> many quantitativegenetics models (e.g., Bulmer 1980), we assume that a largenumber <strong>of</strong> loci interact independently and additively todetermine phenotypic fitness, and ignore <strong>the</strong> effects <strong>of</strong>environment on phenotype.Plants are classified by <strong>the</strong>ir genotype value, x, scaledsuch that pure S. foliosa has genotypes clustered around x=-1, pure S. alterniflora around x=+1, with hybrids spanning<strong>the</strong> intermediate range <strong>of</strong> genotype values. Plants <strong>of</strong>genotype x produce gametes whose genotype is drawn froma Normal distribution with mean x/2. The variance <strong>of</strong> thisdistribution is genotype-specific, under <strong>the</strong> assumption thathybrids can produce genetically more diverse <strong>of</strong>fspring thanei<strong>the</strong>r parental lineage (Fig. 2a). When pollen <strong>of</strong> genotype ysuccessfully combines with an ovule <strong>of</strong> type z, <strong>the</strong> resultingseed has genotype y+z.The population-level model is formulated in terms <strong>of</strong><strong>the</strong> total area occupied by plants <strong>of</strong> genotype x in year t,denoted P t (x). The area occupied by plants in year t+1 isgiven by <strong>the</strong> sum <strong>of</strong> vegetative growth <strong>of</strong> adult plants in yeart and <strong>the</strong> area occupied by new seedlings. Ma<strong>the</strong>maticallythis is expressed byP t+1 (x)=G(x) P t (x)+S(x) N t (x),where G(x) is <strong>the</strong> (genotype-dependent) vegetative growthrate, S(x) is <strong>the</strong> survival <strong>of</strong> seedlings <strong>of</strong> genotype x, and N t (x)is <strong>the</strong> total number <strong>of</strong> seeds produced in year t <strong>of</strong> type x. Thenumber <strong>of</strong> seeds produced <strong>of</strong> type x is <strong>the</strong> product <strong>of</strong> <strong>the</strong>frequency <strong>of</strong> pollen produced by male parents <strong>of</strong> genotype y(proportional to <strong>the</strong> genotype-specific pollen production rateM), <strong>the</strong> number <strong>of</strong> ovules from female parents <strong>of</strong> genotype z,and <strong>the</strong> probability that pollen <strong>of</strong> type y and ovules <strong>of</strong> type zcombine to produce a seed <strong>of</strong> type x, summed over allpossible genotype values <strong>of</strong> <strong>the</strong> male and female parents.The above model is used to test <strong>the</strong> effects <strong>of</strong> elevatedhybrid growth rate, seedling survival and pollen production(G, S and M respectively) on <strong>the</strong> total area occupied bySpartina <strong>of</strong> all genotypes, and <strong>the</strong> frequency <strong>of</strong> hybridgenotypes in <strong>the</strong> population. For comparison, <strong>the</strong> model isalso run under a null scenario whereby all genotypes areequally fit. The functional dependence <strong>of</strong> an elevated traitvalue on a plant’s genotype is depicted in Fig. 2b. Twolevels <strong>of</strong> seedling recruitment are also considered: low (asmight be expected at a site subject to strong tidal action) andhigh (e.g. in a sheltered inlet or marsh restoration site).RESULTSWhen <strong>the</strong> model is run under <strong>the</strong> assumption that allSpartina genotypes are equally fit, <strong>the</strong> population does notundergo super-exponential growth. However, even in <strong>the</strong>absence <strong>of</strong> any fitness advantage to <strong>the</strong> hybrid, <strong>the</strong> frequency<strong>of</strong> pure S. foliosa plants in <strong>the</strong> population suffers a slowdecline, suggesting that given sufficiently long time, <strong>the</strong>hybrid swarm could dominate.If <strong>the</strong> baseline level <strong>of</strong> seedling recruitment is low, anelevated vegetative growth rate <strong>of</strong> hybrids can produce <strong>the</strong>observed faster-than-exponential population growth.Elevated seedling survival <strong>of</strong> hybrids also results in superexponentialgrowth, but at a much less dramatic rate.Conversely, in regions where seedling recruitment is high,elevated hybrid seedling survival produces faster population- 118 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and Spreadexpansion than elevated hybrid growth rate. Elevated hybridgrowth rates and seedling survival both result in a rapiddecline in <strong>the</strong> frequency <strong>of</strong> native genotypes: essentially as<strong>the</strong> area covered by hybrid cordgrass increases, <strong>the</strong>probability <strong>of</strong> hybrid pollen fertilizing <strong>the</strong> native alsoincreases. This ‘pollen-swamping' suggests that ultimatelyall new recruits to <strong>the</strong> Spartina population will be hybrid,resulting in extinction <strong>of</strong> <strong>the</strong> native through introgression.In both recruitment scenarios, <strong>the</strong> effect <strong>of</strong> elevatedrates <strong>of</strong> hybrid pollen production on <strong>the</strong> rate <strong>of</strong> invasion isnegligible, assuming <strong>the</strong> hybrids have no o<strong>the</strong>r fitnessadvantage. However, <strong>the</strong> frequency <strong>of</strong> native S. foliosagenotypes declines at a much faster rate than in <strong>the</strong> nullmodel (where all genotypes produce <strong>the</strong> same amount <strong>of</strong>viable pollen). If elevated hybrid pollen production iscoupled with an increased hybrid growth rate or seedlingsurvival, <strong>the</strong> resulting invasion proceeds faster than it wouldfor elevated hybrid growth rates or seedling survival alone.The effects <strong>of</strong> <strong>the</strong> differing life-history processes on <strong>the</strong> rate<strong>of</strong> invasion <strong>of</strong> hybrid Spartina genotypes are summarized inTable 1.DISCUSSIONHere we have shown that a model which allows forelevated fitness traits (growth, seedling survival, pollenproduction) in hybrid Spartina can explain its explosivepropagation through San Francisco Bay. As well aspredicting a huge increase in <strong>the</strong> total area covered bySpartina in <strong>the</strong> Bay, <strong>the</strong> model also shows a sharp decline in<strong>the</strong> frequency <strong>of</strong> pure S. foliosa genotypes over <strong>the</strong> nextcentury, as an increasingly higher percentage <strong>of</strong> seedlingshave hybrid origin. If this invasion continues unchecked, <strong>the</strong>native S. foliosa will become extinct in <strong>the</strong> Bay. O<strong>the</strong>rmodels <strong>of</strong> invasion and hybridization predict <strong>the</strong> ultimateextinction <strong>of</strong> <strong>the</strong> species (Huxel 1999; Wolf et al. 2001). Thework presented here extends previous modeling efforts bylinking <strong>the</strong> population dynamics and genetics <strong>of</strong> <strong>the</strong>invasion, making transparent <strong>the</strong> roles <strong>of</strong> geneticrecombination and selection acting on favourable phenotypictraits.Hybrid Spartina needs to be eradicated to ensure <strong>the</strong>persistence <strong>of</strong> <strong>the</strong> native S. foliosa in San Francisco Bay.Super-exponential population growth <strong>of</strong> hybrids may bedriven by different life-history processes depending on <strong>the</strong>level <strong>of</strong> local recruitment. In accordance with recentmodeling work on <strong>the</strong> control <strong>of</strong> S. alterniflora in WillapaBay, Washington (Taylor and Hastings 2004), this suggeststhat control efforts at sites with differing levels <strong>of</strong> seedlingrecruitment should be targeted at different stage classes <strong>of</strong>Spartina. At low recruitment sites, control should beTable 1: Effects <strong>of</strong> elevated life-history traits and parental compatibilityon <strong>the</strong> rate <strong>of</strong> invasion <strong>of</strong> hybrid Spartina.TraitAll genotypesequally fitElevated growth rate(G)Elevated seedlingsurvival (S)Elevated pollenproduction (M)Effect on growth rateExponential growth; increase infrequency <strong>of</strong> hybridsSuper-exponential; effect greatestwhen seedling recruitment lowSuper-exponential; effect greatestwhen seedling recruitment highExponential growth; rapid increasein hybrid frequency; superexponentialonly in conjunction wi<strong>the</strong>levated S or Mtargeted at <strong>the</strong> removal <strong>of</strong> fast-growing isolated clones. Atsites with potentially a high level <strong>of</strong> recruitment, such asmarsh restoration sites, control should focus on <strong>the</strong> removal<strong>of</strong> seedlings, or herbicide treatment <strong>of</strong> hybrid Spartinameadows during <strong>the</strong> pollination season.ACKNOWLEDGMENTSWe gratefully ackowledge funding from NSFBiocomplexity Grant AESRJ85. RJH is grateful to CazTaylor and John Lambrinos for helpful discussions.REFERENCESAyres, D.R., D.R. Strong, and P. Baye, 2003. Spartina foliosa(Poaceae) – a common species on <strong>the</strong> road to rarity? Madrono50:209-213.Ayres, D.R., D.L. Smith, K. Zaremba, S.Klohr, and D.R. Strong,2004. Spread <strong>of</strong> exotic cordgrasses and hybrids (Spartina sp.) in<strong>the</strong> tidal marshes <strong>of</strong> San Francisco Bay, California, USA. BiologicalInvasions 6: 221-231.Bulmer, M.G. 1980. The ma<strong>the</strong>matical <strong>the</strong>ory <strong>of</strong> quantitative genetics.Oxford University Press, UK.Hall, R.J., A. Hastings, and D.R. Ayres, 2006. Explaining <strong>the</strong> explosion:modelling hybrid invasions. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> RoyalSociety B 273:1385-1389.Huxel, G.R. 1999. Rapid displacement <strong>of</strong> native species by invasivespecies: effects <strong>of</strong> hybridization. Biological Conservation89:143-152.Taylor, C.M., and A. Hastings, 2004. Finding optimal controlstrategies for an invasive grass using a density-structured model.Journal <strong>of</strong> Applied Ecology 41:1049-1057.Turchin, P. 2003. Complex population dynamics. Princeton UniversityPress, USA.Williamson, M. 1996. Biological invasions. Chapman and Hall,London, UK.Wolf, D.E., N. Takebayashi, and L.H. Rieseberg, 2001. Predicting<strong>the</strong> risk <strong>of</strong> extinction through hybridization. Conservation Biology15:1039-1053.- 119 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadMODELING THE SPREAD AND CONTROL OF SPARTINA ALTERNIFLORA IN A PACIFICESTUARYC.M. TAYLOR 1 ,A.HASTINGS 2 ,H.G.DAVIS 3 ,J.C.CIVILLE 4,6 AND F.S. GREVSTAD 51 Department <strong>of</strong> Ecology and Evolutionary Biology, Tulane University, 6823 St. Charles Ave., New Orleans, LA 70118;caz@tulane.edu2 Department <strong>of</strong> Environmental Science and Policy, University <strong>of</strong> California, One Shields Ave., Davis, CA 95616-87553 Center for Population Biology, University <strong>of</strong> California, Davis, One Shields Ave., Davis, CA 95616; hgdavis@sonic.net4 Section <strong>of</strong> Evolution & Ecology, University <strong>of</strong> California, Davis, One Shields Ave., Davis, CA 956166 Current address: 2731 Be<strong>the</strong>l St. NE, Olympia, WA 98506; jciville@comcast.net5 Olympic Natural Resources Center, University <strong>of</strong> Washington, 2907 Pioneer Road, Long Beach, WA 98331Results from both a spatially-explicit simulation model and a spatially-implicit analytical model showthat an Allee effect can slow <strong>the</strong> spread <strong>of</strong> an invasive plant, Spartina alterniflora at sites within a Pacificcoast estuary. The average rate <strong>of</strong> spread with <strong>the</strong> Allee effect is about 20% per year. Removing <strong>the</strong> Alleeeffect results in an average rate <strong>of</strong> spread <strong>of</strong> about 30% per year. The analytical model which partitions<strong>the</strong> population according to density classes, instead <strong>of</strong> <strong>the</strong> more usual age or stage classes was used inconjunction with a genetic algorithm to investigate density-based eradication strategies. We ask whe<strong>the</strong>rit is more efficient to first remove low-density plants at <strong>the</strong> edge <strong>of</strong> an invasion site that produce fewerpropagules but spread rapidly by rhizomes or to remove high-density plants that spread slowly butproduce most <strong>of</strong> <strong>the</strong> new recruits. We explored <strong>the</strong> consequences <strong>of</strong> <strong>the</strong> Allee effect and different annualbudget levels on <strong>the</strong> optimal strategy. We found that <strong>the</strong> optimal strategy was dependent on annualbudget levels. At low budgets, it was necessary to remove <strong>the</strong> low-density areas first to achieveeradication but if a high budget is available <strong>the</strong>n <strong>the</strong> optimal strategy is to prioritise high-density areas.Without an Allee effect <strong>the</strong> optimal strategy would always prioritize <strong>the</strong> removal <strong>of</strong> <strong>the</strong> fast-growing lowdensityareas. Given uncertainty in future budgets, we recommend a strategy that prioritizses <strong>the</strong> removal<strong>of</strong> low density plants over <strong>the</strong> high density plants. The reproductive Allee effect in this system is notsufficiently strong to outweigh <strong>the</strong> importance <strong>of</strong> <strong>the</strong> rapid vegetative spread.Keywords: Allee effect, Spartina alterniflora, Willapa Bay, invasive species, control strategy,eradication, pollen limitationThis paper describes two models that we developed tostudy <strong>the</strong> invasion <strong>of</strong> Spartina alterniflora into Willapa Bay inWashington State, USA. The first is a spatially explicitsimulation model that we used to investigate population levelconsequences <strong>of</strong> an Allee effect caused by pollen limitation.The second is a derived non-spatial analytical model that weused to investigate eradication strategies.As is described by o<strong>the</strong>rs at this conference (Davis et al.;Civille et al. this issue), S. alterniflora (smooth cordgrass,hereafter referred to as Spartina) was introduced to Willapabay just over 100 years ago. It was introduced into a smallnumber <strong>of</strong> locations around <strong>the</strong> bay and <strong>the</strong>n spread itselfaround <strong>the</strong> whole estuary. Individuals become established asseedlings <strong>the</strong>n grow, vegetatively, into circular clones.Eventually <strong>the</strong> clones merge toge<strong>the</strong>r to form meadows. By2002, many <strong>of</strong> <strong>the</strong> mudflats <strong>of</strong> Willapa Bay had beencompletely converted to cordgrass meadow and almost all <strong>of</strong><strong>the</strong> mudflats have been invaded to some degree, although <strong>the</strong>reare still extensive areas <strong>of</strong> mostly unvegetated mudflat. Atypical mudflat is in transition, a meadow has already formed,uncoalesced clones are scattered at <strong>the</strong> edge <strong>of</strong> <strong>the</strong> meadowand <strong>the</strong> rest <strong>of</strong> <strong>the</strong> mudflat is uninvaded.When we measured seedset in <strong>the</strong>se clones we found thatseed production in <strong>the</strong> meadows is vastly greater than (morethan ten times) seed production in <strong>the</strong> uncoalesced clones(Davis et al. 2004a). This reduced seed production whendensity is low is an example <strong>of</strong> an Allee effect, definedgenerally as a positive relationship between any component <strong>of</strong>fitness (in this case seed production) and conspecific density(Stephens et al. 1999). Later experiments demonstrated thatpollen limitation is <strong>the</strong> mechanism that causes <strong>the</strong> Allee effectin <strong>the</strong> Willapa Bay population <strong>of</strong> S. alterniflora (Davis et al.2004b).SIMULATION MODEL OF SPARTINA SPREADWe calculated, from data collected in <strong>the</strong> field, that <strong>the</strong>meadows produce an average <strong>of</strong> approximately 500 seeds persquare meter (m -2 ) and that <strong>the</strong> clones only produceapproximately 12 seeds m -2 . Data from GIS maps from 1994and 1997 (Civille 2005) was used to estimate seed dispersaldistances, establishment probability and vegetative growth- 121 -

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinarates and, using <strong>the</strong>se parameters, we built a spatially explicitsimulation model <strong>of</strong> <strong>the</strong> spread <strong>of</strong> Spartina. Full details <strong>of</strong> <strong>the</strong>model and its parameterization are given in Taylor et al. 2004.The model represents one square kilometer <strong>of</strong> mudflat,arranged as a square lattice <strong>of</strong> 1000 by 1000 cells; each cell isone square meter and can ei<strong>the</strong>r be occupied by Spartina orvacant. The simulated invasion starts from a single occupiedcell in <strong>the</strong> center <strong>of</strong> <strong>the</strong> lattice, representing a single 1-m 2clone, and is run for 100 years. In each year <strong>the</strong> plants growvegetatively into empty neighboring sites and produce arandom number <strong>of</strong> seeds (distributed according to a Poissondistribution), <strong>the</strong> seeds disperse exponentially away from <strong>the</strong>parents and have a small probability <strong>of</strong> becoming establishedclones. When two clones become adjacent to one ano<strong>the</strong>r, thatpatch <strong>of</strong> Spartina is classified as a meadow. We have twolevels <strong>of</strong> seed production; cells that are classified as clonesproduce a small number <strong>of</strong> seeds; cells that are classified asmeadows produce a larger number. We created a variant <strong>of</strong> <strong>the</strong>model in which we removed <strong>the</strong> Allee effect by setting <strong>the</strong>seed production <strong>of</strong> clones to <strong>the</strong> same as <strong>the</strong> seed production <strong>of</strong>meadows.We found that <strong>the</strong> Allee effect dramatically slows <strong>the</strong>invasion. From 380 runs <strong>of</strong> <strong>the</strong> model, with parameters variedover <strong>the</strong>ir estimated ranges, we calculated an average increasein area occupied <strong>of</strong> about 20% per year with <strong>the</strong> Allee effect;<strong>the</strong> area occupied by Spartina doubles approximately every 5years. Removing <strong>the</strong> Allee effect and repeating those runs gavean average <strong>of</strong> 30% increase per year, a doubling time <strong>of</strong> lessthan 4 years (Taylor et al. 2004) (Fig. 1). This result isconsistent with <strong>the</strong>oretical investigations <strong>of</strong> <strong>the</strong> consequences<strong>of</strong> an Allee effect in invasions (Lewis and Kareiva 1993, Wangand Kot 2001, Wang et al. 2002).ANALYTICAL MODEL OF SPARTINA SPREADWe <strong>the</strong>n created a non-spatial analytical model <strong>of</strong> thissame process (Taylor et al. 2004). We kept some <strong>of</strong> <strong>the</strong>structure <strong>of</strong> <strong>the</strong> simulation model by using as <strong>the</strong> mainvariables <strong>the</strong> area occupied by seedlings (S), <strong>the</strong> area occupiedby clones (C) and <strong>the</strong> area occupied by meadows (M). O<strong>the</strong>rvariables are <strong>the</strong> number <strong>of</strong> individual clones (N C ) and <strong>the</strong>number <strong>of</strong> meadows (N M ).The main equations <strong>of</strong> this model are:S t+1 = f C C t + f M M t (1)C t+1 = S t + (1-)g c C t (2)M t+1 = g c C t + g M M t (3)Equation 1 says that seedlings are created from <strong>the</strong> seedsproduced by clones and seeds produced by meadows. Thefecundity <strong>of</strong> <strong>the</strong> clones (f C ) is much smaller than <strong>the</strong> fecundity<strong>of</strong> <strong>the</strong> meadows (f M ) when <strong>the</strong>re is an Allee effect. We remove<strong>the</strong> Allee effect in <strong>the</strong> same way as in <strong>the</strong> simulation model, bysetting f C equal to fM. Equation 2 says that <strong>the</strong> area occupiedFig. 1: Mean area occupied by Spartina as predicted by 380 runs <strong>of</strong> <strong>the</strong> simulationmodel with and without <strong>the</strong> Allee effect. The upper line (squares)shows <strong>the</strong> predicted area occupied when <strong>the</strong>re is no Allee effect and <strong>the</strong>lower line (circles) shows <strong>the</strong> predicted area occupied with <strong>the</strong> Allee effect.A modified form <strong>of</strong> this figure was originally published in Taylor et al. 2004,Ecology 85(12).by clones increases by seedlings becoming clones and also byvegetative spread <strong>of</strong> existing clones (at growth rate g C ) anddecreases by a proportion <strong>of</strong> <strong>the</strong> area occupied by clones ()coalescing into o<strong>the</strong>r patches <strong>of</strong> Spartina and becomingmeadow. Equation 3 describes <strong>the</strong> dynamics <strong>of</strong> area occupiedby meadow which increases by <strong>the</strong> clones coalescing intomeadow at rate and also by vegetative growth <strong>of</strong> existingmeadow (growth rate g M ). These equations are only a partialdescription <strong>of</strong> <strong>the</strong> model. The full model includes equations for<strong>the</strong> dynamics <strong>of</strong> <strong>the</strong> number <strong>of</strong> clones (N C ) and <strong>the</strong> number <strong>of</strong>meadows (N M ) and also describes how <strong>the</strong> parameters (,f C ,f M ,g C , and g M ) change with <strong>the</strong> area occupied by clones andmeadows and number <strong>of</strong> clones and meadows. A fulldescription <strong>of</strong> this model is provided in (Taylor et al. 2004).Several comparisons <strong>of</strong> <strong>the</strong> analytical model with <strong>the</strong>simulation model give us confidence that <strong>the</strong> two modelspredict <strong>the</strong> same dynamics. We use <strong>the</strong> analytical model to findoptimal density-based eradication strategies. Full details <strong>of</strong> thiswork are given in Taylor and Hastings 2004, and we give abrief summary here.MANAGEMENT STRATEGIESEfforts to control or eradicate Spartina in Willapa Baybegan in <strong>the</strong> early 1990s and it soon became obvious thateradicating this plant this was not going to be easy or cheap.Spartina can be removed mechanically by mowing or diggingand it can be killed with herbicide but all <strong>of</strong> <strong>the</strong> methodsrequire specialized equipment and repeated treatments (Patten,this issue). Large scale efforts to control this invasion began in2003 and current plans predict that <strong>the</strong> invasion will beeradicated within ten years if <strong>the</strong> current level <strong>of</strong> control ismaintained (Murphy 2003).- 122 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadFig. 2: Grey area shows area occupied by Spartina invasion <strong>of</strong> different agesas predicted by <strong>the</strong> analytical model and black area shows minimum areathat needs to be removed each year for eradication to succeed within tenyears.We used <strong>the</strong> model to explore two questions about control.First, how much Spartina has to be removed each year in orderto clear this invasion? Second, is it better to target <strong>the</strong> highdensity meadows that are producing most <strong>of</strong> <strong>the</strong> seed but notspreading vegetatively very quickly; or is it more effective totarget <strong>the</strong> low density clones that produce very little seed(because <strong>of</strong> <strong>the</strong> Allee effect) but spread vegetatively relativelymuch more quickly because <strong>the</strong>y are surrounded by open mud?We assume a fixed annual budget and that <strong>the</strong>re is a fixed costper square meter <strong>of</strong> removing Spartina; this means that a fixedmaximum area can be removed each year. We define a controlstrategy as (1) T t : <strong>the</strong> actual area <strong>of</strong> Spartina removed in eachyear t and (2) X t <strong>the</strong> proportion <strong>of</strong> <strong>the</strong> area removed that ismeadow in each year t. For a control program that lasts tenyears <strong>the</strong>re will be ten values <strong>of</strong> T, one for each year and tenvalues <strong>of</strong> X. We only consider control <strong>of</strong> clones and meadows(not seedlings) so (1 - X t ) is <strong>the</strong> proportion <strong>of</strong> <strong>the</strong> area removedthat is clones. We model control <strong>of</strong> clones by subtracting T t (1 -X t ) from <strong>the</strong> left hand side <strong>of</strong> equation 2 and we model control<strong>of</strong> meadows by subtracting T t X t from <strong>the</strong> left hand side <strong>of</strong>equation 3 (Taylor and Hastings 2004).In order to be considered successful <strong>the</strong> control strategyhas to completely eradicate <strong>the</strong> invasion within ten years. For asuccessful strategy, <strong>the</strong>re are two o<strong>the</strong>r objectives. The first isto minimize <strong>the</strong> total cost which is <strong>the</strong> same as minimizing <strong>the</strong>total area removed over <strong>the</strong> ten years (since we are assuming afixed cost per square meter). The second is to minimize <strong>the</strong>risk that this one site <strong>of</strong> <strong>the</strong> invasion poses to o<strong>the</strong>r sites inWillapa Bay. We assume that <strong>the</strong> risk <strong>of</strong> seeds escaping andcolonizing o<strong>the</strong>r mudflats in <strong>the</strong> bay is linearly proportional to<strong>the</strong> total number <strong>of</strong> seeds produced during <strong>the</strong> ten years <strong>of</strong>control. In order to minimize both cost and risk, we minimized<strong>the</strong> product <strong>of</strong> <strong>the</strong> two quantities. We used a genetic algorithmto find <strong>the</strong> optimal strategy; for details <strong>of</strong> <strong>the</strong> optimization seeTaylor and Hastings (2004). In <strong>the</strong> case when <strong>the</strong> annualA: Low BudgetB: High BudgetFig. 3: A The top left graph shows <strong>the</strong> optimal strategy (“clone first”) for eradication <strong>of</strong> a 40 year old invasion when 3600 m 2 can be removed per year for up to tenyears. White bars show <strong>the</strong> fraction <strong>of</strong> <strong>the</strong> budget applied to removal <strong>of</strong> clones and grey bars show <strong>the</strong> fraction <strong>of</strong> <strong>the</strong> budget applied to removal <strong>of</strong> meadows. Thebottom left graph shows area occupied by clones and meadows as control strategy is implemented. B. The top right graph shows <strong>the</strong> optimal strategy (“meadowfirst”) for eradication <strong>of</strong> a 40 year old invasion when 5000 m 2 can be removed per year for up to ten years. As in A, white bars show removal <strong>of</strong> clones and greybars show removal <strong>of</strong> meadows. The bottom right graph shows area occupied by clones and meadows as this control strategy is implemented. Details <strong>of</strong> <strong>the</strong> optimizationare given in text and in Taylor and Hastings (2004).- 123 -

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinabudget is lower than is necessary to eradicate <strong>the</strong> invasion, <strong>the</strong>optimization algorithm is unable to find a strategy thatremoves all <strong>the</strong> Spartina but instead finds <strong>the</strong> strategy thatminimizes <strong>the</strong> area occupied at <strong>the</strong> end <strong>of</strong> <strong>the</strong> control period.We calculated <strong>the</strong> minimum annual budget in terms <strong>of</strong>area removed that is necessary to eradicate an invasion withinten years when starting control after 40 to 100 years <strong>of</strong>unchecked invasion. The results are shown in Fig. 2. Thenecessary annual budget depends on <strong>the</strong> starting time and size<strong>of</strong> <strong>the</strong> invasion but roughly speaking it is necessary to removeannually at least <strong>the</strong> equivalent <strong>of</strong> about 15 to 20% <strong>of</strong> <strong>the</strong>starting area <strong>of</strong> <strong>the</strong> invasion.We found that <strong>the</strong> optimal allocation <strong>of</strong> resources(removal <strong>of</strong> clones versus meadows) depends on <strong>the</strong> annualbudget level (Fig. 3). For a 40 year old invasion that occupies~17,000 m 2 , if <strong>the</strong> budget is low (in this case removing at most3600m 2 per year; an area equivalent to about 22% <strong>of</strong> <strong>the</strong> initialinvasion), <strong>the</strong> optimal strategy is to remove clones firstallowing <strong>the</strong> meadows to initially expand and <strong>the</strong>n removemeadows once <strong>the</strong> clones are largely removed. If <strong>the</strong> budget ishigh (removing at most 5000 m 2 per year; an area equivalent to30% <strong>of</strong> <strong>the</strong> initial invasion), <strong>the</strong> optimal strategy is to remove<strong>the</strong> meadows first, allowing <strong>the</strong> clones to expand and startremoving clones in a later year. We also found that <strong>the</strong> clonefirststrategy is optimal when minimizing cost only andignoring <strong>the</strong> risk posed by escaping seeds and that <strong>the</strong> clonefirststrategy is <strong>the</strong> only possible way to eradicate an invasionwhen <strong>the</strong> budget is low. The meadow-first strategy can only beused when <strong>the</strong> budget is high and it is optimal because itsubstantially reduces <strong>the</strong> risk <strong>of</strong> escaping seed although it doesnot substantially reduce <strong>the</strong> cost. We also found that <strong>the</strong> clonefirststrategy is optimal when <strong>the</strong>re is no Allee effect since; inthis case, <strong>the</strong> clones produce as much seed as <strong>the</strong> meadows butgrow faster vegetatively (Taylor and Hastings 2004).CONCLUSIONSWe conclude that <strong>the</strong> Allee effect caused by pollenlimitation has dramatically slowed down <strong>the</strong> spread <strong>of</strong> thisplant and affects <strong>the</strong> optimal eradication strategy. The optimalcontrol strategy illustrates <strong>the</strong> importance <strong>of</strong> vegetative spreadin this plant. To control an invasion with limited or uncertainresources, <strong>the</strong> best and only viable strategy is to prioritize <strong>the</strong>removal <strong>of</strong> <strong>the</strong> fast growing, low density clones even though<strong>the</strong>y produce very few <strong>of</strong> <strong>the</strong> new seedlings. With higherbudgets and taking into account <strong>the</strong> risk <strong>of</strong> seed escaping tocolonize o<strong>the</strong>r sites, <strong>the</strong> optimal strategy is to prioritizeremoval <strong>of</strong> <strong>the</strong> slower-growing, high density meadows.ACKNOWLEDGMENTSWe would like to thank Don Strong, Debra Ayres,Miranda Wecker and John Lambrinos for helpful discussions.Field work was assisted by Washington State Department <strong>of</strong>Natural Resources and Washington State Fish and WildlifeService. This work was funded by <strong>the</strong> National ScienceFoundation Biocomplexity Grant #DEB0083583 (P.I. AlanHastings)REFERENCESCiville, J.C. 2005. Spatial and temporal analyses <strong>of</strong> an estuarineinvasion: Spartina alterniflora in Willapa Bay, WA. DoctoralDissertation. University <strong>of</strong> California at Davis, Davis, CA.Davis, H.G., C.M. Taylor, J.C. Civille, and D. R. Strong. 2004a. AnAllee effect at <strong>the</strong> front <strong>of</strong> a plant invasion: Spartina in a Pacificestuary. Journal <strong>of</strong> Ecology 92:321-327.Davis, H.G., C.M. Taylor, J.G. Lambrinos, and D.R. Strong. 2004b.Pollen limitation causes an Allee effect in a wind-pollinatedinvasive grass (Spartina alterniflora). PNAS 101:13804-13807.Lewis, M.A., and P. Kareiva. 1993. Allee dynamics and <strong>the</strong> spread <strong>of</strong>invading organisms. Theoretical Population Biology 43:141-158.Murphy, K C. 2003. Report to <strong>the</strong> Legislature: Progress <strong>of</strong> <strong>the</strong> 2003Spartina eradication program. PUB 805-110 (N/1/04), WashingtonState Department <strong>of</strong> Agriculture, Olympia.Stephens, P.A., W.J. Su<strong>the</strong>rland, and R. P. Freckleton. 1999. What is<strong>the</strong> Allee effect? Oikos 87:185-190.Taylor, C.M., H.G. Davis, J.C. Civille, FS. Grevstad, and A. Hastings.2004. Consequences <strong>of</strong> an Allee effect on <strong>the</strong> invasion <strong>of</strong> a Pacificestuary by Spartina alterniflora. Ecology 85:3254-3266.Taylor, C.M., and A. Hastings. 2004. Finding optimal controlstrategies for invasive species: a density-structured model forSpartina alterniflora. Journal <strong>of</strong> Applied Ecology 41:1049-1057.Wang, M.H., and M. Kot. 2001. Speeds <strong>of</strong> invasion in a model withstrong or weak Allee effects. Ma<strong>the</strong>matical Biosciences 171:83-97.Wang, M.H., M. Kot, and M.G. Neubert. 2002. Integrodifferenceequations, Allee effects, and invasions. Journal <strong>of</strong> Ma<strong>the</strong>maticalBiology 44:150-168.- 124 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 2: Spartina Distribution and SpreadHYBRID CORDGRASS (SPARTINA) AND TIDAL MARSH RESTORATION IN SAN FRANCISCO BAY:IF YOU BUILD IT, THEY WILL COMED. R. AYRES 1 AND D. R. STRONG 21,2Evolution and Ecology, University <strong>of</strong> California, Davis, CA 956161drayres@ucdavis.edu; 2 drstrong@ucdavis.eduKeywords: invasive species, salt pondsRestoration sites built on former salt ponds (Fig. 1) presentan ideal combination <strong>of</strong> biotic and abiotic conditions forSpartina regeneration by seed. First, <strong>the</strong>y are unvegetated soseedlings grow unhindered by competition with establishedplants. Second, <strong>the</strong>y are graded to present a range <strong>of</strong> intertidalelevations enhancing <strong>the</strong> opportunity for colonization at elevationsthat are nei<strong>the</strong>r too saline, nor receive too much tidalinnundation for cordgrass seedling germination and growth.<strong>Third</strong>, tidal waters enter through levee breaches (see Fig. 1),muting <strong>the</strong> force <strong>of</strong> <strong>the</strong> waves and reducing seedling loss dueto physical removal by tidal action. And finally, many <strong>of</strong> <strong>the</strong>restored marshes in <strong>the</strong> sou<strong>the</strong>rn and eastern regions <strong>of</strong> <strong>the</strong>Bay are near marshes invaded by hybrids, or are actually fedby tidal water which travels though hybrid Spartina-chokedchannels (i.e. Alameda Creek and Alameda Flood ControlChannel; se e # in Fig. 2), thus ensuring a tidally-borne seedbank rich in floating hybrid Spartina seed.The result is that recent restoration sites in <strong>the</strong> area havebeen quickly colonized by large numbers <strong>of</strong> hybrid Spartina(Table 1). This pattern can be expected to continue as morenew marshes are opened until hybrid cordgrass is controlled.Hybrid cordgrass thus presents a substantial challenge tomarsh restoration that is particularly relevant in light <strong>of</strong> <strong>the</strong>planned restoration <strong>of</strong> 25 square miles <strong>of</strong> salt ponds in <strong>the</strong>South Bay (Fig. 2).Fig. 1. Cogswell Marsh in Hayward, CA is a former salt pond opened to tidal action (note levee breaches along<strong>the</strong> shoreline and within marsh denoted by ^ ) in 1980. Non-working salt ponds are at <strong>the</strong> left and top <strong>of</strong> <strong>the</strong>photo (photos provided by California Coastal Conservancy).- 125 -

Chapter 2: Spartina Distribution and Spread<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaFig. 2. Twenty-five-square mile South Bay Salt Pond Restoration Project which will result in large-scale restoration <strong>of</strong> wetlands from <strong>the</strong> San Mateo Bridge to<strong>the</strong> sou<strong>the</strong>rn end <strong>of</strong> SF Bay. Pound sign (#) denotes hybrid Spartina invaded creeks and channels in <strong>the</strong> area; asterisk (*) is <strong>the</strong> Eden Landing restoration site(map from http://www.southbayrestoration.org/images/map.gif).Table 1. Percentage <strong>of</strong> hybrid plants as <strong>of</strong> 2004, determined by genetic analysis, in restored tidal marshes opened to tidalaction. Question marks indicate uncertainty in opening dates and subsequent invasion by hybrid cordgrass.Marsh Year Opened Acreage % hybrid plantsCogswell 1980 250 94Oro Loma 1997 364 73Arrowhead 1998 70 100Eden Landing (*Fig. 2) 2004? 836 ??South Bay Salt Ponds ?? 15,100 ??- 126 -

CHAPTER THREEEcosystem Effects <strong>of</strong>Invasive Spartina

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaASSESSMENT OF THE POTENTIAL CONSEQUENCES OF LARGE-SCALE ERADICATION OFSPARTINA ANGLICA FROM THE TAMAR ESTUARY, TASMANIAM. SHEEHAN 1 AND J.C. ELLISON 21Department <strong>of</strong> Primary Industries, PO Box 3100, Bendigo DC, Victoria, Australia; mat<strong>the</strong>w.sheehan@dpi.vic.gov.au2School <strong>of</strong> Geography and Environmental Studies, University <strong>of</strong> Tasmania, Locked bag 1376, Launceston, Tasmania;Joanna.Ellison@utas.edu.auSpartina anglica is a vigorous exotic perennial salt marsh grass typically inhabiting <strong>the</strong> upperintertidal zone <strong>of</strong> temperate estuaries. Following its introduction into <strong>the</strong> Tamar Estuary, Tasmania,patterns <strong>of</strong> sediment deposition and erosion have been significantly altered. This paper presents <strong>the</strong>background to an interdisciplinary approach used to determine <strong>the</strong> potential impacts <strong>of</strong> <strong>the</strong> widescale eradication <strong>of</strong> S. anglica, from what is now Australia’s largest infestation. Analysis <strong>of</strong> <strong>the</strong>seinterdisciplinary lines <strong>of</strong> inquiry have provided a greater understanding <strong>of</strong> <strong>the</strong> biogeomorphologicalresponses to restoration attempts within intertidal zones and have provided a sound basis on whichto formulate and implement future management <strong>of</strong> Spartina.Keywords: Spartina anglica, Estuaries, salt marsh, intertidal zone, geomorphology, sward, sedimentation,erosion, accretion, eradication, surveyingINTRODUCTIONSpartina anglica, (rice grass or cordgrass) is a vigorousexotic perennial salt marsh grass typically inhabiting<strong>the</strong> upper intertidal zone <strong>of</strong> temperate estuaries. Spartinaanglica influences marsh development through <strong>the</strong> ability <strong>of</strong><strong>the</strong> canopy to promote sediment deposition (Shi et al. 1995),and <strong>the</strong> ability <strong>of</strong> <strong>the</strong> dense rhizomes and roots to increase<strong>the</strong> erosion resistance <strong>of</strong> <strong>the</strong> substrate (Brown 1998a; Brownet al. 1998b; Van Eerdt 1985). These properties, along withits ability to establish and spread rapidly are largely <strong>the</strong>reason for its deliberate introduction to temperate estuariesthroughout <strong>the</strong> world (Chapman 1960; Lee and Partridge1983; Ranwell 1967).Accelerated accretion rates following <strong>the</strong> establishment<strong>of</strong> S. anglica have been observed in <strong>the</strong> Ne<strong>the</strong>rlands (VanEerdt 1985), China (Chung 1990), New Zealand (Lee andPartridge 1983) and in Poole Harbour, United Kingdom(Long et al. 1999). However, compaction and settling havebeen found to be important components <strong>of</strong> sedimentationthat may, in some instances result in negligible change intopographic height <strong>of</strong> mudbanks (Lee and Partridge 1983).Fur<strong>the</strong>rmore, studies <strong>of</strong> Louisiana marshes suggest that rootgrowth and below-ground accumulation <strong>of</strong> organic matter,ra<strong>the</strong>r than inorganic matter governs <strong>the</strong> maintenance <strong>of</strong> saltmarshes in <strong>the</strong> vertical plane (Hatton 1983; Nyman 1995;Turner 2004).This paper provides <strong>the</strong> background to an interdisciplinaryand precautionary study that was conducted between 2002and 2007 as part <strong>of</strong> a PhD research project aimed at assessing<strong>the</strong> impacts <strong>of</strong> <strong>the</strong> wide scale eradication <strong>of</strong> S. anglica from<strong>the</strong> Tamar Estuary, Tasmania. Methodology and results <strong>of</strong>this study have been reported elsewhere (Sheehan & Ellison,2007; Sheehan, 2008).BackgroundThe Tamar Estuary (Fig. 1) extends some 71 kilometers(km) inland from Bass Strait through to its tidal extent at<strong>the</strong> city <strong>of</strong> Launceston. The Tamar is a semi-diurnal mesotidalsystem with tides ranging from 1.3 meters (m) at LowHead, to 4 m at Launceston. The Tamar is well supplied withsediment from <strong>the</strong> South and North Esk Rivers, which, in <strong>the</strong>narrow and poorly flushed estuary, tend to accumulate as finegrainedsilt deposits in <strong>the</strong> upper reaches <strong>of</strong> <strong>the</strong> system bothwithin <strong>the</strong> channel and on adjacent mudflats and shoals. Theaverage rate <strong>of</strong> siltation has been estimated at 30,000 cubicmeters (m 3 ) per year, though this may vary between 10,000and 90,000 depending on variation in river flow (Foster etal. 1986). Sedimentation in <strong>the</strong> Tamar’s upper reaches hasbeen an issue <strong>of</strong> long standing concern, both for reasons <strong>of</strong>shipping channel maintenance and environmental quality.Introduction <strong>of</strong> SpartinaSpartina anglica was introduced in 1947 at Windermere(Fig. 1) in an attempt to stabilize sediments and safeguardagainst fur<strong>the</strong>r siltation <strong>of</strong> <strong>the</strong> shipping channel at a timewhen commercial water traffic travelled <strong>the</strong> length <strong>of</strong> <strong>the</strong>estuary (Phillips 1975; Pringle 1993; Ranwell 1967). Mudflatsand shoals in <strong>the</strong> upper Tamar were prone to severe erosionduring <strong>the</strong> combination <strong>of</strong> flood tide and north-west winds.It was thought that stabilizing <strong>the</strong> mudflats would promotevertical accretion, defining <strong>the</strong> channel to enhance scour, andhence reducing <strong>the</strong> likelihood <strong>of</strong> fur<strong>the</strong>r siltation.In 1997, surveys showed that S. anglica covered some420 hectares (ha) (4.2 km 2 ) <strong>of</strong> intertidal zone within <strong>the</strong>- 129 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaFig. 2: Physiographic units <strong>of</strong> <strong>the</strong> Tamar Estuary intertidalzone.Fig. 1: Tamar Estuary, Tasmania.Tamar; however, recent surveys indicate that it continues tospread throughout <strong>the</strong> lower estuary (Hedge 1998).Although <strong>the</strong> introduction <strong>of</strong> S. anglica has improvednavigability in <strong>the</strong> Tamar Estuary (Pringle 1993; Wells 1995),it is now considered a pest species as <strong>the</strong> current generation<strong>of</strong> natural resource managers and land-holders consider itsprogressive invasion to be a major threat to <strong>the</strong> integrity <strong>of</strong>inter-tidal coastal ecosystems and wetlands <strong>of</strong> internationalimportance (Doody 1990; Wells 1995).The present management strategy for S. anglica within<strong>the</strong> Tamar is one <strong>of</strong> containment ra<strong>the</strong>r than eradication dueto <strong>the</strong> complex nature <strong>of</strong> <strong>the</strong> Estuary as outline above. Thereis, however, community and industry pressure to eradicate S.anglica from <strong>the</strong> banks <strong>of</strong> <strong>the</strong> Tamar.GEOMORPHOLOGY OF THE INTERTIDAL ZONEThe intertidal zone within <strong>the</strong> Tamar consists <strong>of</strong> fourdistinct physiographic units (Fig. 2):1) Narrow mudflats <strong>of</strong> recent alluvial deposits largely <strong>of</strong>terrestrial origin entering <strong>the</strong> system via <strong>the</strong> North Esk and <strong>the</strong>South Esk Rivers. These have accumulated predominantly in<strong>the</strong> upper estuary, however this unit can also be found in <strong>the</strong>mid and lower intertidal zones <strong>of</strong> <strong>the</strong> mid estuary or in largerembayments that receive sediment from minor tributaries;2) A laterally extensive area <strong>of</strong> outcropping substratewhich has wea<strong>the</strong>red in situ to form a narrow boulder beachthroughout <strong>the</strong> mid and some parts <strong>of</strong> <strong>the</strong> lower Tamar. Thisformation consists largely <strong>of</strong> Dolerite but also some Miocenebasalt shorelines at Whirlpool reach;3) Extensive beaches within <strong>the</strong> mid Tamar consisting<strong>of</strong> a thin veneer <strong>of</strong> Tertiary sands and gravels underlain byclays. The clays are likely to be lacustrine deposits <strong>of</strong> <strong>the</strong>Miocene while <strong>the</strong> sand and gravel are indicative <strong>of</strong> highenergyfluvial environments, which would have been presentduring <strong>the</strong> Pleistocene; and- 130 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaFig. 3: Native salt marsh communities <strong>of</strong> <strong>the</strong> lower salt marsh being progressivelyinvaded by S. anglica clones, establishing seaward <strong>of</strong> <strong>the</strong> nativevegetation.Fig. 6: Spartina anglica colonization <strong>of</strong> a dolerite boulder beach at EastArm.Source: T. Colesa) b)Fig. 4: The intertidal zone <strong>of</strong> <strong>the</strong> Tamar at Paper Beach in 1956 (a), showing<strong>the</strong> sand and gravel substrate, and in 2004 (b) Spartina marsh extendingsome 140 m seaward.Fig. 7: Mature Spartina marsh at Lone Pine point, previously a DoleriteBoulder beach.Fig. 5: One <strong>of</strong> <strong>the</strong> few remaining sandy intertidal zones <strong>of</strong> Tamar, <strong>the</strong> upperintertidal zone is typically sand and gravels grading into mudflat.4) Marine sands, limited to <strong>the</strong> lower estuary close to <strong>the</strong>mouth at Low head.VEGETATION OF THE INTERTIDAL ZONEPrior to <strong>the</strong> establishment <strong>of</strong> S. anglica, much <strong>of</strong><strong>the</strong> intertidal zone, particularly in <strong>the</strong> mid Tamar, wasunvegetated. Enteromorpha and o<strong>the</strong>r algae were foundextensively on mudflats at all elevations within <strong>the</strong> intertidalzone throughout <strong>the</strong> estuary. The absence <strong>of</strong> highernative salt marsh plants is probably due to <strong>the</strong>ir inabilityto colonize <strong>the</strong> various intertidal geologies and promote<strong>the</strong> accumulation <strong>of</strong> fine silts. Native salt marsh vegetation,such as Sarcocornia quinqueflora, Sclerostegia arbuscula,and Suaeda australis is limited to a narrow fringe belowFig. 8: A prograding Spartina marsh, with <strong>the</strong> coalescence <strong>of</strong> isolatedclumps characteristic <strong>of</strong> Spartina colonization.<strong>the</strong> high water mark and in sheltered embayments, wi<strong>the</strong>xtensive salt marshes occurring only near Bell Bay, onrelatively sandy substrates in <strong>the</strong> lower estuary (Fig. 3). Theability <strong>of</strong> S. anglica to establish at elevations lower thannative salt marsh plants provides a valuable competitiveadvantage. Spartina anglica has successfully colonizedin isolated clumps, coalesced to form laterally extensiveswards seaward <strong>of</strong> <strong>the</strong> native salt marsh, and progressivelymove landward into <strong>the</strong> native vegetation.- 131 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaIMPACTS OF SPARTINA INVASIONThe introduction <strong>of</strong> S. anglica to <strong>the</strong> Tamar Estuary hasbrought about a dramatic and rapid change to <strong>the</strong> physiography<strong>of</strong> <strong>the</strong> intertidal zone, illustrated well at a photo pointestablished in 1956 at Paper beach (Fig. 4a and b). The colonization<strong>of</strong> what was essentially a vacant niche has transformed<strong>the</strong> gently grading sandy intertidal zones, as shown in Fig. 5,and hard rock intertidal zones (Fig. 6) into laterally extensivemuddy terraces (Figs. 7 and 8).The ecological impacts <strong>of</strong> Spartina invasion in temperateestuaries <strong>of</strong> South Eastern Australia are poorly understood.Studies <strong>of</strong> benthic macro-invertebrate communities in <strong>the</strong>Little Swanport Estuary, Tasmania suggest that Spartinasignificantly increases macro-invertebrate species richnessand total species abundance when compared to previouslynon-vegetated intertidal areas (Hedge 1997). Additionally,macro-invertebrate communities <strong>of</strong> Spartina marshesshowed remarkable similarity to those <strong>of</strong> native salt marshes(Hedge and Kriwoken 2000). It is suggested <strong>the</strong>refore thatSpartina invasion has provided a niche <strong>of</strong> s<strong>of</strong>t substrate anddense protection, favoring macro-invertebrates and somebirds such as <strong>the</strong> purple swamphen (Porphyrio porphyrio).It is likely that this subsequently caused a displacement <strong>of</strong>species that formerly inhabited or utilized <strong>the</strong> intertidal zone.Species assemblage and utilization <strong>of</strong> <strong>the</strong> intertidal zone priorto Spartina is not well documented, but likely was relativelyrich in fish and bird species.DISCUSSIONThis paper has provided background for <strong>the</strong> researchproject that assessed <strong>the</strong> alteration to <strong>the</strong> intertidal zone <strong>of</strong> <strong>the</strong>Tamar Estuary. Using transect based topographic surveys andcoring, <strong>the</strong> volume <strong>of</strong> material trapped under Spartina wascalculated to be 1,193,441 m 3 , comprised <strong>of</strong> approximately17% Spartina-derived organic matter and 83% silts andclays. Based on historical pr<strong>of</strong>iles, sedimentation rates since<strong>the</strong> introduction <strong>of</strong> S. anglica have been estimated at between8.7 and 52.4 millimeters per year (mm yr-1).From <strong>the</strong> analysis <strong>of</strong> 80 cores from four sites, Spartinatrappedsediment was found to contain levels <strong>of</strong> cadmium,copper, lead and zinc elevated above background levels.However, <strong>the</strong>se generaly were below trigger values <strong>of</strong> <strong>the</strong>ANZECC/ARMCANZ (2000) interim sediment qualityguidelines. It is considered unlikely that released sedimentswould impact on water quality or health <strong>of</strong> biota with respectto trace metals or organic contamination.Behaviour <strong>of</strong> sediment with respect to erosion rates,sediment redeposition and causative hydrodynamics werealso monitored within a test area from which Spartina coverwas removed. It has been demonstrated that <strong>the</strong> eradication<strong>of</strong> S. anglica will result in elevation loss from <strong>the</strong> Spartinamarsh surface at a rate six times greater than in vegetatedmarshes. The study assumed that this elevation loss is causedby liberation <strong>of</strong> sediments. The rate <strong>of</strong> elevation loss orerosion is likely to increase once <strong>the</strong> dead S. anglica root matdecomposes and <strong>the</strong> surface cohesion and sediment-bindingcapacity is diminished. Elevation loss increases by a factor <strong>of</strong>1.06 with every 10 m from <strong>the</strong> high water bank. Erosion ratesin <strong>the</strong> outer 40 m <strong>of</strong> <strong>the</strong> marsh were also significantly greaterthan <strong>the</strong> remainder <strong>of</strong> <strong>the</strong> marsh at both sites, suggesting thata process o<strong>the</strong>r than S. anglica removal is contributing to thatretreat <strong>of</strong> <strong>the</strong> lower marsh.Analysis <strong>of</strong> <strong>the</strong>se interdisciplinary lines <strong>of</strong> inquiryhave allowed for a greater understanding <strong>of</strong> <strong>the</strong>biogeomorphological responses to restoration attempts withinintertidal zones and have provided a sound basis on which t<strong>of</strong>ormulate and implement future management <strong>of</strong> Spartina.Spartina eradication is recommended for <strong>the</strong> lower estuary(type 2 marshes) only, where trapped sediment volumesare significantly smaller and tidal flushing is greatest. Thiswould enable significant areas <strong>of</strong> sand/gravel intertidal zonesto recover and reduce <strong>the</strong> likelihood <strong>of</strong> fur<strong>the</strong>r downstreamexpansion <strong>of</strong> Spartina swards. Retaining type 1 marshes in<strong>the</strong> upper estuary will prevent <strong>the</strong> remobilization <strong>of</strong> sedimentsthat contain <strong>the</strong> highest concentrations <strong>of</strong> contaminants, andwill retain <strong>the</strong> marshes for <strong>the</strong> ecological role <strong>the</strong>y currentlyperform.ACKNOWLEDGMENTSThis study was funded by <strong>the</strong> Australian ResearchCouncil Linkage Grant LP0214145, with support from <strong>the</strong>Department <strong>of</strong> Primary Industries and Water, Tasmania and<strong>the</strong> Rice Grass Advisory Group.REFERENCESANZECC/ARMCANZ, 2000. Australian and New Zealand guidelinesfor fresh and marine water quality. Australian and NewZealand Environment and Conservation Council/ Agricultureand Resource Management Council <strong>of</strong> Australia and NewZealand, Canberra, ACT.Brown, S. 1998. Sedimentation on a Humber salt marsh. In: Cramp,A. ed., Sedimentary Processes in <strong>the</strong> Intertidal Zone. GeologicalSociety, London.Brown, S., E.A. Warman, S. McGrorty, M. Yates, R.J. Pakeman,L.A. Boorman, J.D. Goss-Custard and A.J. Gray. 1998. SedimentFluxes in Intertidal Biotopes: BIOTA II. Marine PollutionBulletin 37(3-7):173-181.Chapman, V.J., 1960. Salt marshes and salt deserts <strong>of</strong> <strong>the</strong> world.Leonard Hill, London and New York.Chung, C., 1990. Twenty-five years <strong>of</strong> introduced Spartina anglicain China. In: Gray, A.J. and P.E.M. Benham, eds. Spartinaanglica: a research review, ITE research publication. NaturalEnvironment Research Council, Institute <strong>of</strong> Terrestrial Ecology,London.Doody, J.P., 1990. Spartina – friend or foe? A conservation viewpoint. In: Gray, A.J. and P.E.M. Benham, eds. Spartina anglica: aresearch review, ITE research publication. Natural EnvironmentResearch Council, Institute <strong>of</strong> Terrestrial Ecology, London.Foster, D.N., R. Nittim J. Walker. 1986. Tamar River siltation study.Technical Report, No. 85/07, University <strong>of</strong> NSW Water ResearchLaboratory.Hatton, R.S., R.D. Delaune and W.H. Patrick, Jr. 1983.Sedimentation, accretion, and subsidence in marshes <strong>of</strong> BaratariaBasin, Louisiana. Limnology and Oceanography 28: 494-502.- 132 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaHedge, P., 1997. Impact <strong>of</strong> Spartina anglica to intertidal faunain Little Swanport Estuary, Tasmania. Graduate Diploma<strong>of</strong> Environmental studies (Honours) Thesis, Centre forEnvironmental Studies, University <strong>of</strong> Tasmania, Hobart.Hedge, P., 1998. Strategy for <strong>the</strong> management <strong>of</strong> Rice Grass(Spartina anglica) in Tasmania, Australia., Department <strong>of</strong>Primary Industries, Water and Environment, Tasmania,Australia.Hedge, P. and L.K. Kriwoken. 2000. Evidence for <strong>the</strong> effects <strong>of</strong>Spartina anglica invasion on benthic macr<strong>of</strong>auna in LittleSwanport Estuary, Tasmania. Austral Ecology 25: 150-159.Lee, W.G. and T.R. Partridge. 1983. Rates <strong>of</strong> spread <strong>of</strong> Spartinaanglica and sediment accretion in <strong>the</strong> New River Estuary,Invercargill, New Zealand. New Zealand Journal <strong>of</strong> Botany21:231-236.Long, A.J., R.G. Scaife and R.J. Edwards. 1999. Pine pollen inintertidal sediments from Poole Harbour, UK: Implications forlate Holocene accretion rates and sea-level rise. Quaternary<strong>International</strong> 55:3-16.Nyman, J.A., R.D. Delaune, S.R. Pezeshki, W.H. Patrick. 1995.Organic matter fluxes and marsh stability in a rapidly submergingestuarine marsh. Estuaries 18(1B):207-218.Phillips, A.W., 1975. The establishment <strong>of</strong> Spartina in <strong>the</strong> TamarEstuary, Tasmania. Papers and <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> Royal Society<strong>of</strong> Tasmania 109:65-75.Pringle, A., 1993. Spartina anglica ation and physical effects in<strong>the</strong> Tamar Estuary, Tasmania 1971-91. Papers and <strong>Proceedings</strong><strong>of</strong> <strong>the</strong> Royal Society <strong>of</strong> Tasmania 127:1-10.Ranwell, S., 1967. World resources <strong>of</strong> Spartina Townsenii (sensulato) and economic use <strong>of</strong> Spartina marshland. Journal <strong>of</strong> AppliedEcology 4:239-256.Sheehan, M.R. and J.C. Ellison. 2007. Baseline survey <strong>of</strong> TamarEstuary shoreline banks Report for NRM North, Launceston.Sheehan, M.R. 2008. Assessment <strong>of</strong> <strong>the</strong> potential impacts <strong>of</strong> largescaleeradication <strong>of</strong> Spartina anglica from <strong>the</strong> Tamar Estuary,Tasmania. Unpublished PhD <strong>the</strong>sis, University <strong>of</strong> Tasmania,Launceston.Shi, Z., J.S. Pethick, and K. Pye. 1995. Flow structure in <strong>the</strong>various heights <strong>of</strong> a Salt marsh canopy: A Laboratory FlumeStudy. Journal <strong>of</strong> Coastal Research 11(4):1204-1209.Turner, R.E., E.M. Swenson, C.S. Milan, J.M. Lee, T.A. Oswald.2004. Below-ground biomass in healthy and impaired saltmarshes. Ecological Research 19(1):29-35.Van Eerdt, M., 1985. The influence <strong>of</strong> vegetation on erosion andaccretion in salt marshes <strong>of</strong> <strong>the</strong> Oosterschelde, The Ne<strong>the</strong>rlands.Vegetation 62:362-373.Wells, A., 1995. Rice grass in Tasmania - an overview. In: Rash,J.E., R.C. Williamson and S.J. Taylor, eds. <strong>Proceedings</strong> <strong>of</strong><strong>the</strong> Australasian <strong>Conference</strong> on Spartina control. VictorianGovernment Publication, Melbourne, Australia.- 133 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaSPARTINA INVASION CHANGES INTERTIDAL ECOSYSTEM METABOLISM IN SANFRANCISCO BAYA.C. TYLER 1,2 AND E.D. GROSHOLZ 11 Department <strong>of</strong> Environmental Science and Policy, University <strong>of</strong> California, Davis, CA 95616;annachristinatyler@gmail.com2 Current address: School <strong>of</strong> Biological and Medical Sciences, Rochester Institute <strong>of</strong> Technology, Rochester, NY 14623In San Francisco Bay, Atlantic smooth cordgrass, Spartina alterniflora, and its hybrids haveinvaded unvegetated mudflats and native marshes formerly dominated by Sarcocorniapacifica and Spartina foliosa. This recent, rapid invasion has dramatically changedecosystem processes and food web structure. We measured sediment fluxes <strong>of</strong> carbondioxide (CO 2 ) to determine microalgal gross primary production (GPP), sediment respirationand net sediment metabolism in native intertidal areas (mudflats, S. pacifica and S. foliosa)and in adjacent areas invaded by hybrid Spartina. Sediment microalgal GPP and microalgalchlorophyll a were substantially higher in all native habitats than in <strong>the</strong> hybrid. At <strong>the</strong> sametime, sediment respiration rates were generally higher in native vegetation than in hybriddominatedhabitats, but substantially lower on <strong>the</strong> mudflats than elsewhere. Net sedimentmetabolism switched from autotrophy to heterotrophy following invasion <strong>of</strong> mudflats.However, <strong>the</strong> higher respiration rate in native vegetation, especially S. pacifica, relative tohybrid areas, suggests slower decomposition <strong>of</strong> hybrid detritus. This result is corroboratedby litterbag decomposition rates and indicates a build-up and/or export <strong>of</strong> refractory organicmatter following hybrid invasion. The switch from a microalgal-dominated system to arefractory detritus-dominated system has clear implications for support <strong>of</strong> higher trophiclevels within <strong>the</strong> intertidal zone <strong>of</strong> San Francisco Bay.Keywords: ecosystem metabolism, carbon cycling, Spartina alterniflora, Sarcocornia sp.,Spartina foliosa, benthic microalgaeINTRODUCTIONOne <strong>of</strong> <strong>the</strong> most serious threats to natural ecosystemsand <strong>the</strong> maintenance <strong>of</strong> ecosystem services is <strong>the</strong> invasion <strong>of</strong>non-native plant species (Drake et al. 1989; Vitousek et al.1997). These threats may include <strong>the</strong> extinction <strong>of</strong> nativespecies, loss <strong>of</strong> functional native diversity, changes innutrient cycling and organic matter storage and loss <strong>of</strong>habitat. One <strong>of</strong> <strong>the</strong> most dramatic invasions in coastalsystems in western North America has been <strong>the</strong> invasion <strong>of</strong>Spartina alterniflora (smooth cordgrass), a native <strong>of</strong> westernAtlantic marshes. This invasion has fostered large-scalealterations in ecosystem processes in commercially andecologically important estuaries in both California andWashington.S. alterniflora was first introduced into San FranciscoBay by <strong>the</strong> U.S. Army Corps <strong>of</strong> Engineers in 1973 for marshrestoration (Faber 2000). Subsequent hybridization with <strong>the</strong>native cordgrass, Spartina foliosa, resulted in a highlysuccessful hybrid population (henceforth hybrid Spartina)that has colonized ~500 acres, in south and central portions<strong>of</strong> <strong>the</strong> Bay (Daehler and Strong 1997; Ayres et al. 2004).Hybrid Spartina occupies a wide tidal range that includeshistorically unvegetated mudflats and native marshes oncedominated by S. foliosa and Sarcocornia pacifica (Callawayand Josselyn 1992).The Spartina invasion has greatly altered <strong>the</strong> cycling <strong>of</strong>carbon (C) in <strong>the</strong> intertidal zone. Hybrid Spartina formsdense clones, with an average aboveground biomass <strong>of</strong> 1.8 ±0.3 kilograms per square meter (kg m -2 ) at <strong>the</strong> end <strong>of</strong> <strong>the</strong>growing season (fall; Tyler et al. 2007). Much <strong>of</strong> <strong>the</strong>aboveground portion senesces during <strong>the</strong> winter, but <strong>the</strong>substantial belowground roots and rhizomes (mean = 4.2 ±0.8 kg m -2 ) persist year-round (Tyler et al. 2007). Thecarbon to nitrogen (C:N) ratio <strong>of</strong> Spartina is higher than <strong>the</strong>native vegetation, particularly when compared tomicroalgae, and <strong>the</strong> slow decomposition <strong>of</strong> this refractorydetritus has resulted in substantial deposits <strong>of</strong> belowgroundorganic matter (Neira et al. 2005). We do not know howoverall ecosystem production and respiration are changed asa result <strong>of</strong> this invasion, but this is clearly important inunderstanding <strong>the</strong> overall impact <strong>of</strong> invasion on estuarineecosystem function.The lower productivity and low stature <strong>of</strong> nativesou<strong>the</strong>rn California salt marsh plants, which results in morelight reaching <strong>the</strong> sediment surface, may promote microalgal- 135 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaproductivity that is generally greater than that <strong>of</strong> <strong>the</strong> vascularplants (Zedler 1984). In contrast, primary production inAtlantic and Gulf coast marshes is generally dominated by S.alterniflora (Zedler 1980). Studies <strong>of</strong> salt marsh food webson both <strong>the</strong> Atlantic and Pacific coasts show that microalgaemay contribute up to 50% <strong>of</strong> <strong>the</strong> C assimilated byinvertebrates and higher trophic levels (Page 1995; Deeganand Garritt 1997; Kwak and Zedler 1997; Page 1997). Theimpact <strong>of</strong> Spartina invasion on microalgal productivitywithin vegetated marshes <strong>of</strong> San Francisco Bay is not wellunderstood. However, because algae are more readilyassimilated than detrital Spartina, <strong>the</strong> invasion <strong>of</strong> Spartinamay ultimately reduce <strong>the</strong> availability <strong>of</strong> primary productionto higher trophic levels (i.e., consumers).The results presented here are a preliminary analysis <strong>of</strong><strong>the</strong> impact <strong>of</strong> invasive Spartina on microalgal production,sediment respiration and net sediment metabolism. Futureanalysis <strong>of</strong> <strong>the</strong>se data, in combination with detailed estimates<strong>of</strong> vascular plant production and decomposition rates willultimately give us a more complete understanding <strong>of</strong> howSpartina influences overall ecosystem functioning in <strong>the</strong>intertidal areas <strong>of</strong> San Francisco Bay.METHODSSediment CO 2 fluxes and benthic chlorophyll a (Chl a)were measured in native- and hybrid-dominated areas at foursites in South San Francisco Bay in December 2003 andMarch, June and September 2004 (Fig. 1). Two sites,Cogswell marsh (Hayward Regional Shoreline) and SanMateo marsh, represent areas formerly dominated bypickleweed, S. pacifica. The Elsie Roemer Bird Sanctuaryon Alameda Island and <strong>the</strong> San Lorenzo marsh at Robert’sLanding are mudflats currently being invaded by hybridSpartina. Also at San Lorenzo, areas <strong>of</strong> native S. foliosa areoverrun with hybrid Spartina.Sediment CO 2 fluxes were measured in polycarbonatecores as described in (Neubauer et al. 2000). Briefly,polycarbonate cores (9.3 centimeters inside diameter [cmI.D.]) were inserted 5 cm into <strong>the</strong> sediment (headspaceapproximately 800 cubic centimeters [cm 3 ]). Incurrent andexcurrent tubes were fitted into <strong>the</strong> sealed lid and connectedto a LiCor LI820 Infrared CO 2 Gas Analyzer. Air flow wasmaintained at approximately 500 millileters per minute (mlmin -1 ) using a small electric air pump. Prior to eachsampling period <strong>the</strong> Gas Analyzer was calibrated using CO 2standards (Scott Specialty Gases, 0 parts per million [ppm]and 1,007 ppm CO 2 ). All measurements were conductedbetween approximately 10 am and 2 pm during low tide.Each core was darkened using plastic pots coated withaluminum foil for a minimum <strong>of</strong> 30 minutes prior to makingmeasurements. We recorded CO 2 concentrations in <strong>the</strong> coreevery two seconds for 4-6 minutes in <strong>the</strong> dark and <strong>the</strong>nremoved <strong>the</strong> darkening pot. After waiting ano<strong>the</strong>r 4-6minutes, <strong>the</strong> CO 2 concentrations in <strong>the</strong> light were recorded.SBOSMHBAIALARLAOLACOGALCFig. 1. South San Francisco Bay showing sampling sites: Elsie RoemerBird Sanctuary (ALA), Robert’s Landing (RLA), Cogswell (COG) andSan Mateo (SMH).Three replicates were performed in each area at each site.Fluxes were estimated based on <strong>the</strong> slope <strong>of</strong> <strong>the</strong> change inconcentration over time using <strong>the</strong> equation:dC VJ = •dt Awhere J is <strong>the</strong> flux rate in micromoles per square meter perhour (μmol m -2 h -1 ), A is <strong>the</strong> core area, V is <strong>the</strong> headspacevolume, C is <strong>the</strong> concentration <strong>of</strong> CO 2 and t is time.Light at <strong>the</strong> sediment surface and at <strong>the</strong> top <strong>of</strong> <strong>the</strong>canopy was recorded every two minutes during <strong>the</strong> samplingperiod using a Li-Cor model LI1400 meter with a 4πspherical quantum sensor. Sediment temperature adjacent toeach core was measured using an analog soil temperatureprobe. Following each set <strong>of</strong> measurements two small (1.1cm I.D. x 0.5 cm depth) cores were taken for Chl a analysis.These samples were kept frozen until sonication <strong>of</strong> <strong>the</strong>sediment with cold 90% acetone, overnight extraction in <strong>the</strong>freezer and standard spectrophotometric measurement.Hourly gross primary production (GPP) was calculatedfrom <strong>the</strong> difference between CO 2 fluxes in <strong>the</strong> light and <strong>the</strong>dark by assuming that sediment respiration was <strong>the</strong> same in<strong>the</strong> light and <strong>the</strong> dark. Daily GPP was calculated bymultiplying hourly GPP by <strong>the</strong> numbers <strong>of</strong> hours <strong>of</strong> light on<strong>the</strong> day <strong>of</strong> measurement. Net daily sediment metabolism was- 136 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartinacalculated by summing daily GPP and hourly respiration x24. Data were pooled across sampling dates and statisticalcomparisons were made (ANOVA) between GPP, sedimentrespiration, light at <strong>the</strong> sediment surface, and Chl ameasured in hybrid Spartina and native areas for each <strong>of</strong> <strong>the</strong>three native conditions (mudflats, S. foliosa, S. pacifica).Mudflats and S. pacifica were represented by two sites andS. foliosa by a single site (San Lorenzo).RESULTSWe found significant differences for several variablesbetween native and Spartina hybrid-invaded habitats inSouth San Francisco Bay.GPPMicroalgal GPP was greater in native habitats than inhybrid Spartina in all cases (Fig. 2A). This difference wassignificant for <strong>the</strong> mudflat-hybrid comparison and <strong>the</strong> S.pacifica-hybrid comparison, where GPP was more than twotimes greater in S. pacifica.Sediment RespirationHybrid Spartina had significantly greater sedimentrespiration rates than those found on <strong>the</strong> mudflat (Fig. 2B).In contrast, both S. foliosa and S. pacifica had higherrespiration rates, but only significantly so for S. pacifica.Net Sediment MetabolismInvaded mudflats switched from net autotrophic(positive net metabolism) to net heterotrophic (negative netmetabolism) upon invasion by hybrid Spartina (Fig. 2C).The native habitats, S. foliosa and S. pacifica, had higher netmetabolism than <strong>the</strong> hybrid. Again, only S. pacifica wassignificantly different from <strong>the</strong> hybrid.LightSpartina hybrid invasion resulted in a significantreduction in light availability at <strong>the</strong> sediment surface relativeto mudflats (18% <strong>of</strong> available light) and S. foliosa (61% in S.foliosa; 26% in hybrid; Fig. 3A). Light availability wassimilar in S. pacifica and hybrid habitats (16% and 18% <strong>of</strong>available light, respectively).Benthic Chl aNative habitats had higher Chl a than hybrid habitats inall cases, but only significantly so relative to S. pacifica (Fig.3B).DISCUSSIONThe invasion <strong>of</strong> S. alterniflora and its hybrids into <strong>the</strong>native mudflats and marshes <strong>of</strong> San Francisco Bay hasclearly changed both microalgal production and sedimentrespiration. As a result net metabolism <strong>of</strong> <strong>the</strong> sediments isdramatically different than it was prior to invasion, with <strong>the</strong>trajectory <strong>of</strong> change depending on <strong>the</strong> initial conditions.The consistent decrease in microalgal GPP and Chl aupon invasion may be due to a combination <strong>of</strong> factors,including light and nutrient availability and grazing. TheA. MICROALGAL GROSS PRIMARY PRODUCTIONmg C m -2 d -18006004002000MF HYB FOL HYB SAR HYBB. SEDIMENT RESPIRATION0mg C m -2 d -1-500-1000-1500-2000-2500-3000-3500MF HYB FOL HYB SAR HYBC. NET SEDIMENT METABOLISMmg C m -2 d -15000-500-1000-1500-2000-2500MF HYB FOL HYB SAR HYBFig. 2. Microalgal gross primary production (2A), sediment respiration(2B) and net sediment metabolism (2C) measured in mudflats (MF), S.foliosa (FOL), S. pacifica (SAR) and adjacent hybrid areas. Starsrepresent significant differences between areas (p < 0.05) based onANOVA.substantial decrease in available light relative to anunvegetated mudflat is likely important. However, lightavailability is similar between hybrid and S. pacificahabitats, suggesting <strong>the</strong> importance <strong>of</strong> ano<strong>the</strong>r mechanism.The current estimate <strong>of</strong> GPP is likely an overestimate <strong>of</strong>actual GPP because our measurements were made at <strong>the</strong>time <strong>of</strong> peak solar insolation and at low tide. Futurerefinement <strong>of</strong> <strong>the</strong>se results will include an adjustment <strong>of</strong>GPP based on photosyn<strong>the</strong>sis-irradiance curves for eachhabitat type and <strong>the</strong> reduction in photosyn<strong>the</strong>sis observed- 137 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaA. LIGHTmmol photons m -2 s -1B. BENTHIC CHLOROPHYLL amg Chl a m -225002000150010005000250200150100500MF HYB FOL HYB SAR HYBMF HYB FOL HYB SAR HYBFig. 3. Light measured at <strong>the</strong> sediment surface (3A) and benthicchlorophyll a (3B) measured in mudflats (MF), S. foliosa (FOL), S.pacifica (SAR) and adjacent hybrid areas. Stars represent significantdifferences between areas (p < 0.05) based on ANOVA.during flooding, which can be 25–49% (Holmes and Mahall1982; Pinckney and Zingmark 1993). In spite <strong>of</strong> <strong>the</strong>limitations <strong>of</strong> our current GPP estimates, <strong>the</strong>re is a clear andconsistent decline in microalgal abundance. This decline hasvery important ramifications for higher trophic levels.Infaunal invertebrate abundance and diversity declined 75%upon invasion <strong>of</strong> mudflats at <strong>the</strong> Alameda site (Neira et al.2005). While <strong>the</strong>re are a number <strong>of</strong> factors that maycontribute to this loss, including a lack <strong>of</strong> available space,slower water velocities, and predation (Neira et al. 2005),<strong>the</strong> availability <strong>of</strong> microalgae as a food source is also a keycontributor (Levin et al. 2006; Grosholz et al. 2009).The massive increase in belowground biomass andorganic matter that occurs when mudflats are invaded bySpartina is <strong>the</strong> likely cause <strong>of</strong> higher respiration rates inhybrid habitats. Both microbial and root respiration shouldbe higher in invaded areas. However, higher respirationrates in native vegetation relative to hybrid areas are moredifficult to interpret. We did not observe an increase inbelowground biomass in hybrid habitats relative to S. foliosaor S. pacifica. Therefore, root respiration is likely to besimilar among <strong>the</strong>se three habitats. However, in separatelitterbag experiments, we found that both S. foliosa and S.pacifica decompose more rapidly than hybrid Spartina(Tyler, unpub. data). This is consistent with <strong>the</strong> higher C:Nratio <strong>of</strong> hybrid Spartina relative to native vegetation.Because <strong>the</strong> hybrid detritus is refractory and <strong>the</strong>reby a poorqualityfood source for microbes and invertebrates, wewould expect <strong>the</strong> low sediment respiration rates that we didindeed observe. When decomposition rates are low, organicmatter will remain intact and will build up in <strong>the</strong> system overtime. The build up <strong>of</strong> this detritus is also supported by <strong>the</strong>finding that Spartina detritus does not contribute to highertrophic levels as readily as native plant detritus (Brusati2004; Brusati and Grosholz 2007, 2008).The difference in net sediment metabolism <strong>of</strong> hybridsediments relative to S. foliosa and S. pacifica is drivenlargely by <strong>the</strong> high respiration rates found in native habitats.Hybrid invasion acts to decrease <strong>the</strong> overall metabolic rateand create an excess inventory <strong>of</strong> organic matter. Themudflats, however, switched from net autotrophy to ne<strong>the</strong>terotrophy upon invasion, due to higher respiration andlower microalgal photosyn<strong>the</strong>sis. In this case, <strong>the</strong> invasionacts to eliminate a valuable food source (microalgae) andreplace it with a poor quality food source (Spartina detritus).Future refinements to this model <strong>of</strong> ecosystemmetabolism following Spartina hybrid invasion will includevascular plant production, above-ground decomposition <strong>of</strong>vascular plant detritus and <strong>the</strong> adjustments to microalgalGPP discussed above. From this, we will be able toestimate, on an annual basis, <strong>the</strong> very significant impact <strong>of</strong>invasive Spartina on ecosystem function in San FranciscoBay. At this point, we can conclude that this invasion hasresulted in <strong>the</strong> replacement <strong>of</strong> mudflats and marshescontaining abundant microalgae and bioavailable detritus bya habitat with lower microalgal productivity and refractorydetritus. This change has very important ramifications forhigher trophic levels and for <strong>the</strong> health <strong>of</strong> <strong>the</strong> estuary as awhole.ACKNOWLEDGMENTSFunding for this work came from a NSF BiocomplexityGrant (DEB-0083583). We are particularly grateful to U.Mahl, R. Blake, C. Sorte, N. Christensen and C. Love forfield and laboratory assistance and to <strong>the</strong> C. Goldman lab for<strong>the</strong> use <strong>of</strong> lab facilities.REFERENCESAyres, D.R., D.L. Smith, K. Zaremba, S. Klohr, and D.R. Strong.2004. Spread <strong>of</strong> exotic cordgrasses and hybrids (Spartina sp.) in<strong>the</strong> tidal marshes <strong>of</strong> San Francisco Bay, California, USA.Biological Invasions 6:221-231.Brusati, E.D. 2004. Effects <strong>of</strong> native and hybrid cordgrass onbenthic invertebrate communities and food webs. Ph.D.Dissertation, University <strong>of</strong> California, Davis, Davis, California,USA.Brusati, E.D., and E.D. Grosholz. 2007. Effect <strong>of</strong> native andinvasive cordgrass on Macoma petalum density, growth, and- 138 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartinaisotopic signatures. Estuarine, Coastal and Shelf Science 71:517-222.Brusati, E. D., and E. D. Grosholz. 2008. Does invasion <strong>of</strong> hybridcordgrass change estuarine food webs? Biological Invasions11:917-926.Callaway, J.C., and M.N. Josselyn. 1992. The introduction andspread <strong>of</strong> smooth cordgrass (Spartina alterniflora) in South SanFrancisco Bay. Estuaries 15:218-226.Daehler, C.C., and D.R. Strong. 1997. Hybridization betweenintroduced smooth cordgrass (Spartina alterniflora, Poaceae)and native California cordgrass (Spartina foliosa) in SanFrancisco Bay, California, USA. American Journal <strong>of</strong> Botany85:607-611.Deegan, L.A., and R.H. Garritt. 1997. Evidence for spatialvariability in estuarine food webs. Marine Ecology ProgressSeries 147:31-47.Drake, J.A., H.A. Mooney, F. DiCastri, R.H. Groves, F.J. Kruger,M. Rejmanek, and M. Williamson. 1989. Biological Invasions:A Global Perspective. John Wiley & Sons, Inc., Chichester, UK.Faber, P.M. 2000. Grass Wars: Good intentions gone awry.California Coast and Ocean 16:14-17.Grosholz, E.D., L.A. Levin, A.C. Tyler, and C. Neira. 2009.Changes in community structure and ecosystem functionfollowing Spartina alterniflora invasion <strong>of</strong> Pacific estuaries. In:Silliman, B.R., E.D. Grosholz, and M.D. Bertness, eds. HumanImpacts on Salt Marshes: A Global Perspective. University <strong>of</strong>California Press, Berkeley, California, USA.Holmes, R., and B. Mahall. 1982. Preliminary observations on <strong>the</strong>effects <strong>of</strong> flooding and dessication upon <strong>the</strong> net photosyn<strong>the</strong>ticrates <strong>of</strong> high intertidal estuarine sediments. Limnology andOceanography 27:954-958.Kwak, T.J., and J.B. Zedler. 1997. Food web analysis <strong>of</strong> sou<strong>the</strong>rnCalfiornia coastal wetlands using multiple stable isotopes.Oecologia 110:262-277.Levin, L.A., C. Neira, and E.D. Grosholz. 2006. Invasive cordgrassmodifies wetland trophic function. Ecology 87:419-432.Neira, C., L.A. Levin, and E.D. Grosholz. 2005. Benthicmacr<strong>of</strong>aunal communities <strong>of</strong> three sites in San Francisco Bayinvaded by hybrid Spartina, with comparison to uninvadedhabitats. Marine Ecology Progress Series 292:111-126.Neubauer, S., W. Miller, and I. Anderson. 2000. Carbon cycling ina tidal freshwater marsh ecosystem: a carbon gas flux study.Marine Ecology Progress Series 199:13-20.Page, H.M. 1995. Variation in <strong>the</strong> natural abundance <strong>of</strong> 15 N in <strong>the</strong>halophyte, Salicornia virginica, associated with groundwatersubsidies <strong>of</strong> nitrogen in a sou<strong>the</strong>rn California salt-marsh.Oecologia 104:181-188.Page, H.M. 1997. Importance <strong>of</strong> vascular plant and algalproduction to macro-invertebrate consumers in a sou<strong>the</strong>rnCalifornia salt marsh. Estuarine, Coastal and Shelf Science45:823-834.Pinckney, J.L., and R.G. Zingmark. 1993. Modeling <strong>the</strong> annualproduction <strong>of</strong> intertidal benthic microalgae in estuarineecosystems. Journal <strong>of</strong> Phycology 29:396-407.Tyler, A.C., J.G. Lambrinos, and E.D. Grosholz. 2007. Nitrogeninputs promote <strong>the</strong> spread <strong>of</strong> an invasive marsh grass. EcologicalApplications 17:1886-1898.Vitousek, P.M., H.A. Mooney, J. Lubchenco, and J.M. Melillo.1997. Human domination <strong>of</strong> Earth's ecosystems. Science277:494-499.Zedler, J.B. 1980. Algal mat productivity: comparisons in a saltmarsh. Estuaries 3:122-131.Zedler, J.B. 1984. The Ecology <strong>of</strong> Sou<strong>the</strong>rn California Coastal SaltMarshes: A Community Pr<strong>of</strong>ile. Report No. FWS/OBS-81/54,U.S. Fish and Wildlife Service, Washington, DC, USA.- 139 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaMECHANISTIC PROCESSES DRIVING SHIFTS IN BENTHIC INFAUNAL COMMUNITIESFOLLOWING HYBRID SPARTINA TIDAL FLAT INVASIONC. NEIRA 1 , E.D. GROSHOLZ 2 AND L.A. LEVIN 31 Integrative Oceanography Division, Scripps Institution <strong>of</strong> Oceanography, La Jolla, California 92093-0218, USA;cneira@coast.ucsd.edu2 Department <strong>of</strong> Environmental Science and Policy, University <strong>of</strong> California, Davis, One Shields Ave., Davis, CA 95616,USA; tedgrosholz@ucdavis.edu3 Integrative Oceanography Division, Scripps Institution <strong>of</strong> Oceanography, La Jolla, California 92093-0218, USA;llevin@ucsd.eduSpartina alterniflora x foliosa hybrids are perennial cordgrasses that have rapidly invaded mudflatsand marshes in central and sou<strong>the</strong>rn San Francisco Bay. Recent studies conducted by <strong>the</strong> authors at<strong>the</strong> Elsie Roemer Bird Sanctuary in Alameda (San Francisco Bay) showed a 75% reduction inmacr<strong>of</strong>aunal densities and shift in macr<strong>of</strong>aunal composition. Here we identify <strong>the</strong> mechanisms thatunderlie such observed changes in macr<strong>of</strong>aunal community structure following tidal flat invasion by<strong>the</strong> hybrid Spartina. Specifically we performed a series <strong>of</strong> in situ manipulative experiments toexamine hybrid Spartina canopy influence on water motion and water flow speed, larval flux,animal transport, and predation, as mediators <strong>of</strong> change for sediments and macrobenthos. Overall,hybrid Spartina exerted a strong influence on <strong>the</strong> hydrodynamic regime, reducing water flow that, inturn, influences flux <strong>of</strong> recruiting larvae, transport <strong>of</strong> o<strong>the</strong>r benthos, and <strong>the</strong> input <strong>of</strong> organic matterand sediment deposition. Habitat modification results in poor survivorship <strong>of</strong> surface-feeding taxavia sulfide toxicity, altered predation pressure and changed food availability. All <strong>the</strong>se mechanismscan play key, possibly synergistic roles in structuring Spartina-invaded ecosystems.Keywords: tidal flat invasion, Spartina alterniflora, hybrid Spartina, benthos, macr<strong>of</strong>auna,mechanistic processesINTRODUCTIONSan Francisco Bay is one <strong>of</strong> <strong>the</strong> most heavily invadedestuaries in <strong>the</strong> world with nearly 250 non-native orintroduced species (Cohen and Carlton 1998). One <strong>of</strong> <strong>the</strong>most serious invasions has been that <strong>of</strong> <strong>the</strong> Atlanticcordgrass Spartina alterniflora and its hybrids (hereafter“hybrid Spartina”). The genetic background <strong>of</strong> <strong>the</strong> plantswas confirmed by molecular genetic analysis by D. Ayres(unpublished results). Hybrid Spartina has invaded morethan 800 hectares (ha) <strong>of</strong> mudflats and marshes in centraland south San Francisco Bay (Ayres et al. 2004). Thisinvasive cordgrass is converting mudflats at low tidal levelsinto dense, nearly monotypic meadows (Ayres et al. 2003,2004). Of extreme concern is that this plant’s invasion intounvegetated tidal flats will result in a loss <strong>of</strong> open foragingarea for shorebirds and fishes, flow reductions, highersedimentation rates, changes in light penetration, andreduction <strong>of</strong> benthic algal production (Zipperer 1996;Daehler and Strong 1996; Stenzel et al. 2002; Grosholz et al.2009).To date very little is known about <strong>the</strong> impact <strong>of</strong> <strong>the</strong>hybrid Spartina invasion on sediment properties,macr<strong>of</strong>aunal communities, and ecosystem functioning. Whatwe do know results from recent studies <strong>of</strong> three sites in SanFrancisco Bay, including <strong>the</strong> Elsie Roemer Bird Sanctuary inAlameda (San Francisco Bay, CA, USA) (Neira et al. 2005),which has experienced hybrid Spartina invasion for <strong>the</strong> past30 years. At this site macr<strong>of</strong>aunal densities were 75% lowerin hybrid Spartina-invaded sediments than on adjacent,uninvaded tidal flats (Neira et al. 2005). Biomass was 57%lower in invaded sediments (Levin et al. 2006). We alsoobserved important shifts in species composition in <strong>the</strong>hybrid-invaded patches relative to tidal flats (Neira et al.2005). Surface feeders such as Gemma gemma (Bivalvia),Corophium spp. and Grandidierella japonica (Amphipoda),and Tharyx sp. and Eteone sp. (Polychaeta) were negativelyaffected by Spartina invasion, exhibiting reduced densities.Subsurface-deposit feeders such as capitellid polychaetesand tubificid oligochaetes were less affected or unaffected(Neira et al. 2005). Because surface-feeding taxa are moreaccessible to epibenthic consumers than capitellidpolychaetes and oligochaetes that live deeper in <strong>the</strong>sediment, <strong>the</strong> loss <strong>of</strong> surface feeding taxa could havepr<strong>of</strong>ound implications for higher trophic levels and henceaffect <strong>the</strong> whole ecosystem. Therefore, <strong>the</strong> aim <strong>of</strong> <strong>the</strong>following work was to experimentally document <strong>the</strong> causes<strong>of</strong> <strong>the</strong> benthic changes identified in <strong>the</strong> initial mensurativestudy.METHODSThe study site was located on Alameda Island (SanFrancisco Bay) along <strong>the</strong> shoreline adjacent to Elsie RoemerBird Sanctuary (37 o 45’35”N; 122 o 28’48”W). Detailed- 141 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinadescription <strong>of</strong> <strong>the</strong> study site is provided in Neira et al.(2005). Using a paired sampling design, we established 10blocks (2x2 meters (m)), approximately 10 m inside <strong>the</strong>hybrid Spartina meadow, and 10 similar blocks on <strong>the</strong>adjacent open tidal flat approximately 5-10 m from <strong>the</strong>meadow edge. We performed in situ manipulativeexperiments to identify what mechanisms are responsible forobserved shifts in faunal community structure followinghybrid Spartina invasion.Water motion: To evaluate <strong>the</strong> influence <strong>of</strong> <strong>the</strong> hybridSpartina canopy on relative water motion, we used <strong>the</strong>gypsum dissolution technique (Doty 1971). Pre-weighedgypsum cards were deployed about 10 cm above <strong>the</strong> bottomfor 6, 3 and 4 days during April and June 2002 in vegetatedand unvegetated tidal flat habitats. It is assumed thatdissolution rate <strong>of</strong> <strong>the</strong> gypsum, grams per day (g d -1 ), isproportional to water velocity (Porter et al. 2000). Watervelocity in both habitats was measured on several days nearmaximum ebb and flood tides just below <strong>the</strong> water surfacewith a Marsh-McBirney Flow Meter 2000.Larval flux: The influence <strong>of</strong> <strong>the</strong> hybrid Spartinacanopy on larval flux was measured in vegetated andunvegetated habitats using Geukensia demissa (mussel)shells as a hard substrate for settlement. Barnacle larvalabundance over time was used as a proxy for larval flux.Three dowels, each with one attached Geukensia shell, wereinserted into <strong>the</strong> sediment (1 m 2 plot) in each block <strong>of</strong> bothhybrid Spartina habitat and tidal flat (i.e., 60 in total).Mussels remained 10-20 cm above <strong>the</strong> sediment.Sediment deposition: We investigated <strong>the</strong> effects <strong>of</strong>Spartina plant canopy on short-term sediment depositionrates by deploying petri-dish sediment traps (Reed 1992) inboth vegetated and unvegetated habitats (10 replicates each)during low tide. Sediment traps were composed <strong>of</strong> GF/FWhatman filters (9-cm diameter) supported on <strong>the</strong> petridishes affixed on <strong>the</strong> sediment surface. Filters were removedand replaced 24 hours (h) later (one tidal cycle) during twosuccessive days in July 2002. Filters were gently rinsed withdistilled water to remove salts on pre-weighed aluminumdishes, oven-dried at 60 o C, and re-weighed. Sedimentdeposition rate was calculated as mass deposited per trap andexpressed in milligrams per centimeter squared per day (mgcm -2 d -1) . Percent organic matter was determined by massloss after ignition at 500 o C for 4 h. Mud content wasestimated after wet sieving (63 micrometers(μm)) <strong>the</strong>sediment deposited on traps and weighing both fractions (>63 μm and < 63 μm) after drying at 60°C.Animal transport: The influence <strong>of</strong> hybrid Spartinacanopy on animal-transport dynamics was evaluated usingpassive tube traps made <strong>of</strong> polypropylene (22 cm tall x 2.7cm diameter). To prevent escape <strong>of</strong> animals from <strong>the</strong> tube,15 milliliters (ml) <strong>of</strong> a dense brine solution (>90 parts perthousand (ppt)) was added; <strong>the</strong> remained volume was filledwith filtered seawater. Ten replicate tube traps weredeployed on <strong>the</strong> unvegetated tidal flat about 10 cm above <strong>the</strong>sediment surface, supported on PVC pipe inserted into <strong>the</strong>sediment. In <strong>the</strong> hybrid Spartina habitat, tube traps wereinserted into <strong>the</strong> sediment leaving <strong>the</strong>ir opening about 5 cmabove <strong>the</strong> bottom. Sediment was collected from traps inplastic jars and preserved in 8% buffered formalin.Sediment properties: In order to explore how hybridSpartina ultimately affects <strong>the</strong> sediment ecosystem, weexamined some key sediment properties in vegetated andunvegetated habitats (10 replicates each). Porewater salinity(top 3 cm) was measured with a hand-held refractometer,sediment redox potential (top 1 cm) was measured with aportable Mettler Toledo redox meter. Plexi-glass cores (18.1cm 2 , 0-6 cm depth) were taken for total organic content(determined after combustion at 500°C x 4h); syringe coreswere taken for chlorophyll a (an estimate <strong>of</strong> sedimentmicroalgal biomass) and determined according to Plante-Cuny (1973) after extraction with 90% acetone. Sedimentporosity was determined according to Buchanan (1984).Porewater sulfide was measured concurrently by A.C. Tyler(UC Davis) based on Cline (1969) with a few modifications.Macr<strong>of</strong>auna were collected from both habitats with cores(18.1 cm 2 , 0-6 cm depth) and sieved through a 0.3-mm meshsieve.Plant removal experiment: We also performed anexperiment to determine if <strong>the</strong> invaded habitats would returnto <strong>the</strong> original unvegetated condition and resembleunvegetated tidal flats. We clipped all above-groundSpartina plant material to <strong>the</strong> sediment level in ten replicate2x2 m areas with adjacent controls. In both removal andcontrols, we compared sediment properties and macr<strong>of</strong>aunalcommunities among habitats after 90 days.Sediment transplant experiment: We conducted atransplant experiment to determine whe<strong>the</strong>r unvegetatedsediment and fauna from <strong>the</strong> tidal flat would begin to looklike those in <strong>the</strong> invaded areas when moved into <strong>the</strong> areainvaded by hybrid Spartina. In summer 2002, intactsediment (625 cm 2 x 10 cm depth) was moved from <strong>the</strong> tidalflat (about 7 m from <strong>the</strong> edge) to Spartina-invaded habitat(10 m inside <strong>the</strong> meadow). Tidal flat sediment removed andreplaced served as a control treatment. Each treatmentcontained ten replicate blocks (20 in total).Predation: In order to assess if <strong>the</strong> shifts observed inmacr<strong>of</strong>aunal composition in <strong>the</strong> initial transplant experimentwere <strong>the</strong> result <strong>of</strong> changes in predation pressure followingSpartina-hybrid invasion, we performed controlled andreplicated in situ predator exclusion experiments beginningon September 10, 2003. We used European green crabs(Carcinus maenas), which are common (introduced) marshpredators in San Francisco Bay (Cohen et al. 1995). Intactuninvaded tidal flat sediment and associated fauna weretransplanted to hybrid Spartina habitat as in <strong>the</strong> transplantexperiment. However, this time we enclosed one individualC. maenas (carapace width 40 mm) in each <strong>of</strong> eightreplicated enclosures (0.3 x 0.3 m x 0.5 m height) made <strong>of</strong>galvanized hardware cloth (0.63-cm mesh). Crabs were- 142 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartinacollected from adjacent tidal flats and placed directly intoenclosures. A total <strong>of</strong> eight replicate blocks with fivetreatments per block were established (details in Neira et al.2006). All cages had a removable top for measurement <strong>of</strong>environmental variables such as temperature, salinity, lightpenetration, and redox potential. At <strong>the</strong> termination <strong>of</strong> <strong>the</strong>experiment (4 weeks), we recorded sediment temperature,salinity, light, and redox potential. For macr<strong>of</strong>aunalanalyses, we collected a single-core (18.1 cm 2 , 0-6 cm) from<strong>the</strong> center <strong>of</strong> each experimental and control enclosure. Aftersieving on a 0.3 mm mesh sieve, retained organisms weresorted, identified to species level and counted. An additionalcore was taken for analysis <strong>of</strong> organic matter content.RESULTSRelative water flow: Overall relative water velocity wasreduced by more than 50% in hybrid Spartina habitatrelative to unvegetated tidal flats. The plant canopy reducedmean flow speed from 3-4 cm s -1 in unvegetated tidal flat toless than 0.7 cm per second (s -1 ) in vegetated patches. Waterflux as estimated by mass loss <strong>of</strong> gypsum cards (pooledacross <strong>the</strong> three measurement periods) was 14.5 ±0.9 g d -1 in<strong>the</strong> tidal flat and 6.6 ±0.5 g d -1 in <strong>the</strong> vegetated patches(paired t test, P < 0.0001, Table 1). Gypsum cards placed on<strong>the</strong> tidal flat were strongly eroded which revealed <strong>the</strong>dynamic and variable nature <strong>of</strong> <strong>the</strong> water regime at ElsieRoemer.Larval flux: The hybrid Spartina canopy structurereduced larval flux relative to <strong>the</strong> unvegetated tidal flat.After two months, <strong>the</strong> number <strong>of</strong> barnacles per shell wasnearly 10 times higher in tidal flats (32.1 ±6.4) than inSpartina hybrid habitat (3.6 ±1.3 (paired t test, P = 0.022,Table 1).Sediment deposition: We recorded higher rates <strong>of</strong>sedimentation in <strong>the</strong> hybrid Spartina habitat (20.4 mg cm -2d -1 ) than on <strong>the</strong> unvegetated tidal flat (14.2 mg cm -2 d -1 )(paired t test, P = 0.005). The fine sediment fraction (< 63μm) deposited on traps placed in <strong>the</strong> hybrid habitat wasgreater than 45%, but was less than 13% in tidal flats (Table1).Animal transport: We found nearly two times as manyanimals entrained in traps deployed in <strong>the</strong> hybrid Spartinahabitat (4.9 ±0.6 per individual trap per day (ind trap -1 d -1 ))than on <strong>the</strong> unvegetated tidal flats (2.6 ±0.5 ind trap -1 d -1 )(paired t test, P = 0.019, Table 1. In addition, <strong>the</strong> faunalassemblage differed between habitats (ANOSIM, P =0.013). This may be <strong>the</strong> result <strong>of</strong> <strong>the</strong> canopy structureslowing water flow and reducing turbulence, thus allowinggreater deposition <strong>of</strong> animals in <strong>the</strong> collectors placed in <strong>the</strong>vegetated patches.Sediment properties: Most <strong>of</strong> <strong>the</strong> alterations weresimilar to those found during mensurative studies performedpreviously at this site (Neira et al. 2005). Relative tounvegetated tidal flat, sediments <strong>of</strong> unclipped hybridSpartina -vegetated patches were muddier (48.6% vs 10.7%Table 1. Summarized results <strong>of</strong> manipulative experiments performed at Elsie RoemerEnvironment variables / Macr<strong>of</strong>auna Tidal flat Spartina hybrid(Mean ± 1 SE)Water columnWater motion (weight loss <strong>of</strong> gypsum cards) (g d -1 ) 14.5 (±0.9) 6.6 (±0.5)Water speed (cm s -1 ) 3-4 0.3-0.7Larval flux (barnacle shell -1 ) 32.1 (±6.4) 3.6 (±1.3)Sediment deposition rate (mg cm -2 d -1 ) 14.2 (±) 20.4 (±)TOM <strong>of</strong> deposited sediment (%) 1.9 (±0.4) 13.9 (±0.8)Mud content (

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaobserved in unmanipulated sediments <strong>of</strong> hybrid Spartinahabitat relative to tidal flats (Neira et al. 2005).Predation: The sediment surface in cages containingcrabs was disturbed with obvious depressions created byforaging crabs. The sediments also had lower organic matterand chlorophyll a content relative to those cages withoutcrabs. No differences were found in salinity, temperatureand light penetration between treatments. Total macr<strong>of</strong>aunaldensity in experimental enclosures containing crabs was 2.2times lower than those without crabs. Density <strong>of</strong> selectedsurface-feeding species such as G. japonica, Corophium spp.and G. gemma declined by 57%, 87% and 67%, respectivelyin enclosures with crabs. O<strong>the</strong>r taxa that also declined wereSphaerosyllis californiensis, turbellarians, a juvenilearenicolid species and juvenile capitellids (Neira et al.2006).DISCUSSION AND CONCLUSIONOur results show that composition and structure <strong>of</strong> tidalflat macr<strong>of</strong>aunal benthic communities are stronglyinfluenced by hybrid Spartina invasion. Specifically, wefound that Spartina reduces water flow. These, in turn,influence flux <strong>of</strong> recruiting larvae, transport <strong>of</strong> o<strong>the</strong>rbenthos, <strong>the</strong> input <strong>of</strong> organic matter, and sedimentdeposition.These physical changes interacted with chemicalchanges including increased porewater sulfideconcentrations and more negative redox potential levels(Neira et al. 2006, 2007). We also found changes in growthand survival via predation (Neira et al. 2006) and foodavailability (Levin et al. 2006), which can play key, possiblysynergistic, roles in structuring Spartina-invadedecosystems. The presence <strong>of</strong> hybrid Spartina in open tidalflats exerts a strong influence on <strong>the</strong> composition anddistribution <strong>of</strong> benthic invertebrates in Elsie Roemer (Neiraet al. 2005, 2006).The reduction <strong>of</strong> macr<strong>of</strong>aunal densities in hybridSpartina relative to <strong>the</strong> naturally unvegetated tidal flats isconsistent with results for invasive S. anglica (Jackson 1985)and S. alterniflora in o<strong>the</strong>r estuaries (Zipperer 1996), butcontrasts with existing paradigms about positive vegetationeffects on marine macrobenthos (e.g., Hedge and Kriwoken2000; Netto and Lana 1999). In addition, not all systemsrespond in <strong>the</strong> same way to <strong>the</strong> hybrid Spartina invasion(Neira et al. 2005). For example, at o<strong>the</strong>r sites such asRoberts Landing (15-year invasion), hybrid Spartina habitatdiffered from tidal flat sediments in composition but notabundance. At a third San Francisco Bay site, in San Mateo,where a Salicornia marsh is being invaded (8-10 years),sediment properties were similar, and no differences weredetected in densities or proportions <strong>of</strong> surface- orsubsurface-deposit feeders, but <strong>the</strong> proportion <strong>of</strong>carnivores/omnivores and grazers increased in <strong>the</strong> hybridSpartina habitat (Neira et al. 2005). At Elsie Roemer, most<strong>of</strong> <strong>the</strong> species shown to have reduced density in hybridSpartina-invaded tidal flats are surface-feeding animals,such as amphipods (Corophium spp., Grandidierellajaponica) bivalves (Gemma gemma) and polychaetes(Tharyx spp.). Thus, where hybrid Spartina has invaded tidalflats it is likely to cause not only changes in <strong>the</strong> habitatstructure, but also to shift macr<strong>of</strong>aunal feeding modes fromsurface-microalgae feeders to subsurface detritivores. Theshift from surface-feeding taxa can have negative ecologicalimplications for higher trophic levels. Birds and fishesdepend more on larger, surface-feeding species (Simenstadand Thom 1995), which decline in abundance with <strong>the</strong>invasion <strong>of</strong> Spartina. However, <strong>the</strong> full impact <strong>of</strong> <strong>the</strong>Spartina invasion on birds and fishes in San Francisco Bayhas yet to be measured.We can conclude that <strong>the</strong> invasion <strong>of</strong> hybrid Spartinahas resulted in substantial changes in benthic communitiesby modifying <strong>the</strong> physical and chemical environment.However, we find that not all ecosystems respond identicallyfollowing hybrid Spartina invasion in San Francisco Bay.Hybrid Spartina can have differing and complex effects on<strong>the</strong> sediment environment and associated fauna dependingon <strong>the</strong> location, type <strong>of</strong> habitat involved, age <strong>of</strong> invasion,and local hydrodynamics. The processes underlying <strong>the</strong>variable responses to Spartina invasion, and <strong>the</strong> differentrates <strong>of</strong> recovery following invasion, are factors that shouldbe considered in planning for Spartina eradication.ACKNOWLEDGMENTSWe acknowledge G. Mendoza, R. Blake, C. Tyler, S.Maezumi, C. Whitcraft, U. Mahl, E. Brusati, N. Rayl, N.Christensen, P. McMillan, P. Colombano, J. Gonzalez, S.Norton, and D. Chiang who kindly assisted in experimentsetup, sample collection in <strong>the</strong> field and lab analysis. We aregrateful to C. Tyler who provided sulfide data. We thank C.Nordby, E. Brusati and R. Blake for providing compliancewith California Clapper Rail permit requirements. Supportwas provided by National Science FoundationBiocomplexity Program (DEB 0083583) to L.A.L. andE.D.G.A full manuscript <strong>of</strong> this study was published inEcological Applications (Neira et al. 2006); see referencebelow.REFERENCESAyres, D.R., D.R. Strong and P. Baye. 2003. Spartina foliosa(Poaceae) - A common species on <strong>the</strong> road to rarity. Madroño50:209-213.Ayres, D.R., D.L. Smith, K. Zaremba, S. Klohr and D.R. Strong.2004. Spread <strong>of</strong> exotic cordgrasses and hybrids (Spartina sp.) in<strong>the</strong> tidal marshes <strong>of</strong> San Francisco Bay. Biological Invasions6:221-231.Buchanan, J.B. 1984. Sediment analysis. In: Holme N.A., andMcIntyre, A.D., eds. Methods for <strong>the</strong> study <strong>of</strong> marine benthos, p.41-65, 2nd Ed. Blackwell Scientific Publications, Oxford,London.Cline, J.D. 1969. Spectrophotometric determination <strong>of</strong> hydrogensulfide in natural waters. Limnology and Oceanography 14:454-485.- 144 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaCohen, A.N., J.T. Carlton and M.C. Fountain. 1995. Introduction,dispersal, and potential impacts <strong>of</strong> <strong>the</strong> green crab Carcinusmaenas in San Francisco Bay, California. Marine Biology122:225-237.Cohen, A.N. and J.T. Carlton. 1998. Accelerating invasion rate in ahighly invaded estuary. Science 279:555-558.Daehler, C.C. and D.R. Strong. 1996. Status, prediction andprevention <strong>of</strong> introduced cordgrass Spartina spp. invasions inPacific estuaries, USA. Biological Conservation 78:51-58.Doty, M.S. 1971. Measurements <strong>of</strong> water movement in reference tobenthic algae growth. Botanica Marina 14:32-35.Grosholz, E.D., L.A. Levin, A.C. Tyler, and C. Neira. 2009.Changes in community structure and ecosystem functionfollowing Spartina alterniflora invasion <strong>of</strong> Pacific estuaries, p.23-40. In: Silliman, B.R., Grosholz, E.D., and Bertness, M.D.,eds. Human impacts on salt marshes: a global perspective.University <strong>of</strong> California Press, Berkeley and Los Angeles,California.Hedge, P. and L.K. Kriwoken. 2000. Evidence for effects <strong>of</strong>Spartina anglica invasion on benthic macr<strong>of</strong>auna in LittleSwanport estuary, Tasmania. Austral Ecology 25:150-159.Jackson, D. 1985. Invertebrate populations associated withSpartina anglica salt-marsh and adjacent intertidal mud flats.Estuarine Brackishwater Science Association Bulletin 40:8-14.Levin, L.A., C. Neira and E.D. Grosholz. 2006. Invasive cordgrassmodifies wetland and trophic function. Ecology 87:419-432.Neira, C., L.A. Levin and E.D. Grosholz. 2005. Benthicmacr<strong>of</strong>aunal communities <strong>of</strong> three sites in San Francisco Bayinvaded by hybrid Spartina, with comparison to uninvadedhabitats. Marine Ecology Progress Series 292:111-126.Neira, C., E.D. Grosholz, L.A. Levin and R. Blake. 2006.Mechanisms generating modification <strong>of</strong> benthos following tidalflat invasion by a Spartina hybrid. Ecological Applications16:1391-1404.Neira, C., L.A. Levin, E.D. Grosholz and G. Mendoza. 2007.Influence <strong>of</strong> invasive Spartina growth stages on associatedmacr<strong>of</strong>aunal communities. Biological Invasions 9:975-993.Netto, S.A., and P.C. Lana. 1999. The role <strong>of</strong> above- and belowgroundcomponents <strong>of</strong> Spartina alterniflora (Loisel) and detritusbiomass in structuring macrobenthic associations <strong>of</strong> ParanaguaBay (SE, Brazil). Hydrobiologia 400:167-177.Nichols, F.H., and J.K. Thompson. 1985. Persistance <strong>of</strong> anintroduced mudflat community in South San Francisco Bay,California. Marine Ecology Progress Series 24:83-97.Plante-Cuny, M.R. 1973. Reserches sur la production primairebentique en milieu marin tropical. I. Variations de la productionprimaire et des teneurs en pigments photosynthétiques surquelques fonds sableux. Valeur des résultats obtenus par laméthode du 14 C. Cahiers O.R.S.T.O.M., sér. Océanographique.11:317-348.Porter, E.K., L.P. Sanford and S.E. Suttles. 2000. Gypsumdissolution is not a universal integrator <strong>of</strong> “water motion”.Limnology and Oceanography 45:145-158.Reed, D.J. 1992. Effects <strong>of</strong> weirs on sediment deposition inLouisiana coastal marshes. Environmental Management 16:55-65.Simenstad, C.A., and R.M. Thom. 1995. Spartina alterniflora(smooth cordgrass) as an invasive halophyte in Pacific NorthwestEstuaries. Hortus Northwest 6: 9-12 and 38-40.Stenzel, L.E., C.M. Hickey, J.E. Kjelmyr and G.W. Page. 2002.Abundance and distribution <strong>of</strong> shorebirds in <strong>the</strong> San FranciscoBay area. Western Birds 33:69-98.Zipperer, V.T. 1996. Ecological effects <strong>of</strong> <strong>the</strong> introducedcordgrass, Spartina alterniflora, on <strong>the</strong> benthic communitystructure <strong>of</strong> Willapa Bay, Washington. M.S. Thesis. University<strong>of</strong> Washington, Seattle, Washington.- 145 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaSPARTINA ALTERNIFLORA INVASIONS IN THE YANGTZE RIVER ESTUARY,CHINA:ASYNOPSISB. LI 1 ,C-Z.LIAO, X-D. ZHANG, H-L. CHEN,Q.WANG, Z-Y. CHEN, X-J. GAN, J-H. WU,B.ZHAO, Z-J. MA,X-L. CHENG,L-F.JIANG, Y-Q. LUO, AND J-K. CHENCoastal Ecosystems Research Station <strong>of</strong> <strong>the</strong> Yangtze River Estuary, Ministry <strong>of</strong> Education Key Laboratory for BiodiversityScience and Ecological Engineering, The Institute <strong>of</strong> Biodiversity Science, Fudan University, #220 Handan Road, Shanghai200433, China1 For correspondence: bool@fudan.edu.cnSpartina alterniflora was first found in <strong>the</strong> Yangtze River estuary in <strong>the</strong> mid 1990s, and has nowbecome <strong>the</strong> most abundant vascular plant in <strong>the</strong> estuarine marshlands. We have investigated <strong>the</strong>potential consequences <strong>of</strong> Spartina alterniflora invasions to <strong>the</strong> salt marshes on two large islands in<strong>the</strong> estuary, Chongming and Jiuduansha, over <strong>the</strong> past five years, focusing on effects on biodiversityand ecosystem processes <strong>of</strong> <strong>the</strong> marshlands resulting from <strong>the</strong> invasion. We here summarize <strong>the</strong>major findings from our previous work and provide <strong>the</strong> relevant literature.Keywords: Biodiversity, ecosystem processes, plant invasions, saltmarshes, Spartina alterniflora,Yangtze River estuaryINTRODUCTIONThe Yangtze River estuary with two national naturereserves is an important ecoregion as it is <strong>the</strong> home for manyeconomically and ecologically important species, and servesas an important stopover site for migratory birds on <strong>the</strong> EastAsian-Australasian Flyway (Chen et al. 2003; Ma et al.2004). However, in common with o<strong>the</strong>r estuaries in <strong>the</strong>world (Cohen and Carlton 1998; Grosholz 2002), <strong>the</strong>Yangtze River estuary is seriously threatened by exoticspecies invasions. Of all <strong>the</strong> invasive exotic species, smoothcordgrass (Spartina alterniflora) introduced from NorthAmerica, has become <strong>the</strong> most harmful exotic plant to <strong>the</strong>salt marshes, leading to multifold consequences to <strong>the</strong>estuary. Over <strong>the</strong> past five years, we have investigated <strong>the</strong>population spread <strong>of</strong> S. alterniflora on two major islands in<strong>the</strong> estuary, Chongming and Jiuduansha, and examined <strong>the</strong>ecosystem-level effects <strong>of</strong> its invasion. This synopsis isbased on <strong>the</strong>se investigations. Detailed reports werepublished elsewhere. (See References for <strong>the</strong> full list.)SPREAD AND DISTRIBUTIONS. alterniflora was intentionally introduced fromAmerica to China in 1979 (An et al. this volume), and was<strong>the</strong>n spread to <strong>the</strong> Yangtze River estuary in <strong>the</strong> mid 1990sboth by natural dispersal and intentional introductions. For<strong>the</strong> purposes <strong>of</strong> ecological engineering, i.e., rapid sedimentaccretion, S. alterniflora was <strong>the</strong>n intentionally introduced totwo large islands in <strong>the</strong> estuary, Jiuduansha (in 1997) andChongming (in 2001 and 2003). On each island a nationalnature reserve had been set aside for conserving nativebiodiversity and maintaining ecosystem integrity. In fact,this invasive plant has invaded almost all <strong>the</strong> mudflats andsalt marshes in <strong>the</strong> Yangtze River estuary (Fig. 1A), S. alternifloraei<strong>the</strong>r colonized tidal mudflats (Fig. 1D) or replacednative plants like Scirpus mariqueter (Fig. 1E) (Chen et al.2004) and Phragmites australis (Wang et al. 2006b), and hasbecome one <strong>of</strong> <strong>the</strong> most abundant species in estuarineecosystems during a period <strong>of</strong> just over 10 years. Our recentdata obtained through remote sensing show that in 2005 S.alterniflora monocultures accounted for 49.4% and 37.0% <strong>of</strong><strong>the</strong> vegetated area in Dongtan (Fig. 1B) and Jiuduansha (Fig.1C) marshlands respectively (Li et al. 2009). Range expansion<strong>of</strong> S. alterniflora still continues in estuarine wetlands.Successful invasions <strong>of</strong> S. alterniflora in <strong>the</strong> YangtzeRiver estuary were <strong>the</strong> result <strong>of</strong> interactions between biotic,abiotic and human factors (Wang et al. 2006a). The “enemyrelease hypo<strong>the</strong>sis” that is used to explain successful invasions<strong>of</strong> exotic plants elsewhere might also apply to those <strong>of</strong> S.alterniflora in <strong>the</strong> Yangtze River estuary. Spartina alterniflorahas a number <strong>of</strong> superior traits such as fast growth, a highlyefficient use <strong>of</strong> resources, a high tolerance to salt and a welldevelopedbelowground system that make it a superiorcompetitor or invader and a potent ecosystem engineer (Li etal. 2009). It is for this latter reason that S. alterniflora waswidely used for erosion control and sediment accretion alongshorelines in <strong>the</strong> Yangtze River estuary (Wang et al. 2006a),which might have directly led to its rapid spread in <strong>the</strong>marshlands. Frequent reclamation in <strong>the</strong> Yangtze Riverestuary has made <strong>the</strong> estuarine ecosystem susceptible to S.alterniflora as areas outside <strong>the</strong> dike exhibit conditions typical<strong>of</strong> early saltmarsh succession, which favor S. alterniflorara<strong>the</strong>r than native plants. Environmental changes (saltwaterintrusion and eutrophication) caused by human activities- 147 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaFig. 1A) Locations <strong>of</strong> Jiuduansha and Chongming Dongtan in <strong>the</strong> Yangtze River estuary, on each <strong>of</strong> which a national nature reserve was set up in 2005 (Li etal. 2009); B) and C) Distribution <strong>of</strong> S. alterniflora in Chongming Dongtan and Jiuduansha marshlands in 2005; D) Conversion <strong>of</strong> tidal mudflats to S.alterniflora meadows; and E) Replacement <strong>of</strong> S. mariqueter with S. alterniflora.might also facilitate fur<strong>the</strong>r invasions in <strong>the</strong> estuary (Wang etal. 2006a, b).PLANT COMMUNITIESOne <strong>of</strong> <strong>the</strong> threats <strong>of</strong> invasive exotic plants to nativeecosystems is <strong>the</strong>ir competitive effects on native plantbiodiversity. Our experimental studies showed that S.alterniflora had great competitive effects on native plantspecies in <strong>the</strong> estuary, e.g., S. mariqueter, which is endemicto <strong>the</strong> Yangtze River estuary (Chen et al. 2004; 2005b) andP. australis (Wang et al. 2006b), which resulted in <strong>the</strong>reduced abundance and even local extinction <strong>of</strong> <strong>the</strong>se nativeplants. In particular, <strong>the</strong> decline <strong>of</strong> S. mariqueter abundancehas been dramatic, which may have important consequencesto shorebird communities because this sedge serves as foodfor shorebirds and creates favorable habitats for <strong>the</strong> birds(Ma et al. 2009; Li et al. 2009).SOIL BIOTIC COMMUNITIESNematodes are important components <strong>of</strong> <strong>the</strong> soilecosystem and play important roles in ecosystemfunctioning. The changes in composition <strong>of</strong> plantcommunities are believed to affect <strong>the</strong> structure <strong>of</strong> nematodecommunities (Yeates 1999). We compared <strong>the</strong> structure <strong>of</strong>soil nematode communities among native (S. mariqueter andP. australis) and S. alterniflora communities on Chongming,Jiuduansha Islands and Nanhui (Chen et al. 2007b). Therewere no significant differences among <strong>the</strong> three plantcommunities in total density <strong>of</strong> nematodes, <strong>the</strong> number <strong>of</strong>genera, or in <strong>the</strong> diversity indices. However, S. alterniflorainvasions altered <strong>the</strong> trophic structure <strong>of</strong> nematodecommunities in soil. The relative abundance <strong>of</strong> bacterivoressignificantly increased, whereas that <strong>of</strong> plant feeders andalgal feeders declined in S. alterniflora communities,compared with that in P. australis and S. mariquetercommunities. Our fur<strong>the</strong>r experimental study demonstratedthat <strong>the</strong> changes in nematode communities were <strong>the</strong> result <strong>of</strong>altered litter quality due to <strong>the</strong> replacement <strong>of</strong> native plantsby S. alterniflora (Chen et al. 2007a).Similarly, we compared <strong>the</strong> structure <strong>of</strong> macrobenthicinvertebrate communities between S. mariqueter and S.alterniflora communities at Chongming Dongtan (Chen etal. 2005a). No significant difference in total density <strong>of</strong>macrobenthic invertebrates was detected between <strong>the</strong> twoplant communities although <strong>the</strong> relative abundance <strong>of</strong> somecommon species was altered, ei<strong>the</strong>r increased or decreased.One <strong>of</strong> our recent studies also showed that S. alternifloraprovided compatible habitats for <strong>the</strong> native crab Sesarmadehaani by <strong>of</strong>fering moderate environmental conditions- 148 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina(e.g., mild temperature in summer) (Wang et al. 2008). Likenematodes, <strong>the</strong> relative abundance <strong>of</strong> major trophicfunctional groups <strong>of</strong> macrobenthic invertebrates was alsoaltered when S. mariqueter was replaced by S. alterniflora.Spartina alterniflora communities had a higher proportion <strong>of</strong>suspensivores but a lower proportion <strong>of</strong> herbivores anddetritivores than S. mariqueter communities (Chen et al.2005a).In addition, we compared <strong>the</strong> structure and diversity <strong>of</strong><strong>the</strong> bacterial communities in rhizosphere soils <strong>of</strong> S.alterniflora, P. australis and S. mariqueter throughconstructing 16S ribosomal DNA (rDNA) clone libraries.Our results showed that <strong>the</strong> shift <strong>of</strong> species composition <strong>of</strong>plant communities from native plants to S. alternifloracaused considerable changes in bacterial composition <strong>of</strong>rhizosphere soils in estuarine salt masrhes (Wang et al.2007).The above changes in soil biota caused by S.alterniflora invasions could have great effects on ecosystemprocesses in <strong>the</strong> soils that are associated with <strong>the</strong>seorganisms as decomposers.ARTHROPOD COMMUNITIESPlants serve as both food and habitats for manyarthropods, so subtle shifts in plant community compositionmay lead to considerable changes in <strong>the</strong> arthropodcommunity. In order to understand <strong>the</strong> possible effects <strong>of</strong> S.alterniflora invasions on arthropods, we examined <strong>the</strong>community structure and diets <strong>of</strong> arthropods in an estuarinesalt marsh previously dominated by native P. australis atChongming Dongtan through net sweeping and plantharvesting methods and stable isotope analysis (Wu et al.2009). We found that diversity indices were not significantlydifferent between exotic and native plant communities, but<strong>the</strong> total abundance <strong>of</strong> insects estimated through plantharvesting was found to be lower in S. alternifloramonocultures than that in P. australis monocultures.Community structure <strong>of</strong> insects in S. alternifloramonocultures was dissimilar to that in P. australismonocultures and P. australis–S. alterniflora mixtures.Moreover, stable isotope analysis showed that althoughsome native arthropods (perhaps generalists) shifted <strong>the</strong>irdiets, most native taxa did prefer P. australis to S.alterniflora even in S. alterniflora monocultures. Spartinaalterniflora invasions resulted in reduced abundances <strong>of</strong>some arthropds, and increased <strong>the</strong> dominance <strong>of</strong> o<strong>the</strong>rs thatfed preferentially on S. alterniflora. The experimentalevidence provided showed that invasive plants can change<strong>the</strong> community structure and diets <strong>of</strong> native arthropods,which will eventually alter arthropod food webs, and affect<strong>the</strong> integrity and functioning <strong>of</strong> native ecosystems within anature reserve that has been set aside for conserving <strong>the</strong>native biodiversity and maintaining <strong>the</strong> ecosystem integrity.BIRD COMMUNITIESThe conversion <strong>of</strong> mudflats to meadows resulting fromS. alterniflora invasions had significant impact on birds <strong>of</strong>Charadriidae and Scolopacidae, which might have beenattributable to <strong>the</strong> reduction in food resources and physicalalterations <strong>of</strong> habitats for birds (Li et al. 2009). Never<strong>the</strong>less,S. alterniflora may also provide habitat for certain landbirds.Recent bird surveys indicated that <strong>the</strong> Japanese marshwarbler Megalurus pryeri, a newly-recorded species in <strong>the</strong>Yangtze River estuary, nested exclusively in S. alternifloracommunities (Gan et al. 2006). This might be because <strong>the</strong>dense vegetation benefits <strong>the</strong> construction <strong>of</strong> nests. Inaddition, M. pryeri largely fed on <strong>the</strong> invertebrates in S.alterniflora communities (Gan 2009). It is likely that <strong>the</strong>rapid population increase <strong>of</strong> <strong>the</strong> Japanese marsh warbler wasrelated to <strong>the</strong> rapid spread <strong>of</strong> S. alterniflora in <strong>the</strong> YangtzeRiver estuary. However, <strong>the</strong> effects <strong>of</strong> S. alterniflora onlocal landbird communities still remain largely unexplored,and fur<strong>the</strong>r studies are needed.CARBON AND NITROGEN CYCLESS. alterniflora had great productive potential in <strong>the</strong>marshlands in <strong>the</strong> Yangtze River estuary. The results weobtained in Jiuduansha marshlands showed that exotic S.alterniflora had much greater primary productivity thannatives S. mariqueter and P. australis (Liao et al. 2007)because <strong>the</strong> exotic had greater leaf area index (LAI) and alonger photosyn<strong>the</strong>tic season than <strong>the</strong> native species (Jianget al. 2009). At <strong>the</strong> same time, decomposition rates <strong>of</strong> S.alterniflora litter, particularly <strong>the</strong> belowground litter, werelower than those <strong>of</strong> S. mariqueter and P. australis litter dueto <strong>the</strong> lower litter quality <strong>of</strong> S. alterniflora (Liao et al. 2008).Therefore, larger stocks <strong>of</strong> carbon and nitrogen were foundin <strong>the</strong> ecosystems dominated by S. alterniflora than in thosedominated by S. mariqueter and P. australis. Our fur<strong>the</strong>ranalysis <strong>of</strong> stable carbon isotopes also confirmed that <strong>the</strong>replacement <strong>of</strong> S. mariqueter by S. alterniflora significantlyincreased soil organic carbon and total soil nitrogen (Chenget al. 2006), especially soil labile carbon, recalcitrant carbon,and soil recalcitrant nitrogen contents in <strong>the</strong> upper soil layers(0–60 cm) (Cheng et al. 2008). The ecosystem carbon andnitrogen cycles altered by S. alterniflora invasions might be<strong>of</strong> limited significance on a large scale in relation to <strong>the</strong>range <strong>of</strong> S. alterniflora in <strong>the</strong> Yangtze River estuary, butmight have potentially far-reaching impact on <strong>the</strong> adjacentecosystems. In fact, S. alterniflora functioned as <strong>the</strong> primaryenergy source for certain nektons in <strong>the</strong> marshlands in <strong>the</strong>Yangtze River estuary (Quan et al. 2007). Potentialcascading effects on estuarine food webs <strong>of</strong> increasedecosystem productivity due to S. alterniflora invasions needsfur<strong>the</strong>r exploration.- 149 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaCONCLUSIONSThe invasion history <strong>of</strong> S. alterniflora in <strong>the</strong> YangtzeRiver estuary is relatively short, but this invasive plant hasbeen found to cause considerable impact on structure andfunctioning <strong>of</strong> native ecosystems although it might havepositive effects on certain species. However, <strong>the</strong> responses<strong>of</strong> native ecosystems to plant invasions may be <strong>of</strong> longerduration than a few decades, thus long-term monitoring isneeded to understand <strong>the</strong> ecosystem changes in response toS. alterniflora invasions. In <strong>the</strong> meantime, most <strong>of</strong> ourstudies have examined <strong>the</strong> effects <strong>of</strong> S. alterniflora invasionson biodiversity and processes <strong>of</strong> native ecosystems.Therefore, future studies are clearly needed to examine how<strong>the</strong> change <strong>of</strong> one ecosystem component affects or isaffected by those <strong>of</strong> o<strong>the</strong>r components, as examined byLevin et al. (2006) in San Francisco Bay (USA), and to link<strong>the</strong> changes <strong>of</strong> biodiversity caused by S. alterniflorainvasions to <strong>the</strong> alterations <strong>of</strong> ecosystem processes. Finally,considering that most <strong>of</strong> <strong>the</strong> marshlands invaded by S.alterniflora that we studied are protected for conserving <strong>the</strong>native biodiversity and maintaining <strong>the</strong> ecosystem integrityin <strong>the</strong> Yangtze River estuary, S. alterniflora invasions needto be managed appropriately.ACKNOWLEDGMENTSWe thank <strong>the</strong> students at <strong>the</strong> Ecology Lab <strong>of</strong> BiologicalInvasions at Fudan University who <strong>of</strong>fered assistance in <strong>the</strong>filed. The entire work described here was financiallysupported by National Basic Research Program <strong>of</strong> China(Grant no.: 2006CB403305), Natural Science Foundation <strong>of</strong>China (Grant No.: 30670330 and 30370235), <strong>the</strong> Ministry <strong>of</strong>Education <strong>of</strong> China (Grant No.: 105063), and <strong>the</strong> Scienceand Technology Department <strong>of</strong> Shanghai (Grant No.:04DZ19304) to Bo Li.REFERENCESChen, H.L., Li B., Fang C.M., Chen J.K., and Wu J.H. 2007a. Exoticplant influences soil nematode communities through litterinput. Soil Biology and Biochemistry 39: 1782-1793.Chen, H.L., Li B., Hu J.B., Chen J.K., and Wu J.H. 2007b. Effects<strong>of</strong> Spartina alterniflora invasion on benthic nematode communitiesin <strong>the</strong> Yangtze Estuary. Marine Ecology-Progress Series336: 99-110.Chen, J.K., Ma Z.J. and Li B. (ed). 2003. Comprehensive Surveyson Shanghai Jiuduansha Wetland Nature Reserve, <strong>the</strong> YangtzeRiver Estuary. Beijing: Science Press.Chen, Z.Y., Li B., Zhong Y., and Chen J.K. 2004. Local competitiveeffects <strong>of</strong> introduced Spartina alterniflora on Scirpus mariqueterat Dongtan <strong>of</strong> Chongming Island, <strong>the</strong> Yangtze River estuaryand <strong>the</strong>ir potential ecological consequences. Hydrobiologia528: 99-106.Chen, Z.Y., Fu C.Z., Wang H.Y., Li B., Wu J.H., and Chen J.K.2005a. Effects <strong>of</strong> Spartina alterniflora invasions on <strong>the</strong> macrobenthicinvertebrate community in saltmarshes at ChongmingDongtan , <strong>the</strong> Yangtze River Estuary. Wetland Science, 3: 1-7.Chen, Z.Y., Li B. and Chen J.K. 2005b. Growth-related traits andrelative competitive ability <strong>of</strong> invasive Spartian alterniflora andnative Scirpus mariqueter. Biodiversity Science 13: 130-136.Cheng, X.L., Luo Y.Q., Chen J.Q., Lin G.H., Chen J.K. and Li B.2006. Short-term C 4 plant Spartina alterniflora invasions change<strong>the</strong> soil carbon in C 3 plant-dominated tidal wetlands on a growingestuarine island. Soil Biology and Biochemistry 38: 3380–3386.Cheng, X.L., Chen J.Q., Luo Y.Q., Henderson R., Zhang Q.F.,Chen J.K. and Li B. 2008. Assessing <strong>the</strong> effects <strong>of</strong> short-termSpartina alterniflora invasion on labile and recalcitrant C and Npools by means <strong>of</strong> soil fractionation and stable C and N isotopes.Geoderma 145: 177-184.Cohen, A.N. and Carlton J.T. 1998. Accelerating invasion rate in ahighly invaded estuary. Science 279: 555-558.Gan, X.J. 2009. Habitat selection <strong>of</strong> saltmarsh birds at ChongmingDongtan in <strong>the</strong> Yangtze River Estuary as influenced by Spartinaalterniflora invasions. Ph.D. dissertation, Fudan University,Shanghai, China.Gan, X.J., Zhang K.J., Tang S.M., Li B. and Ma Z.J. 2006. Threenew records <strong>of</strong> birds in Shanghai: Locustella pleskei (Pleske’swarbler), Megalurus pryeri (Japanese swamp warbler) and Acrocephalusconcinens (blunt-winged paddyfield warbler). Journal<strong>of</strong> Fudan University (Natural Science) 45: 417–419.Grosholz, E. 2002. Ecological and evolutionary consequences <strong>of</strong>coastal invasions. Trends in Ecology and Evolution 17: 22-27.Jiang, L.F., Luo Y.Q., Chen J.K. and Li B. 2009. Ecophysiologicalcharacteristics <strong>of</strong> invasive Spartina alterniflora and native speciesin salt marshes <strong>of</strong> Yangtze River estuary, China. Estuarine,Coastal and Shelf Science 81:74-82.Levin, L.A, Neira C., and Grosholz E.D. 2006. Invasive cordgrassmodifies wetland trophic function. Ecology 87: 419-432.Li B., Liao C.Z., Zhang X.D., Chen H.L, Wang Q., Chen Z.Y., GanX.J., Wu J.H., Zhao B., Ma Z.J., Cheng X.L., Jiang L.F. andChen J.K. 2009. Spartina alterniflora invasions in <strong>the</strong> YangtzeRiver estuary, China: an overview <strong>of</strong> current status and ecosystemeffects. Ecological Engineering 35:511-520.Liao, C.Z., Luo Y.Q., Jiang L.F., Zhou X.H., Wu X.W., Fang C.M.,Chen J.K. and Li B. 2007. Invasion <strong>of</strong> Spartina alterniflora enhancedecosystem carbon and nitrogen stocks in <strong>the</strong> Yangtze Estuary,China. Ecosystems 10: 1351-1361.Liao, C.Z., Luo Y.Q., Fang C.M., Chen J.K. and Li B. 2008. Litterpool sizes, decomposition, and nitrogen dynamics in Spartina alterniflora-invadedand native coastal marshlands <strong>of</strong> <strong>the</strong> YangtzeEstuary. Oecologia 156: 589-600.Ma, Z.J., Li B., Jing K., Tang S.M. and Chen J.K. 2004. Are artificialwetlands good alternatives to natural wetlands for waterbirds?-A case study on Chongming Island, China. Biodiversityand Conservation 13: 333-350.Ma, Z.J., Wang Y., Gan X.J., Li B., Cai Y.T. and Chen J.K. 2009.Waterbird population changes in <strong>the</strong> wetlands at ChongmingDongtan in <strong>the</strong> Yangtze River estuary, China. EnvironmentalManagement 43:1187-1200.Quan, W.M., Fu C.Z., Jin B.S., Luo Y.Q., Li B., Chen J.K. and WuJ.H. 2007. Tidal marshes as energy sources for commerciallyimportant nektonic organisms: stable isotope analysis. MarineEcology-Progress Series 352: 89–99.Wang, J.Q., Zhang X.D., Nie M., Fu C.Z., Chen J.K. and Li B.(2008) Exotic Spartina alterniflora provides compatible habitatsfor native estuarine crab Sesarma dehaani in <strong>the</strong> Yangtze Riverestuary. Ecological Engineering 34: 57-64.Wang, M., Chen J.K. and Li B. 2007. Characterization <strong>of</strong> bacterialcommunity structure and diversity in rhizosphere soils <strong>of</strong> threeplants in rapidly changing salt marshes using 16S rDNA. Pedosphere17: 545-556.- 150 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaWang, Q., An S.Q., Ma Z.J., Zhao B., Chen J.K. and Li B. 2006a.Invasive Spartina alterniflora: biology, ecology and management.Acta Phytotaxonomica Sinica 44: 559–588.Wang, Q., Wang C.H., Zhao B., Ma Z.J., Luo Y.Q., Chen J.K. andLi B. 2006b. Effects <strong>of</strong> growing conditions on <strong>the</strong> growth <strong>of</strong> andinteractions between salt marsh plants: implications for invasibility<strong>of</strong> habitats. Biological Invasions 8: 1547-1560.Wu, Y.T., Wang C.H., Zhang X.D., Zhao B., Jiang L.F., Chen J.K.and Li B. 2009. Effects <strong>of</strong> saltmarsh invasion by Spartina alternifloraon arthropod community structure and diets. BiologicalInvasions 11:635-649.Yeates, G.W. 1999. Effects <strong>of</strong> plant on nematode community structure.Annual Review <strong>of</strong> Phytopathology 37: 127-149.- 151 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaTHE ROLE OF SPARTINA ANGLICA PRODUCTION IN BIVALVE DIETS IN NORTHERNPUGET SOUND, WA, USAC.E. HELLQUIST 1 AND R.A. BLACKSchool <strong>of</strong> Biological Sciences, Washington State University, Pullman, WA 991641 Current address: Department <strong>of</strong> Biological Sciences, State University <strong>of</strong> New York, Oswego, NY 13126;eric.hellquist@oswego.eduThe importance <strong>of</strong> salt marsh productivity to coastal food webs is a question <strong>of</strong> ecological andeconomic importance. In nor<strong>the</strong>rn Puget Sound, Washington, USA, a relatively new source <strong>of</strong>production has become prominent with <strong>the</strong> establishment <strong>of</strong> invasive Spartina anglica (Poaceae:English cordgrass). Spartina anglica has converted native coastal mudflat communities that hadlittle or no emergent vascular vegetation into expansive cordgrass meadows. One consequence <strong>of</strong>Spartina productivity on invaded mudflats may be altered trophic patterns. Three bivalves withdifferent feeding modes (Macoma balthica, Mya arenaria, and Mytilus sp.) that are commonlyfound at <strong>the</strong> edges <strong>of</strong> Spartina meadows were selected to investigate whe<strong>the</strong>r Spartina iscontributing to bivalve diets. We compared <strong>the</strong> stable isotope ratios (δ 13 C and δ 34 S) <strong>of</strong> <strong>the</strong> bivalvesto potential food sources including macroalgae, Spartina, and o<strong>the</strong>r vascular plants. We estimated<strong>the</strong> feasible contributions <strong>of</strong> production to bivalve diets during March 2003 by analyzing <strong>the</strong> results<strong>of</strong> multiple source, mass-balanced linear mixing models as calculated by IsoSource. Our estimatesindicate that Spartina biomass may comprise 37-60% <strong>of</strong> <strong>the</strong> diet <strong>of</strong> Macoma, while dead Spartinabiomass contributes 0-46%. For Mya arenaria a filter-feeding clam, 40-59% <strong>of</strong> its diet may containSpartina biomass, while 0-35% <strong>of</strong> its diet may consist <strong>of</strong> dead Spartina biomass. Mytilus sp., afilter-feeding mussel, had 19-44% <strong>of</strong> its diet originating from Spartina biomass while dead Spartinamay be 0-46% <strong>of</strong> its diet. Spartina consumption by bivalves is consistent with previous isotopicstudies. Although Spartina biomass is considered recalcitrant, <strong>the</strong> immediate proximity <strong>of</strong> <strong>the</strong>consumers to vast quantities <strong>of</strong> Spartina productivity may best explain <strong>the</strong> prevalence <strong>of</strong> Spartina inbivalve diets while o<strong>the</strong>r potential sources have minor estimated contributions. This study providesan initial examination <strong>of</strong> how <strong>the</strong> biomass <strong>of</strong> an invasive plant species is becoming integrated intoestuarine trophic webs.Keywords: Spartina anglica, stable isotopes, mixing models, IsoSource, Macoma, Mytilus, MyaINTRODUCTIONEstuarine salt marshes intercept nutrients and biomassfrom uplands and also export nutrients and biomass intonearshore coastal ecosystems (Deegan and Garritt 1997;Tealand Howes 2000; Valiela et al. 2000). The importance <strong>of</strong> saltmarsh productivity in estuarine ecosystems has been acentral research question with ramifications foreconomically important coastal fisheries (Peterson et al.1985, 1986; Teal and Howes 2000; Valiela et al. 2000).Few studies have addressed <strong>the</strong> role <strong>of</strong> invasive speciesin altering ecosystem processes in marine and estuarineenvironments (Ruiz et al. 1997, Grosholz 2002) despite <strong>the</strong>wealth <strong>of</strong> research examining how invasive species alterlandscapes, nutrient cycling, and species interactions (Macket al. 2000 and references <strong>the</strong>rein). Spartina anglica C. E.Hubbard (Poaceae; English cordgrass) was introduced intonor<strong>the</strong>rn Puget Sound, Washington in <strong>the</strong> early 1960s inSnohomish County (Hacker et al. 2001; Hellquist 2005).Spartina anglica covers nearly 300 solid hectares (ha) <strong>of</strong>Puget Sound intertidal habitat (Hedge et al. 2003). Spartinaanglica colonizes s<strong>of</strong>t sediments and is capable <strong>of</strong> rapidlyconverting mudflat habitats into elevated Spartina meadows.Habitat conversion by Spartina is <strong>of</strong> great concernecologically, economically, and aes<strong>the</strong>tically (Hacker et al.2001; Hedge et al. 2003).There are no native maritime species <strong>of</strong> Spartina in <strong>the</strong>Pacific Northwest <strong>of</strong> North America and thus mudflatecosystems <strong>of</strong> Puget Sound have developed without lowintertidal meadows <strong>of</strong> emergent C 4 vegetation (i.e. S.anglica) along <strong>the</strong>ir periphery. Generally, <strong>the</strong> high carboncontents <strong>of</strong> C 4 plants have been considered to be <strong>of</strong> lownutritional quality (Caswell et al. 1973). Extensive meadows<strong>of</strong> Spartina that colonize mudflats represent a large subsidy<strong>of</strong> low-quality productivity for consumers. At Alice Bay,Washington, S. anglica meadows had over 10 times <strong>the</strong>aboveground biomass as uninvaded mudflat (Hellquist2005). In Willapa Bay, Washington, colonization <strong>of</strong> Zosterajaponica and Spartina alterniflora has increased primaryproductivity by more than 50% in intertidal mudflats(Ruesink et al. 2006).- 153 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaAs noted for introduced Zostera japonica (Hahn 2003),<strong>the</strong> invasion <strong>of</strong> S. anglica could potentially alter detritaldecomposition rates and detrital quality within estuariesdominated by native eelgrass, Zostera marina. LikeSpartina alterniflora that has higher C:N ratios than Z.marina (Ruesink et al. 2006), mudflats colonized by S.anglica have higher C:N ratios than those with nativeproducers (Hellquist 2005). Changes in detrital availabilityand quality may influence <strong>the</strong> feeding patterns <strong>of</strong> residentdetrital consumers. The S. anglica invasion provides aunique opportunity to understand <strong>the</strong> trophic integration <strong>of</strong>an invasive producer in an estuarine ecosystem.Estuarine food webs are complex due to <strong>the</strong>predominance <strong>of</strong> detrital pathways, <strong>the</strong> abundance <strong>of</strong>potential productivity sources for consumption, omnivory,spatial heterogenity, and opportunistic feeding <strong>of</strong> consumersthat may change throughout <strong>the</strong> course <strong>of</strong> <strong>the</strong> year (Deeganand Garritt 1997; Riera et al. 1999). The use <strong>of</strong> multiplestable isotopes (especially δ 13 C and δ 34 S) can providevaluable insight into how Spartina spp. may contribute to<strong>the</strong> diets <strong>of</strong> estuarine consumers (Peterson et al. 1986;Peterson and Howarth 1987; Deegan and Garritt 1997;Valiela et al. 2000; Connolly et al. 2004). For example,13 C/ 12 C isotope ratios can distinguish C 3 from C 4 vegetationor marine algae from terrestrial C 3 vegetation (Fry and Sherr1984, Deegan and Garritt 1997). In marine studies, 34 S/ 32 Sisotope ratios are especially useful to distinguish sources <strong>of</strong>productivity because SO 2– 4 and HS – that are used by plantshave distinct isotopic signatures (Fry et al. 1982; Trust andFry 1992; Connolly et al. 2004).Due to <strong>the</strong> consistency <strong>of</strong> isotopic signatures fromproducers to consumers, trophic relationships can bediscerned from stable isotopic data that o<strong>the</strong>rwise mayremain virtually unknown (Fry and Sherr 1984; Peterson etal. 1985). Once isotopic data <strong>of</strong> consumers and potentialdietary producers is obtained, stable isotopic mixing modelscan be used to estimate <strong>the</strong> proportion <strong>of</strong> sources (producers)that contribute to <strong>the</strong> composition <strong>of</strong> a mixture (consumers;Phillips 2001).We present data that describe <strong>the</strong> potential contributions<strong>of</strong> Spartina biomass to <strong>the</strong> diets <strong>of</strong> three bivalves in nor<strong>the</strong>rnPuget Sound through <strong>the</strong> use <strong>of</strong> δ 13 C and δ 34 S stable isotoperatios and mixing models calculated by <strong>the</strong> computerapplication IsoSource (Phillips and Gregg 2003). As adeposit feeder that scours surface sediments for organicmatter, we expected that <strong>the</strong> clam Macoma balthica wouldhave a Spartina contribution in its diet. For <strong>the</strong> filter feedersMya arenaria (s<strong>of</strong>t-shelled clam) and Mytilus sp. (mussel),we expected Spartina contributions to be minimal or entirelyabsent.MATERIALS AND METHODSSamples were collected in March 2003 at West Pass(Camano Island, Island County, Washington; 48º 15’ 19” N,122º 24’ 56” W). The West Pass and English Boomshoreline is an extensive area <strong>of</strong> mudflat fringed by nativesalt marsh and Spartina meadows located in south SkagitBay, just north <strong>of</strong> <strong>the</strong> original introduction site <strong>of</strong> Spartinaanglica near Stanwood, Washington. This area is dominatedby approximately 100 ha <strong>of</strong> S. anglica meadows that growalong mudflats adjacent to <strong>the</strong> West Pass channel. Producerswere randomly collected along transects and included C 3emergent vascular vegetation (Grindelia integrifolia andSalicornia virginica), submergent vascular vegetation(Zostera marina), macroalgae (Fucus spiralis), and C 4vascular plants. C 4 plants included S. anglica (living leafand standing dead tissue) and <strong>the</strong> native salt marsh grassDistichlis spicata.Three bivalve species (Macoma balthica, Mya arenaria,and Mytilus sp.) were chosen based on <strong>the</strong>ir feeding patternsand co-occurrence with Spartina. Macoma is a surfacedeposit feeder, Mya a filter feeder, and Mytilus is anepifaunal filter feeder (Incze et al. 1982; Dame 1996). BothMacoma and Mya burrow in sediments immediately adjacentto Spartina meadows or burrow among Spartina roots alongtidal channels. Mytilus was found along Spartina meadowstypically using Spartina root masses and stems as anchoringsubstrate. Bivalves and Spartina were randomly sampled on<strong>the</strong> same transect at <strong>the</strong> edge <strong>of</strong> a large tidal channel thatpasses through <strong>the</strong> largest Spartina meadow at West Pass.Following collection, bivalves were held for 24 hours toallow gut contents to clear. Plant and animal samples werefrozen, cleaned to remove sediment, epiphytes, andcarbonates (10% HCl acid washes) and dried in a dryingoven. Samples were ground into a fine powder prior toisotopic analysis. For Macoma all visceral mass tissue wasused for analysis, whereas for Mya and Mytilus adductormuscles were dissected for analysis. Sample sizes forMacoma, Mya, and Mytilus were 12, seven, and sevenindividuals respectively. Producer sample sizes ranged fromthree to seven collections.Samples were analyzed for δ 13 C at <strong>the</strong> Idaho StableIsotope Laboratory, University <strong>of</strong> Idaho, Moscow, Idaho.Continuous-flow stable isotopic analyses were conducted ona Finnigan-MAT, Delta+ isotope ratio mass spectrometer(Thermo Finnigan, Thermo Electron: Waltham,Massachusetts). Samples were flash-combusted in a NC2500 elemental analyzer (CE Instruments, Thermo Electron:Waltham, Massachusetts) interfaced with a Conflo II. Stableisotope ratios are <strong>the</strong> ratio (R) <strong>of</strong> 13 C/ 12 C or 34 S/ 32 Sexpressed in standard delta (δ) notation values in parts permille (‰) where δ 13 C or δ 34 S = [(R sample /R standard)-1 ]*1000.The δ 13 C and δ 15 N reference standard was acetanilide (δ 13 C= -26.07‰ ± 0.11‰; δ 15 N = -0.33 ± 0.17‰). Spinach wasused as an internal standard (δ 13 C = -26.2‰ ± 0.2‰).Sample analysis <strong>of</strong> δ 34 S was conducted by Iso-Analytical Limited (Sandbach, Cheshire, United Kingdom).The reference standard was NBS-127 (barium sulfateδ 34 S V-CDT = 20.3‰; IAEA, Vienna, Austria). Calibration and- 154 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartinacorrection standards were IAEA S-1 (silver sulfide δ 34 S V-CDT= -0.3‰) and Iso-Analytical IA-R025 (barium sulfateδ 34 S V-CDT = +8.53‰). IA-R025 (barium sulfate δ 34 S V-CDT =+8.53‰) and whale baleen (δ 34 S V-CDT = +16.30‰) wereused as internal standards.The IsoSource isotopic mixing model analysisapplication (www.epa.gov/wed/pages/models.htm; Phillipsand Gregg 2003) was used to estimate <strong>the</strong> feasiblecontributions <strong>of</strong> S. anglica and six o<strong>the</strong>r potential sources to<strong>the</strong> diets <strong>of</strong> Macoma, Mya, and Mytilus. IsoSource calculates<strong>the</strong> ranges <strong>of</strong> all possible source contributions when <strong>the</strong>number <strong>of</strong> sources placed in <strong>the</strong> mixing model exceeds n+1,where n is <strong>the</strong> number <strong>of</strong> isotopic tracers (e.g. δ 13 C or δ 34 S;Phillips and Gregg 2003). Therefore, IsoSource allowssource contributions to a mixture to be estimated regardless<strong>of</strong> <strong>the</strong> number <strong>of</strong> sources and isotopic tracers available. In<strong>the</strong>se circumstances unique solutions cannot be obtained dueto <strong>the</strong> use <strong>of</strong> large numbers <strong>of</strong> sources in <strong>the</strong> mixing model(Phillips and Gregg 2003). Since unique solutions cannot bedetermined, IsoSource output provides ranges <strong>of</strong> all possiblesource contributions to <strong>the</strong> mixture based on mass balancecalculations (Table 1; Phillips and Gregg 2003). The morenarrow <strong>the</strong> range <strong>of</strong> <strong>the</strong>se solutions (e.g 30%-50% vs. 10-70%), <strong>the</strong> more reliable <strong>the</strong> interpretation <strong>of</strong> sourcecontributions to <strong>the</strong> mixture can be. Thus, IsoSource is veryuseful for studies <strong>of</strong> trophic relationships where numerousfood sources could exist for a consumer. IsoSource has beenapplied to a variety <strong>of</strong> trophic studies inside and outside <strong>of</strong>estuaries (e.g. Felicetti et al. 2003; Melville and Connolly2003; Ben-David et al. 2004; Newsome et al. 2004).The carbon isotope ratios <strong>of</strong> <strong>the</strong> bivalves were adjustedby a 1‰ fractionation factor based on Peterson and Fry(1987), Vander Zanden and Rasmussen (1999), andMcCutchan et al. (2003). Sulfur appears to have little to n<strong>of</strong>ractionation associated with its movement through trophiclevels. Due to <strong>the</strong> scarcity <strong>of</strong> sulfur fractionation data(McCutchan et al. 2003), <strong>the</strong> typical assumption that <strong>the</strong>re isminimal fractionation <strong>of</strong> sulfur was applied (Peterson andHowarth 1987). A biomass contribution increment <strong>of</strong> 1%and a tolerance interval <strong>of</strong> 0.11‰ was used for all IsoSourceanalyses to insure that <strong>the</strong> full range <strong>of</strong> feasible solutionswas generated (Table 1).RESULTSMacoma (n=12) clustered close to <strong>the</strong> outer edge <strong>of</strong>mixing polygon in close proximity to living and deadSpartina tissue (Figure 1A). Only living tissue <strong>of</strong> Spartinawas considered to be a component <strong>of</strong> <strong>the</strong> diet <strong>of</strong> Macoma inall solutions <strong>of</strong> <strong>the</strong> mixing model with contributions rangingfrom 37-60% (Table 1). Five <strong>of</strong> <strong>the</strong> six productivity sourcessampled could be absent in <strong>the</strong> diet <strong>of</strong> Macoma (Table 1).Four Mya clustered near <strong>the</strong> center <strong>of</strong> <strong>the</strong> mixingpolygon while three o<strong>the</strong>r individuals clustered close toSpartina (Figure 1B). IsoSource indicated that Mya couldreceive 40-59% <strong>of</strong> its diet from living Spartina tissue andTable 1. IsoSource results. Ranges <strong>of</strong> feasible biomass source contributionsfor <strong>the</strong> each bivalve summarized from 34,854 mass balanced mixingmodel solutions (Macoma balthica), 23,996 solutions (Mya arenaria), and56,083 solutions (Mytilus sp.). Each distribution is defined by <strong>the</strong> minimumand maximum values as well as <strong>the</strong> 1 st and 99 th percentiles. The percentilesrepresent 98% <strong>of</strong> all <strong>the</strong> possible IsoSource solutions. The meanpercent <strong>of</strong> each source distribution is cited to show central tendency <strong>of</strong> <strong>the</strong>distribution only and should not be considered a point estimate <strong>of</strong> <strong>the</strong>source contribution to <strong>the</strong> diet <strong>of</strong> <strong>the</strong> bivalve.Source Consumer Biomass %Minimum 1% Mean 99% MaximumSpartina (living leaves) Macoma 37 41 52 59 60Mya 40 44 53 58 59Mytilus 19 24 36 43 44Spartina (standing dead) Macoma 0 0 15 38 46Mya 0 0 8 27 35Mytilus 0 0 11 36 46Distichlis spicata Macoma 0 0 10 26 31Mya 0 0 6 19 24Mytilus 0 0 8 25 32C 3 Vascular Plants Macoma 0 0 5 15 17Salicornia virginica Mya 3 5 14 23 26Grindelia integrifolia Mytilus 5 8 20 32 35Zostera marina Macoma 0 0 9 28 33Mya 0 0 10 32 40Mytilus 0 0 13 42 52Macroalgae Macoma 0 0 9 26 31Fucus spiralis Mya 0 0 9 30 38Mytilus 0 0 12 40 503-26% <strong>of</strong> its diet from C 3 vegetation. All o<strong>the</strong>r sources hadwide ranges <strong>of</strong> potential producer contributions that includedzero as likely dietary contributions (Table 1).Living Spartina biomass and C 3 vegetation were <strong>the</strong>only dietary sources that IsoSource included within all <strong>of</strong> <strong>the</strong>feasible solutions for Mytilus. Living Spartina couldcontribute 19-44% to <strong>the</strong> diet <strong>of</strong> Mytilus, whereas C 3vegetation may contribute 5-35% <strong>of</strong> its diet. All o<strong>the</strong>rpotential sources included 0% as <strong>the</strong> most frequent feasiblecontribution (Table 1). Six <strong>of</strong> <strong>the</strong> seven individuals <strong>of</strong>Mytilus clustered in <strong>the</strong> center <strong>of</strong> <strong>the</strong> mixing polygon whileone individual was more depleted in its δ 34 S signature(Fig. 1C).DISCUSSIONOur mixing models present evidence that Spartinaanglica is consumed by bivalves. The use <strong>of</strong> Spartina byMacoma is particularly convincing based on <strong>the</strong> constrainedmixing model solution and <strong>the</strong> proximity <strong>of</strong> individuals to<strong>the</strong> isotopic signatures <strong>of</strong> living and dead Spartina (Fig. 1A).The mixing model for Mya is also convincing, although <strong>the</strong>δ 13 C values tend to be more depleted than those <strong>of</strong> Macomapotentially indicating smaller Spartina contributions to itsdiet compared to Macoma. With <strong>the</strong> exception <strong>of</strong> oneoutlying individual, Mytilus had <strong>the</strong> most consistent isotopicsignatures for δ 13 C and δ 34 S, and <strong>the</strong> lowest ranges <strong>of</strong>potential Spartina contributions. As a suspension feedingmussel, Mytilus should be drawing its food from <strong>the</strong>- 155 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaFig. 1. Dual isotope (δ 34 Sand δ 13 C) scatterplots <strong>of</strong> <strong>the</strong> three bivalves andsix producers sampled at West Pass, Island County, Washington. Eachopen circle is an individual consumer. A. Macoma balthica (n=12). B.Mya arenaria (n=7). C. Mytilus sp. (n=7). Producers (mean±SE) are <strong>the</strong>same for each panel: C 3 (pooled values <strong>of</strong> Salicornia virginica and Grindeliaintegrifolia: diamond), Fucus spiralis (closed circle), Zostera marina(square), Distichlis spicata (x), Standing dead Spartina (inverted triangle),Living Spartina leaves (triangle). Source sample sizes range from n=3(Zostera) to n=7 (C 3 plants). Error bars are plotted for SE values > 1 ‰.predominant sources <strong>of</strong> carbon available in <strong>the</strong> seston(Ruckelshaus et al. 1993). However, <strong>the</strong> consumption <strong>of</strong>suspended particulate organic matter (SPOM) by <strong>the</strong>sebivalves at West Pass during additional 2003 samplingperiods was surprisingly inconclusive despite using δ 13 C,δ 15 N and δ 34 S to simultaneously estimate dietarycontributions (Hellquist 2005).Previous trophic studies have provided a variety <strong>of</strong>conclusions about <strong>the</strong> importance <strong>of</strong> detrital Spartina inestuarine food webs. These results differ depending onconsumer trophic status and location. For example, filterfeeders including Geukensia demissa (ribbed mussel;Peterson et al. 1985, 1986; Peterson and Howarth 1987),Crassostrea virginica (oyster; Peterson et al. 1986) and Myaarenaria (Peterson et al. 1986; Peterson and Howarth 1987)have been identified as using Spartina biomass in <strong>the</strong>ir diets.The mud snail (Ilyanassa obsoleta) which is a deposit feederalso appears to use Spartina in Massachusetts and NorthCarolina (Peterson et al. 1986; Peterson and Howarth 1987;Currin et al. 1995). Spartina species also contribute to <strong>the</strong>diets <strong>of</strong> fish in Massachusetts (Peterson et al. 1986) andsou<strong>the</strong>rn California (Kwak and Zedler 1997).However, in a study similar to those above, <strong>the</strong> role <strong>of</strong>vascular plants in <strong>the</strong> diets <strong>of</strong> over 50 consumers wasconsidered minimal or nonexistent in a Mississippi estuary(Sullivan and Moncreiff 1990). Instead, macroalgae andzooplankton appeared to be <strong>the</strong> major dietary components <strong>of</strong><strong>the</strong> consumers sampled. Even Geukensia demissa andCrassostrea virginica shown by Peterson et al. (1986) to useSpartina biomass seemed to have little input from vascularplant productivity (Sullivan and Moncreiff 1990).In France, S. anglica appears to be an unused source <strong>of</strong>productivity for bivalves (Riera et al. 1999). Despite <strong>the</strong>availability <strong>of</strong> S. anglica detrital matter, bivalves (Macomabalthica and Mytilus edulis) relied on suspended particulateorganic matter and benthic diatoms as <strong>the</strong>ir primary dietarycomponents. Mytilus edulis had a diet dominated by 65%phytoplankton and 35% diatoms (Riera et al. 1999) whereas<strong>the</strong> diet <strong>of</strong> M. balthica was estimated to be composed <strong>of</strong>76% benthic diatoms and 24% phytoplankton. However,Jackson et al. (1986) used δ 13 C to estimate <strong>the</strong> dietarycontribution <strong>of</strong> S. anglica to M. balthica and <strong>the</strong>ir datasuggested that 34-50% <strong>of</strong> assimilated biomass consumed byM. balthica was Spartina, for a total estimate <strong>of</strong> 0.2-0.3grams <strong>of</strong> carbon per square meter per year (g C m -2 yr -1 )assimilated.Relatively few studies <strong>of</strong> Puget Sound trophic dynamicshave been conducted (e.g., Simenstad and Wissmar 1985;Ruckelshaus et al. 1993). In sou<strong>the</strong>rn Puget Sound estuariesand littoral beaches, detrital carbon was shown to originateprimarily from Zostera spp., epiphytic algae, andmacroalgae (Simenstad and Wissmar 1985). Our data alsoindicate small contributions to bivalve diets from nativemarsh plants with C 3 isotopic signatures (Table 1).Distichlis spicata, a C 4 salt marsh plant that grows inabundance adjacent to <strong>the</strong> Spartina meadows at West Pass,apparently contributes very little to bivalve diets (Table 1).The small contributions <strong>of</strong> Zostera marina to bivalve diets isprobably a function <strong>of</strong> its low abundance in <strong>the</strong> immediatevicinity <strong>of</strong> <strong>the</strong> sampling location.This study indicates that S. anglica biomass is a locallyimportant component <strong>of</strong> bivalve diets in nor<strong>the</strong>rn PugetSound after being in <strong>the</strong> ecosystem for just over 40 years.The extent that Spartina may contribute to bivalve diets inour study is somewhat unexpected due to <strong>the</strong> recalcitrant- 156 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartinanature <strong>of</strong> Spartina tissue (Peterson and Howarth 1987). Of<strong>the</strong> two Spartina tissue types sampled, we were expectingmore constrained IsoSource solutions for standing deadSpartina contributions to bivalve diets since we expecteddead tissue to more closely match <strong>the</strong> detritus available tobivalves (Table 1). The living Spartina tissue samples in thisstudy were newly emerged leaves, whereas <strong>the</strong> deadSpartina samples consisted <strong>of</strong> standing dead Spartina stemsand leaves. Living and dead Spartina biomass were sampledto represent <strong>the</strong> end points <strong>of</strong> a range <strong>of</strong> potential Spartinaisotopic signatures during its life history. The range <strong>of</strong>potential intermediate isotopic values that may occur duringsenescence is apparent as <strong>the</strong> distance between <strong>the</strong>signatures <strong>of</strong> living and dead Spartina (Figs. 1A-C).It may seem contradictory that <strong>the</strong> mixing models in thisstudy indicate <strong>the</strong> use <strong>of</strong> Spartina living tissue when <strong>the</strong>bivalves sampled are sediment dwelling or epifaunaldetritivores. These estimates suggest that as Spartinadecomposes, plant fragments in <strong>the</strong> early stages <strong>of</strong> decaymay be consumed. Early during decay, Spartina would haveits greatest nutritional value and would be at its leastrecalcitrant since C:N ratios more than double from livingtissue to <strong>the</strong> standing dead tissue phase <strong>of</strong> Spartina(Hellquist 2005). Alternatively, <strong>the</strong> use <strong>of</strong> Spartina detritusby <strong>the</strong> bivalves may be <strong>the</strong> result <strong>of</strong> trophic modification <strong>of</strong><strong>the</strong> biomass as it passes through <strong>the</strong> detrital food web(Peterson et al. 1986).The spatial proximity <strong>of</strong> <strong>the</strong> bivalves to <strong>the</strong> edge <strong>of</strong> <strong>the</strong>Spartina meadow probably accounts for <strong>the</strong> higher thanexpected inputs <strong>of</strong> Spartina in <strong>the</strong> bivalve diets. Although asimilar study <strong>of</strong> S. anglica and bivalves did not find arelationship between Spartina consumption and proximity toSpartina (Riera et al. 1999), o<strong>the</strong>r work has linked spatialproximity <strong>of</strong> consumers to Spartina with <strong>the</strong> magnitude <strong>of</strong>Spartina dietary contributions (Peterson et al. 1986; Petersonand Howarth 1987). Local sources <strong>of</strong> organic matter were <strong>of</strong>major dietary importance to <strong>the</strong> consumers sampled in aMassachusetts estuary (Deegan and Garritt 1997). Forexample, consumers in <strong>the</strong> upper estuary relied onfreshwater marsh organic matter and oligohalinephytoplankton, whereas in <strong>the</strong> lower estuary marine sourceswere more important (e.g. benthic diatoms and salt marshvegetation). We sampled bivalves collected from a tidalcreek passing through an extensive Spartina meadow, whereSpartina was <strong>the</strong> only immediate source <strong>of</strong> vascular plantproductivity. It is probably not surprising that <strong>the</strong> bivalveshave such a large potential contribution <strong>of</strong> S. anglica to <strong>the</strong>irdiets. Spartina was also a locally important dietary sourcefor <strong>the</strong> mussel Geukensia demissa when sampled fromwithin a salt marsh tidal channel (Peterson et al. 1985,1986). In Padilla Bay, Washington <strong>the</strong> importance <strong>of</strong> locallyabundant productivity sources also has been described forMytilus edulis (Ruckelshaus et al. 1993).Like Spartina alterniflora in Willapa Bay, Washington(Ruesink et al. 2006), S. anglica produces biomassthroughout <strong>the</strong> growing season that senesceces during <strong>the</strong>autumn and over winter. Thus, S. anglica adds copiousamounts <strong>of</strong> wrack to intertidal habitats. Release <strong>of</strong> biomassinto <strong>the</strong> estuary as a consequence <strong>of</strong> control efforts may alsocontribute to <strong>the</strong> high levels <strong>of</strong> S. anglica in bivalve diets atWest Pass. The West Pass area contains <strong>the</strong> greatestconcentrations <strong>of</strong> S. anglica in Washington (Hacker et al.2001) and has been subjected to intense control efforts. In2003, West Pass and its environs were intensively controlledvia applications <strong>of</strong> herbicide as well as mowing. Thesecontrol treatments kill Spartina during <strong>the</strong> growing seasonand produce Spartina wrack much earlier than typicalsenescence. The less recalcitrant Spartina tissue that diesprematurely due to control activity enters <strong>the</strong> food web up t<strong>of</strong>our to five months earlier than is typical. This plant tissuemay <strong>the</strong>n decay faster during <strong>the</strong> warmer summer and earlyfall months. By March, <strong>the</strong> Spartina isotopic signaturewould be fully incorporated into consumer tissue.The contributions to bivalve diets in this study duringMarch 2003 are subject to some sampling limitations. Thelack <strong>of</strong> samples <strong>of</strong> benthic diatoms and SPOM unfortunatelyexcludes productivity sources that should logicallycontribute to bivalve diets. However, additional stableisotope ratio data from West Pass collected later in 2003using δ 13 C, δ 15 N, and δ 34 S provide additional evidence for<strong>the</strong> consumption <strong>of</strong> S. anglica by bivalves (Hellquist 2005).These data indicate that <strong>the</strong>re are seasonal patterns <strong>of</strong> S.anglica use in bivalve diets (Hellquist 2005). However,during <strong>the</strong>se two additional sampling periods that includesamples <strong>of</strong> sestonic and benthic production, <strong>the</strong> dietary role<strong>of</strong> <strong>the</strong>se sources was unexpectedly inconclusive (Hellquist2005).The calculations <strong>of</strong> IsoSource greatly enhance ourability to estimate source contributions where <strong>the</strong>re are moresources than elements analyzed (Phillips 2001; Phillips andGregg 2003). However, it is crucial to remember that <strong>the</strong>semixing models should be viewed as an index <strong>of</strong> dietaryconsumption, and not discrete point estimates (Ben-Davidand Schell 2001; Phillips and Gregg 2003). The uncertaintyassociated with <strong>the</strong> dietary estimates generated by IsoSource(Table 1) reinforces <strong>the</strong> inexact nature <strong>of</strong> isotopic data incases where <strong>the</strong>re are multiple productivity sources withsimilar isotopic ratios and where consumer omnivory isprevalent.Along with Hahn (2003), Hellquist (2005), and Ruesinket al. (2006), this research illustrates how ecosystemprocesses in Washington estuaries can be altered by invasiveplant species including Zostera japonica, S. anglica, and S.alterniflora. To our knowledge, this study is <strong>the</strong> first use <strong>of</strong>multiple stable isotope ratios to examine trophicrelationships in Puget Sound in relation to invasive Spartina.This study provides evidence that bivalves living in- 157 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaimmediate proximity to invasive S. anglica are using itsbiomass as a source <strong>of</strong> nutrition during winter months. As arelatively new addition to <strong>the</strong> estuarine flora <strong>of</strong> nor<strong>the</strong>rnPuget Sound, not only is S. anglica a productivity sourcethat was previously absent, but it is also an ecosystemengineer that will eventually alter sediment conditions to <strong>the</strong>detriment <strong>of</strong> <strong>the</strong> bivalve populations that currently use itsbiomass for nutrition (Hellquist 2005).ACKNOWLEDGMENTSFunding for this study was provided by: <strong>the</strong> UnitedStates Environmental Protection Agency (US EPA) Office<strong>of</strong> Research and Development, National Center forEnvironmental Research, Science to Achieve Results(STAR) Fellowship for Graduate Environmental Study toCEH; <strong>the</strong> Estuarine Reserves Division, Office <strong>of</strong> Ocean andCoastal Resource Management, National Ocean Service,National Oceanic and Atmospheric Administration, NationalEstuarine Research Reserve System Graduate StudentFellowship (NA07OR0262) at <strong>the</strong> Padilla Bay NationalEstuarine Research Reserve (PBNERR), Mt. Vernon, WA toCEH; <strong>the</strong> Washington State University (WSU) School <strong>of</strong>Biological Sciences Betty W. Higinbotham Trust. We arealso very grateful to <strong>the</strong> University <strong>of</strong> Idaho Stable IsotopeLaboratory (John Marshall and Robert Brander) and Iso-Analytical, Ltd (Sandbach, Cheshire, UK) for sampleanalysis. R. Dave Evans (WSU) provided essentiallaboratory logistical support. Donald Phillips (US EPA,Corvallis, OR) provided critical advice and helpfuldiscussion for IsoSource analyses. Denise Howe, SvenNelson, and Justin Snider provided assistance with samplepreparation. Doug Bulthuis and Sharon Riggs (PBNERR)provided logistical support throughout this research. KyleMurphy (Washington State Department <strong>of</strong> Agriculture)provided Spartina anglica research permitting support.Thanks also to Brian Maricle, Joel Pankey, Paul Rabie, <strong>the</strong>WSU-UI Stable Isotopes in Ecology seminar, and <strong>the</strong>anonymous reviewers for <strong>the</strong>ir helpful comments on drafts<strong>of</strong> this paper.REFERENCESBen-David, M., and D.M. Schell. 2001. Mixing models in analyses<strong>of</strong> diet using multiple stable isotopes: a response. Oecologia127:180-184.Ben-David, M., K. Titus, and L.R. Beier. 2004. Consumption <strong>of</strong>salmon by Alaskan brown bears: a trade-<strong>of</strong>f between nutritionalrequirements and <strong>the</strong> risk <strong>of</strong> infanticide? Oecologia 138:465-474.Connolly, R.M., M.A. Guest, A.J. Melville, and J.M. Oakes. 2004.Sulfur stable isotopes separate producers in marine food-webanalysis. Oecologia 138:161-167.Currin, C.A., S.Y. Newell, and H.W. Pearl. 1995. The role <strong>of</strong>standing dead Spartina alterniflora and benthic microalgae insalt marsh food webs: considerations based on multiple stableisotope analysis. Mar Ecol Prog Ser 121:99-116.Dame, R. F. 1996. Ecology <strong>of</strong> marine bivalves: an ecosystem approach.CRC Press, Boca Raton, FL.Deegan, L.A., and R.H. Garritt. 1997. Evidence for spatial variabilityin estuarine food webs. Mar Ecol Prog Ser 147:31-47.Felicetti, L.A., C.C. Schwartz, R.O. Rye, M.A. Haroldson, K.A.Gun<strong>the</strong>r, D.L. Phillips, and C.T. Robbins. 2003. Use <strong>of</strong> sulfurand nitrogen stable isotopes to determine <strong>the</strong> importance <strong>of</strong>whitebark pine nuts to Yellowstone grizzly bears. Can J Zool81:763-770.Fry, B., R.S. Scanlan, J.K. Winters, and P.L. Parker. 1982. Sulphuruptake by salt grasses, mangroves, and seagrasses in anaerobicsediments. Geochim Cosmochim Acta 46:1121-1124.Fry, B., and E.B. Sherr. 1984. d13C measurements as indicators <strong>of</strong>carbon flow in marine and freshwater ecosystems. Contrib MarSci 27:15-47.Grosholz, E. 2002. Ecological and evolutionary consequences <strong>of</strong>coastal invasions. Trends Ecol Evol 17:22-27.Hacker, S.D., D. Heimer, C.E. Hellquist, T.G. Reeder, B. Reeves,T.J. Riordan, and M.N. Dethier. 2001. A marine plant (Spartinaanglica) invades widely varying habitats: potential mechanisms<strong>of</strong> invasion and control. Biol Invasions 3:211-217.Hahn, D.R. 2003. Alteration <strong>of</strong> microbial community compositionand changes in decomposition associated with an invasive intertidalmacrophyte. Biol Invasions 5:45-51.Hedge, P., L.K. Kriwoken, and K. Patten. 2003. A review <strong>of</strong>Spartina management in Washington State, US. J. Aquat PlantManage 41:82-90.Hellquist, C.E. 2005. Community and trophic consequences <strong>of</strong> <strong>the</strong>introduction <strong>of</strong> Spartina anglica C.E. Hubb. (Poaceae: Englishcordgrass) in estuaries <strong>of</strong> Puget Sound, WA, USA. Ph.D. Dissertation.Washington State University, Pullman, WA.Incze, L.S., L.M. Mayer, E.B. Sherr, and S.A. Macko. 1982. Carboninputs to bivalve mollusks: a comparison <strong>of</strong> two estuaries.Can J Fish Aquat Sci 39:1348-1352.Jackson, D., D.D. Harkness, C.F. Mason, and S.P. Long. 1986.Spartina anglica as a carbon source for salt-marsh invertebrates:a study using d 13 C values. Oikos 46:163-170.Kwak, T.J., and J.B. Zedler. 1997. Food web analysis <strong>of</strong> sou<strong>the</strong>rnCalifornia coastal wetlands using multiple stable isotopes.Oecologia 110:262-277.Mack, R.N., D. Simberl<strong>of</strong>f, W.M. Lonsdale, H. Evans, M. Clout,and F.A. Bazzaz. 2000. Biotic invasions: causes, epidemiology,global consequences, and control. Ecol Appl 10:689-710.McCutchan, J.H., Jr., J.W. M. Lewis, C. Kendall, and C.C.McGrath. 2003. Variation in trophic shift for stable isotope ratios<strong>of</strong> carbon, nitrogen, and sulfur. Oikos 102:378-390.Melville, A.J., and R.M. Connolly. 2003. Spatial analysis <strong>of</strong> stableisotope data to determine primary sources <strong>of</strong> nutrition for fish.Oecologia 136:499-507.Newsome, S.D., D.L. Phillips, B.J. Culleton, T.P. Guilderson, andP.L. Koch. 2004. Dietary reconstruction <strong>of</strong> an early to middleHolocene human population from <strong>the</strong> central California coast:insights from advanced stable isotope mixing models. J ArchaeolSci 31:1101-1115.Peterson, B.J., and B. Fry. 1987. Stable isotopes in ecosystem studies.Annu Rev Ecol Syst 18:293-320.Peterson, B.J., and R.W. Howarth. 1987. Sulfur, carbon, and nitrogenisotopes used to trace organic matter flow in <strong>the</strong> salt-marshestuaries <strong>of</strong> Sapelo Island, Georgia. Limnol Oceanogr 32:1195-1213.Peterson, B.J., R.W. Howarth, and R.H. Garritt. 1985. Multiplestable isotopes used to trace <strong>the</strong> flow <strong>of</strong> organic matter in estuarinefood webs. Science 227:1361-1363.Peterson, B.J., R.W. Howarth, and R.H. Garritt. 1986. Sulfur andcarbon isotopes as tracers <strong>of</strong> salt-marsh organic matter flow.Ecology 67:865-874.Phillips, D.L. 2001. Mixing models in analyses <strong>of</strong> diet using multiplestable isotopes: a critique. Oecologia 127:166-170.- 158 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaPhillips, D.L., and J.W. Gregg. 2003. Source partitioning usingstable isotopes: coping with too many sources. Oecologia136:261-269.Riera, P., L.J. Stal, J. Nieuwenhuize, P. Richard, G. Blanchard, andF. Gentil. 1999. Determination <strong>of</strong> food sources for benthic invertebratesin a salt marsh (Aiguillon Bay, France) by carbon andnitrogen stable isotopes: importance <strong>of</strong> locally produced sources.Mar Ecol Prog Ser 187:301-307.Ruckelshaus, M.H., R.C. Wissmar, and C.A. Simenstad. 1993. Theimportance <strong>of</strong> autotroph distribution to mussel growth in a wellmixed,temperate estuary. Estuaries 16:898-912.Ruesink, J.L., B.E. Feist, C.J. Harvey, J.S. Hong, A.C. Trimble,and L.M. Wisehart. 2006. Changes in productivity associatedwith four introduced species: ecosystem transformation <strong>of</strong> a 'pristine'estuary. Mar Ecol Prog Ser 311: 203-215.Ruiz, G.M., J.T. Carlton, E. Grosholz, and A.H. Hines. 1997.Global invasions <strong>of</strong> marine and estuarine habitats by nonindigenousspecies: mechanisms, extent , and consequences. AmZool 37:621-632.Simenstad, C.A., and R.C. Wissmar. 1985. δ13C evidence <strong>of</strong> <strong>the</strong>origins and fates <strong>of</strong> organic carbon in estuarine and nearshorefood webs. Mar Ecol Prog Ser 22:141-152.Sullivan, M.J., and C.A. Moncreiff. 1990. Edaphic algae are animportant component <strong>of</strong> salt marsh food-webs: evidence frommultiple stable isotope analyses. Mar Ecol Prog Ser 62:149-159.Teal, J.M., and B.L. Howes. 2000. Salt marsh values: retrospectionfrom <strong>the</strong> end <strong>of</strong> <strong>the</strong> century. In: M. P. Weinstein and D. A. Kreeger,eds. Concepts and controversies in tidal marsh ecology,9-19. Kluwer Academic, Dordrecht, The Ne<strong>the</strong>rlands.Trust, B.A., and B. Fry. 1992. Stable sulphur isotopes in plants: areview. Plant Cell Environ 15:1105-1110.Valiela, I., M.L. Cole, J. McClelland, J. Hauxwell, J. Cebrian, andS.B. Joye. 2000. The role <strong>of</strong> salt marshes as part <strong>of</strong> coastal landscapes.In: M. P. Weinstein and D. A. Kreeger [eds.], Conceptsand controversies in tidal marsh ecology, 23-38. Kluwer Academic,Dordrecht, The Ne<strong>the</strong>rlands.Vander Zanden, M.J., and J.B. Rasmussen. 1999. Primary consumerδ 13 C and δ 15 N and <strong>the</strong> trophic position <strong>of</strong> aquatic consumers.Ecology 80:1395-1404.- 159 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaCONTRASTING EFFECTS OF SPARTINA FOLIOSA AND HYBRID SPARTINA ON BENTHICINVERTEBRATESE.D. BRUSATI 1,2 AND E.D. GROSHOLZ 11 Department <strong>of</strong> Environmental Science and Policy, University <strong>of</strong> California, Davis, One Shields Ave., Davis, CA, USA 956162 Present address: California Invasive Plant Council, 1442-A Walnut St. #462, Berkeley, CA USA 94709;edbrusati@cal-ipc.orgIn San Francisco Bay, California, mudflats and native California cordgrass (Spartina foliosa)marshes are being invaded by a hybrid cordgrass formed by hybridization between S. foliosa andintroduced S. alterniflora. We investigated differences in vegetation and sediment structure, benthicinfauna, and food webs within native and invaded Spartina marshes between San Francisco Bay andBodega Bay, California. The greatest impact <strong>of</strong> hybrid Spartina in San Francisco Bay appears to beits alteration <strong>of</strong> habitat structure ra<strong>the</strong>r than food webs. Habitat structure differed significantlybetween native and hybrid Spartina. Hybrid Spartina produced greater biomass both above andbelow ground, and taller stem heights. Spartina foliosa contained significantly higher densities andbiomass <strong>of</strong> infaunal organisms in benthic cores than did mudflats, while densities and biomass <strong>of</strong>infauna in hybrid Spartina were lower than, or did not differ from, mudflats. Stable isotopes <strong>of</strong>carbon and nitrogen were used to examine whe<strong>the</strong>r macr<strong>of</strong>aunal food webs differ between native orhybrid Spartina and mudflats. Some consumers collected within Spartina showed evidence <strong>of</strong> a shiftin carbon isotope ratios indicating a possible increase in Spartina consumption within vegetation;however, <strong>the</strong> pattern was not consistent across species and sites. Due to <strong>the</strong> differences in <strong>the</strong>ireffects on infauna, hybrid Spartina and S. foliosa should not be considered equivalent for marshrestoration projects.Keywords: hybrid Spartina, Spartina foliosa, infauna, food webs, CaliforniaINTRODUCTIONIn San Francisco Bay, hybrids formed between nativeCalifornia cordgrass (S. foliosa) and introduced S.alterniflora accrete sediment, increase elevation compared tosurrounding mudflats, and significantly reduce light levelsunder <strong>the</strong>ir canopies (Neira et al. 2005). Based on tidallevels, hybrid Spartina would also be able to fill large areas<strong>of</strong> shallow outer coast bays if it establishes populations <strong>the</strong>re(Daehler and Strong 1997). The spread <strong>of</strong> hybrid Spartinathreatens flood control channels and habitat for migratingshorebirds, and may impact invertebrates that are food forbirds and fishes.Spartina cordgrasses may ei<strong>the</strong>r facilitate or inhibitinfauna depending on location (e.g., Capehart and Hackney1989, Netto and Lana 1999). The invasion <strong>of</strong> hybridSpartina provides an opportunity to examine howdifferences in structure between two closely-relatedecosystem engineers affect infaunal communities.Understanding differences between S. foliosa and hybridSpartina may help managers predict <strong>the</strong> impacts <strong>of</strong> <strong>the</strong>continued spread <strong>of</strong> hybrid Spartina.This study examined whe<strong>the</strong>r hybrid Spartina isecologically equivalent to native S. foliosa in its impacts oninfaunal and epifaunal invertebrates. We investigated twohypo<strong>the</strong>ses. First, we predicted that differences in structurebetween S. foliosa and hybrid Spartina will be reflected indifferences in infaunal density, biomass, and taxonomiccomposition. Second, we predicted that, due to <strong>the</strong> greateraboveground biomass produced by hybrid Spartina,organisms living within <strong>the</strong> hybrid will show greater use <strong>of</strong>hybrid as a food source than those in S. foliosa or mudflats,based on stable isotopes <strong>of</strong> carbon and nitrogen.METHODSStudy sites included five S. foliosa marshes and twohybrid Spartina marshes in nor<strong>the</strong>rn California. Spartinafoliosa sites included: China Camp State Park (38° 0.37’N,122° 28.66’W) on San Pablo Bay; Bolinas Lagoon (38°19.22’N, 122° 41.73’W); Shields Marsh (38° 5.35’N, 122°50.44’W) and Tom’s Point (38° 13.21’N, 122° 56.86’W) onTomales Bay; Drakes Estero (38° 5.36’N, 122° 55.86’W).Hybrid marshes were located on <strong>the</strong> eastern side <strong>of</strong> SanFrancisco Bay in San Lorenzo (Roberts Landing, 37°40.22’N 122° 09.70’W) and Alameda (Elsie Roemer BirdSanctuary, 37° 45.58’N 122° 28.80’W). The San Lorenzosite is unique in that it contains discrete patches <strong>of</strong> both S.foliosa and hybrid Spartina (genotypes confirmed by D.Ayres, University <strong>of</strong> California, Davis, pers. comm.).- 161 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaWe established ten study quadrats within Spartina ateach site, paired with ten on mudflats. These quadrats wereused for vegetation, sediment, and infauna sampling. Fordetails <strong>of</strong> methods, see Brusati and Grosholz (2006). Wemeasured stem heights and densities, aboveground biomass,and belowground biomass at each site, as well as sedimentcharacteristics such as organic matter content, bulk density,sediment porewater salinity, and oxidation-reductionpotential. Infauna cores (5 centimeters [cm] deep x 5 cmdiameter) were collected in winter and summer 2001-03.Cores were preserved in 8% formalin and organisms werecounted and weighed in <strong>the</strong> laboratory. We used t-tests ontransformed data to compare results between Spartina andmudflats.To analyze <strong>the</strong> effects <strong>of</strong> Spartina on invertebrate foodwebs, we collected species <strong>of</strong> infaunal and epifaunalinvertebrates in S. foliosa, hybrid Spartina, and mudflats.To understand how <strong>the</strong>se plants are incorporated into <strong>the</strong>invertebrate food webs we analyzed organisms for naturallyoccuring abundances <strong>of</strong> <strong>the</strong> stable isotope ratios 13 C and 15 N. The carbon signal reflects a weighted average <strong>of</strong> foodsources, with Spartina 13 C = -14 ‰, which is significantlymore enriched than o<strong>the</strong>r carbon sources in this habitat(Cloern et al. 2002). We predict that if hybrid Spartina isentering <strong>the</strong> food web, consumers collected within invadedareas will have a stronger Spartina signal than those frommudflats, with consumers from S. foliosa showing anintermediate signal.RESULTSHybrid Spartina differs from S. foliosa in its effects onhabitat structure, infaunal communities, and food webs.Hybrid Spartina produces more biomass, and <strong>the</strong>refore moredense structure on mudflats, than S. foliosa both above andbelow ground (Fig. 1). Alameda, where <strong>the</strong> invasion is 30years old, has greater belowground biomass than SanLorenzo, which was invaded in <strong>the</strong> 1990s. We found fewconsistent differences between sediment characteristics <strong>of</strong>hybrid or S. foliosa and mudflats. Spartina foliosa generallyhad significantly higher infaunal densities and biomass thanadjacent mudflats. Winter data are presented here, butsummer samples showed similar patterns. In contrast, hybridSpartina sediments never contained significantly greaterdensities or biomass than mudflats (Fig. 2, numbers abovebars are p-values, n = 10).Food web analysis also shows differences between S.foliosa and hybrid Spartina. Stable isotope results indicated thatsome species show evidence <strong>of</strong> a slight shift in 13 C towardSpartina for individuals collected within S. foliosa. (Fig. 3) Forexample, shore crabs (Hemigrapsus oregonensis) and Europeangreen crabs (Carcinus maenas) living within S. foliosa at ChinaCamp, and H. oregonensis in S. foliosa at Drakes Estero hadisotopic signatures more similar to S. foliosa than thosecollected on mudflats. At Alameda, <strong>the</strong>re was no significantdifference in isotopic signatures <strong>of</strong> C. maenas or Atlantic oysterdrills (Urosalpinx cinerea) from hybrid Spartina and mudflats.When results from all sites are compared, hybrid Spartina doesnot appear to produce a stronger shift in isotope signatures thanS. foliosa (Brusati and Grosholz 2009).DISCUSSIONOur results show that hybrid Spartina is not ecologicallyequivalent to <strong>the</strong> native cordgrass. While both S. foliosa andhybrid Spartina modify estuarine habitat, <strong>the</strong> magnitude <strong>of</strong><strong>the</strong> changes produced by hybrid Spartina are greater,resulting in qualitative differences in <strong>the</strong> infauna. Thestrongest differences were seen in <strong>the</strong> greater height andaboveground biomass <strong>of</strong> hybrid Spartina compared to S.foliosa. The greater aboveground biomass <strong>of</strong> hybrid Spartinais more than would be expected from differences in stemheight and density between <strong>the</strong> two species; it also reflects<strong>the</strong> fact that <strong>the</strong> hybrid’s stems are considerably thicker thanS. foliosa’s. Native S. foliosa faciliates infauna, possibly bystabilizing substrate or providing attachment sites for tubebuildingorganisms. In contrast, <strong>the</strong> dense roots andAboveground Biomass (g/m 2 )25002000150010005000Outer coastNodataSan FranciscoEstuaryBH BL DE SH TP CC SLF SLH AL---------------Spartina foliosa ----------------------- Hybrid2001 2002 2003Individuals/m 27000060000500004000030000200001000000.005Outer Coast0.2630.0020.000San FranciscoEstuary0.7240.114BL DE SM TP CC SL ALMudflat S. foliosa Hybrid0.001Fig. 1. Aboveground biomass <strong>of</strong> S. foliosa and hybrid Spartina 2001-03 (n= 10). Site abbreviations: BH = Bodega Harbor, BL = Bolinas Lagoon, DE= Drake’s Estero, SH = Shields March, TP = Tom’s Pt., CC = China Camp,SLF = Robert’s Landing S. foliosa, SLH = San Lorenzo hybrid, AL =Alameda.Fig. 2. Infaunal densities in S. foliosa exceed those in mudflats, whilehybrid Spartina never contained greater densities than mudflats (numbersabove bars are p-values, n = 10, winter 2001 data shown).- 162 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaB.A. 15 N (‰) 15 N (‰)1815121816141296China CampHemigrapsusBalanus CarcinusBalanus-22 -20 -18 -16 -14 -12 13 C (‰)BalanusCarcinusUrosalpinxAlameda10-20 -18 -16 -14 -12 13 C (‰)hybridCarcinusHemigrapsusS. foliosa mudflatCarcinusUrosalpinxBalanushybridSpartinamudflatS. foliosaFig. 3. Stable Isotope Results. Crabs (H. oregonensis, C. maenas) at ChinaCamp showed possible incorporation <strong>of</strong> S. foliosa (A), while nei<strong>the</strong>r C.maenas nor oyster drills (Urosalpinx cinerea) at Alameda appeared to incorporatehybrid Spartina (B).rhizomes <strong>of</strong> hybrid Spartina pre-empt space needed byinfauna so that invertebrates are replaced by plant material.Our stable isotope results show that <strong>the</strong> high biomassproduced by hybrid Spartina does not translate intoincreased incorporation <strong>of</strong> Spartina in <strong>the</strong> infaunal food web.Hybrid Spartina’s thick stems may require a longer time todecompose than <strong>the</strong> thinner stems <strong>of</strong> S. foliosa. As a result,carbon is held in hybrid Spartina ra<strong>the</strong>r than transferred tohigher trophic levels.The change from <strong>the</strong> relatively open canopy <strong>of</strong> S.foliosa to <strong>the</strong> more denser stems <strong>of</strong> hybrid Spartina mayhave indirect consequences for o<strong>the</strong>r primary producers thatare important to estuarine food webs, especially benthicmicroalgae. On <strong>the</strong> Atlantic coast, microalgal productionwas lower underneath <strong>the</strong> canopy <strong>of</strong> tall form S. alterniflorathan under <strong>the</strong> shorter dwarf form (Sullivan and Currin2000), suggesting a potential decrease in microalgalproduction in nor<strong>the</strong>rn California as hybrid Spartina replacesshorter S. foliosa.Comparing effects <strong>of</strong> native S. foliosa and hybridSpartina may help resource managers predict <strong>the</strong> impact ifhybrid Spartina establishes populations in coastal bays north<strong>of</strong> San Francisco, while data from S. foliosa marshes mayprovide reference information for evaluating restorationprojects. Infauna are major food sources for shorebirds andnative fishes, recycle carbon by breaking down plantdetritus, and move sediment by bioturbation and suspensionfeeding (Levin et al. 2001). Therefore, if replacement <strong>of</strong> S.foliosa with hybrid Spartina leads to relatively lowerinfaunal densities, consequences for <strong>the</strong> food web may reachbeyond <strong>the</strong> direct structural changes caused by <strong>the</strong> plants.ACKNOWLEDGMENTSWe thank <strong>the</strong> following for <strong>the</strong>ir assistance with thisproject. Funding: Canon National Parks Science ScholarsProgram (E.B.), University <strong>of</strong> California CoastalEnvironmental Quality Initiative (E.B.), UC-Davis PublicService Research Group (E.B.), Bodega Marine Laboratory,Sigma Xi (E.B.), Jastro Shields grant (E.B.), NSFBiocomplexity Program (DEB-0083583) (E.G.). Field andlab assistance: D. Ayres, L. Harris, C. Black, R. Blake, A. C.Tyler, N. Rayl, S. Norton, T. Dillon, and J. McCoy. Siteaccess: Point Reyes National Seashore (permit PORE-2001-SCI-0026), Gulf <strong>of</strong> <strong>the</strong> Farallones National MarineSanctuary (permit GFNMS-2001-004), Cypress GrovePreserve, China Camp State Park, Marin County OpenSpace District, and East Bay Regional Park District.REFERENCESBrusati, E.D., and E.D. Grosholz. 2006. Native and introducedecosystem engineers produce contrasting effects on estuarineinfaunal communities. Biological Invasions 8:683-695.Brusati, E.D., and E.D. Grosholz. 2009. Does <strong>the</strong> invasion <strong>of</strong>hybrid cordgrass change estuarine food webs? BiologicalInvasions 11:917-926.Capehart, A.A., and C.T. Hackney. 1989. The potential role <strong>of</strong>roots and rhizomes in structuring salt-marsh benthiccommunities. Estuaries 12:119-122.Daehler, C.C., and D.R. Strong. 1996. Status, prediction, andprevention <strong>of</strong> introduced cordgrass Spartina spp. invasions inPacific estuaries, USA. Biological Conservation 78:51-58.Levin, L.A., D.F. Boesch, A. Covich, C. Dahm, C. Erseus, K.C.Ewel, R.T. Kneib, A. Moldenke, and J.M. Weslawski. 2001. Thefunction <strong>of</strong> marine critical transition zones and <strong>the</strong> importance <strong>of</strong>sediment biodiversity. Ecosystems 4:430-451.Neira, C., L.A. Levin, and E.D. Grosholz. 2005. Benthicmacroalgal communities <strong>of</strong> three sites in San Francisco Bayinvaded by hybrid Spartina with comparison to uninvadedhabitats. Marine Ecology Progress Series 292:111-126.Netto, S.A., and P.C. Lana. 1999. The role <strong>of</strong> above- and belowgroundcomponents <strong>of</strong> Spartina alterniflora (Loisel) and detritusbiomass in structuring macrobenthic associations <strong>of</strong> ParanaguaBay (SE, Brazil). Hydrobiologia 400:167-177.Sullivan, M.J., and C.A. Currin. 2000. Community structure andfunctional dynamics <strong>of</strong> benthic microalgae in salt marshes. In:Weinstein, M.P., and D.A. Kreeger. Concepts and Controversiesin Tidal Marsh Ecology. pp. 81-106. Kluwer AcademicPublishers. Dordrecht, The Ne<strong>the</strong>rlands.- 163 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaIMPACTS OF BENTHIC INVERTEBRATES ON SEDIMENT POREWATER AMMONIUM ANDSULFIDE:CONSEQUENCES FOR SPARTINA SEEDLING GROWTHU.H. MAHL 1,2 ,A.C.TYLER 1,3 AND E.D. GROSHOLZ 11 Department <strong>of</strong> Environmental Science and Policy, University <strong>of</strong> California, Davis, CA 956162 Current address: Department <strong>of</strong> Biology, University <strong>of</strong> Notre Dame, South Bend, IN 46556; umahl@nd.edu3 Current address: School <strong>of</strong> Biological and Medical Sciences, Rochester Institute <strong>of</strong> Technology, Rochester, NY 14623Since its introduction in <strong>the</strong> 1970s, Spartina alterniflora and its hybrids have rapidly invaded <strong>the</strong>intertidal zone <strong>of</strong> San Francisco Bay. We know relatively little about <strong>the</strong> biotic and abiotic factorsthat have ei<strong>the</strong>r facilitated or hindered this invasion. In its native range, nitrogen availability andsediment anoxia are known to limit <strong>the</strong> growth <strong>of</strong> S. alterniflora. Benthic invertebrates may alterporewater concentrations <strong>of</strong> both ammonium and soluble sulfide, and <strong>the</strong>reby indirectly influence S.alterniflora success. In order to better understand how <strong>the</strong> resident macr<strong>of</strong>aunal invertebratecommunity may impact <strong>the</strong> expansion <strong>of</strong> S. alterniflora into <strong>the</strong> intertidal zone, we examined howcommon invertebrates influence concentrations <strong>of</strong> ammonium and soluble sulfide in sedimentporewater. We conducted laboratory microcosm experiments to determine how species from threefunctional feeding groups <strong>of</strong> benthic invertebrates (subsurface deposit feeders, surface depositfeeders and surface grazers) affect porewater in unvegetated areas adjacent to Spartina-invadedareas. In a separate experiment, we examined <strong>the</strong> growth <strong>of</strong> S. alterniflora seedlings in <strong>the</strong> presence<strong>of</strong> individual invertebrates. Relative to microcosms without invertebrates, where porewaterammonium was fairly high (greater than 400 micromoles per liter [μM]), we found slightly lowerporewater ammonium in microcosms with Heteromastus filiformis (subsurface deposit feeder) orMacoma petalum (surface deposit feeder). Porewater sulfide was slightly higher in <strong>the</strong> presence <strong>of</strong>H. filiformis only. The significantly greater growth <strong>of</strong> S. alterniflora seedlings in <strong>the</strong> presence <strong>of</strong> M.petalum than in <strong>the</strong> presence <strong>of</strong> H. filiformis suggests that both high sulfide and high ammoniummay have been detrimental to seedling success. Thus, invertebrates may indirectly influence <strong>the</strong>success <strong>of</strong> S. alterniflora seedlings by altering porewater chemistry.Keywords: ammonium, benthic, functional group, invasive species, porewater, Spartina, sulfideINTRODUCTIONAmong <strong>the</strong> greatest threats to natural ecosystems areinvasions by non-native species (Drake et al. 1989;Vitousek et al. 1997). However, <strong>the</strong> mechanisms thatdetermine <strong>the</strong> success or failure <strong>of</strong> new invasions are <strong>of</strong>tenpoorly understood. The invasion <strong>of</strong> Pacific Coast estuariesby Spartina alterniflora and its hybrids (S. alterniflora xnative Spartina foliosa) provides some insight into howSpartina can rapidly change benthic communities throughaltering physico-chemical properties <strong>of</strong> <strong>the</strong> habitat. Theshift from native mudflats and relatively diverse uppermarsh habitats to dense Spartina meadows followinginvasion has pr<strong>of</strong>ound effects on <strong>the</strong> functional diversity <strong>of</strong>benthic invertebrates, nutrient cycling, and o<strong>the</strong>r physicalprocesses (Neira et al. 2005, 2006; Levin et al. 2006).However, we still have little information on what biotic andabiotic factors may contribute to <strong>the</strong> great success <strong>of</strong>Spartina in <strong>the</strong> intertidal zone <strong>of</strong> San Francisco Bay.Determining which factors most strongly influence seedlingestablishment will greatly improve our understanding <strong>of</strong> <strong>the</strong>dynamics <strong>of</strong> <strong>the</strong> invasion.Previous work has shown that nutrient limitation andanoxia affect <strong>the</strong> growth and survival <strong>of</strong> native S.alterniflora in Atlantic coast salt marshes (e.g., Gallagher1975; King et al. 1982; DeLaune et al. 1983; Dai andWiegert 1997) and that nutrient limitation may be especiallysevere during <strong>the</strong> early stages <strong>of</strong> marsh development in bothnative and invaded marshes (Tyler et al. 2003, 2007).Benthic invertebrates can significantly impact <strong>the</strong>concentration <strong>of</strong> solutes in porewater and nutrient cyclingthrough bioturbation, burrow ventilation and consumption <strong>of</strong>organic matter (e.g. Aller 1982, Peterson and Heck 1999;Christensen et al. 2000). However, research has not beendone to determine whe<strong>the</strong>r <strong>the</strong> effects <strong>of</strong> benthiccommunities on <strong>the</strong> concentration <strong>of</strong> nutrients (e.g.,ammonium) and toxic metabolites (e.g., soluble sulfide) inporewater can affect <strong>the</strong> success <strong>of</strong> invasive plants like S.alterniflora.As part <strong>of</strong> an investigation linking benthic communitystructure to <strong>the</strong> establishment <strong>of</strong> S. alterniflora seedlings, weperformed two pilot microcosm experiments. The objective<strong>of</strong> <strong>the</strong> first experiment was to determine whe<strong>the</strong>r commonbenthic invertebrates from three different functional feeding- 165 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinagroups affect porewater ammonium and soluble sulfidewithout plants present. The objective <strong>of</strong> <strong>the</strong> secondexperiment was to determine whe<strong>the</strong>r <strong>the</strong>se same speciesinfluenced <strong>the</strong> growth <strong>of</strong> S. alterniflora seedlings. Wepredicted that <strong>the</strong>se three functional groups would havedifferent effects on porewater chemistry resulting indifferential growth <strong>of</strong> S. alterniflora seedlings.METHODSWe conducted microcosm experiments in anenvironmental chamber with a constant temperature <strong>of</strong> 14°Cand a 12-hour light cycle. All organisms and sedimentswere collected from mudflats near <strong>the</strong> edge <strong>of</strong> a hybridSpartina marsh at <strong>the</strong> Elsie Roemer Bird Sanctuary,Alameda Island, San Francisco Bay, California.Microcosms consisted <strong>of</strong> clear polycarbonate tubes (9.3centimeters inside diameter [cm I.D.] x 30 cm height [H])filled with two layers <strong>of</strong> homogenized, defaunated sediment.The lower layer (12 cm) was seived to 500 micrometers(μm) and <strong>the</strong>n frozen for two weeks to kill any organismsthat passed through <strong>the</strong> seive. The surface layer (3 cm),which was seived to 300 μm, but unfrozen, inoculated <strong>the</strong>core with natural micr<strong>of</strong>lora. Following reconstruction, <strong>the</strong>cores acclimated in <strong>the</strong> environmental chamber for threedays. Organic matter (one gram [g] dried, ground Ulva sp.)was added to <strong>the</strong> surface <strong>of</strong> each microcosm one day prior toorganism addition. To simulate tidal inundation, weconstructed an elaborate, automated system that filled eachchamber with sea water (32 parts per thousand [ppt])halfway through <strong>the</strong> light cycle and drained each chamberhalfway through <strong>the</strong> dark cycle each day.In <strong>the</strong> first experiment, we tested <strong>the</strong> effects <strong>of</strong> threefunctional feeding groups (subsurface deposit feeders,surface deposit feeders, and surface grazers) on porewaterammonium and soluble sulfide concentrations. Ourexperimental design consisted <strong>of</strong> a defaunated control andthree single species treatments (n = 6) representing each <strong>of</strong><strong>the</strong> three functional groups: <strong>the</strong> capitellid polychaeteHeteromastus filiformis (sub-surface deposit feeder), <strong>the</strong>nassariid snail Ilyanassa obsoleta (surface grazer) and <strong>the</strong>tellinid clam Macoma petalum (surface deposit feeder) wi<strong>the</strong>ight replicates <strong>of</strong> each treatment (Table 1). The numbers <strong>of</strong>individuals added to each microcosm were equivalent todensities at <strong>the</strong> Elsie Roemer site (see Neira et al. 2005).We extracted porewater from depths <strong>of</strong> 2, 4, and 7 cm usingTable 1: Design for Experiment 1. Density is <strong>the</strong> number <strong>of</strong> individuals forthat taxa per core (68 cm 2 ). N=6 for all treatments.Treatment Functional Group DensityDefaunated control - -Heteromastus filiformis Subsurface deposit feeder 30Ilyanassa obsolete Surface grazer 2Macoma petalum Surface deposit feeder 3a perforated stainless steel sampling probe (Berg andMcGla<strong>the</strong>ry 2001) at <strong>the</strong> termination <strong>of</strong> <strong>the</strong> experiment (14days). Porewater samples were analyzed for ammoniumusing <strong>the</strong> indophenol blue method (Solorzano 1969) and forsulfide using a modification <strong>of</strong> <strong>the</strong> method described byCline (1969). For statistical analyses <strong>of</strong> porewaterparameters, we calculated a mean value for all three depthsfor each replicate and analyzed differences amongtreatments separately for ammonium and sulfide usingANOVA.In <strong>the</strong> second experiment, we used <strong>the</strong> same invertebratetreatments and defaunated control as in <strong>the</strong> first experiment,but added one S. alterniflora seedling to each microcosm.We used nine replicates for defaunated controls and each <strong>of</strong><strong>the</strong> three macroinvertebrate treatments. The seedlings usedin this experiment were germinated in <strong>the</strong> greenhouse usingseeds from inflorescences collected in Willapa Bay,Washington. Thirty days after germination, <strong>the</strong> seedlingswere transferred to environmental chambers and acclimatedto experimental temperature and light regimes while salinity<strong>of</strong> <strong>the</strong> water in sediments was gradually increased from 0 to35 ppt over a period <strong>of</strong> 10 days. Seedlings were transplantedto microcosms after this acclimation period. After 14 days,we measured change in total leaf length (sum <strong>of</strong> length <strong>of</strong>individual leaves) and seedling biomass (aboveground andbelowground). We analyzed differences among treatmentsusing ANOVA.RESULTSFor <strong>the</strong> first experiment, preliminary analysis suggestedthat <strong>the</strong> affects <strong>of</strong> benthic invertebrates on porewaterammonium and sulfide concentrations differed amongfunctional groups. Sulfide concentrations were highest inmicrocosms containing H. filiformis but differences betweentreatments were not significant (p = 0.381; Fig. 1A). Theammonium concentration was highest in <strong>the</strong> control andlowest in microcosms containing M. petalum, but again <strong>the</strong>differences were not significant (p = 0.285; Fig. 1B). Thesurface grazer I. obsoleta had no obvious effects onammonium or sulfide.Overall, in <strong>the</strong> second experiment, seedlingestablishment was relatively poor, and <strong>the</strong>re was visibleyellowing and dehydration <strong>of</strong> leaves in all treatments,particularly microcosms containing H. filiformis. However,our preliminary analysis indicates that <strong>the</strong> surface depositfeeder treatment (M. petalum) had a positive effect onseedling establishment (Fig. 2A, 2B and 2C). In contrast tocontrols that had no significant growth, <strong>the</strong> total leaf lengthincreased approximately seven cm in microcosms containingM. petalum. Post-hoc Tukey tests indicated significantdifferences between M. petalum and H. filiformis (p =0.048). Aboveground and belowground biomass were alsohigher in microcosms containing M. petalum than in o<strong>the</strong>r- 166 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaA.sulfide (μM)35302520151050B.600ammonium (μM)5004003002001000CTRL HET ILY MACCTRL HET ILY MACFig. 1: Results from Experiment 1 by treatment (CTRL=defaunated control,HET=Heteromastus filiformis, ILY=Ilyanassa obsoleta, andMAC=Macoma petalum). A. Porewater sulfide concentration. B. Porewaterammonium concentration. N=6 for all treatments.treatments, although <strong>the</strong>se differences were not significant(Fig. 2B and 2C).DISCUSSIONOur preliminary results suggest that under laboratoryconditions benthic macroinvertebrates can affect bothporewater ammonium concentrations and Spartinaalterniflora seedling growth. We found that both H.filiformis and M. petalum lowered porewater ammoniumconcentrations relative to controls, while surface grazers hadno obvious effect on porewater ammonium or sulfides.Benthic invertebrates can decrease porewater ammoniumand soluble sulfide concentrations by flushing porewatersolutes from sediments during burrow construction andirrigation (Aller 1982; Christensen et al. 2000) and bystimulating oxidation-reduction reactions (Pelegri andBlackburn 1995; Rysgaard et al. 2000), but <strong>the</strong> magnitude <strong>of</strong>change is limited by <strong>the</strong> depth <strong>of</strong> <strong>the</strong> burrow (e.g., Aller1982; Francois et al. 2002; Michaud et al. 2006).Despite <strong>the</strong> negative effect on porewater ammoniumconcentrations, H. filiformis appeared to have a positiveeffect on soluble sulfide concentrations although this wasnot confirmed statistically. Past work has demonstrated thatinvertebrates can increase sulfate reduction rates in marinesediments (Hansen et al. 1996). A variety <strong>of</strong> mechanismsincluding removal <strong>of</strong> inhibitory metabolites, redistribution <strong>of</strong>particles, secretion <strong>of</strong> labile mucus along burrow walls, andtranslocation <strong>of</strong> labile organic matter from <strong>the</strong> surface to <strong>the</strong>deeper sediment during feeding could be responsible forincreased anaerobic metabolism and sulfide productionA.B.C.change in leaf length (cm)biomass (g)biomass (g)121086420-2-4-60.100.080.060.040.020.000.120.100.080.060.040.020.00CTRL HET ILY MACCTRL HET ILY MACCTRL HET ILY MACFig. 2: Results from Experiment 2 by treatment (CTRL=defaunated control,HET=Heteromastus filiformis, ILY=Ilyanassa obsoleta, MAC=Macomapetalum). A. Change in leaf length <strong>of</strong> Spartina alterniflora seedlings. B.Aboveground biomass <strong>of</strong> S. alterniflora seedlings. C. Belowground biomass<strong>of</strong> S. alterniflora seedlings. * indicates treatment means that differsignificantly from each o<strong>the</strong>r (post-hoc Tukey, p< 0.05). N = 9 for eachtreatment.(Wheatcr<strong>of</strong>t et al. 1994; Marinelli and Boudreau 1996; Allerand Aller 1998). However, it is not clear which mechanismswould account for <strong>the</strong> positive effect <strong>of</strong> H. filiformis onsoluble sulfides or why H. filiformis would have oppositeeffects on ammonium and sulfide.Significant differences between treatments for change intotal leaf length suggest that <strong>the</strong> benthic community has <strong>the</strong>- 167 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinapotential to substantially influence <strong>the</strong> establishment <strong>of</strong> S.alterniflora seedlings at <strong>the</strong> edge <strong>of</strong> <strong>the</strong> expanding marsh.Work from eastern U.S. marshes in <strong>the</strong> native range <strong>of</strong> S.alterniflora has shown growth and establishment can belimited by high porewater sulfide concentrations (Morris andDacey 1984; Bradley and Morris 1990) and by availability<strong>of</strong> ammonium in porewater (Tyler et al. 2003). At <strong>the</strong> sametime, very high ammonium can have a toxic effect on plantsalthough tolerances vary widely between species (Britto andKronzucker 2001). Ammonium toxicity for o<strong>the</strong>r marshspecies has been demonstrated at concentrations above 200μM (Tylova et al. 2008), although to our knowledge this hasnot been demonstrated for S. alterniflora. H. filiformisstimulated an increase in porewater sulfide concentrations, adecrease in porewater ammonium and a loss <strong>of</strong> S.alterniflora leaves (measured as <strong>the</strong> decrease in leaf length).In contrast, <strong>the</strong> low porewater sulfide and ammonium in <strong>the</strong>presence <strong>of</strong> M. petalum appeared to promote seedlingsuccess. In our experiments, concentrations <strong>of</strong> ammoniumexceeded 300 μM and it is possible that it had an inhibitoryeffect on seedling growth. While future work is needed toclarify <strong>the</strong> mechanisms linking benthic invertebratecommunities to seedling success, our results suggest thatboth low porewater sulfide and moderate levels <strong>of</strong> porewaterammonium are requisite for S. alterniflora seedling success.The spread <strong>of</strong> S. alterniflora and its hybrids in WestCoast estuaries may depend in part on complex interactionsbetween infaunal community structure and localbiogeochemical cycling that in turn dictate seedling success.As <strong>the</strong> invasion proceeds, changes in benthic communitystructure may ei<strong>the</strong>r facilitate or inhibit <strong>the</strong> success <strong>of</strong> S.alterniflora and its hybrids. Long-term goals for control,eradication, and restoration depend on an understanding <strong>of</strong><strong>the</strong> processes that control <strong>the</strong> establishment and expansion <strong>of</strong><strong>the</strong>se invaders.ACKNOWLEDGMENTSMany thanks to: Jay Stachowicz for generous use <strong>of</strong> hiswet lab, John Lambrinos for provision <strong>of</strong> seedlings, andRachael Blake, Nicole Christiansen, Nicole Smith, andCrystal Love for all <strong>the</strong>ir hard work. Funding was providedby NSF Biocomplexity Project (DEB-0083583).REFERENCESAller, R.C. 1982. The effects <strong>of</strong> macrobenthos on chemical properties<strong>of</strong> marine sediment and overlying water. In: P.L. McCall andM.J.S. Tevesz, eds. Animal-Sediment Relations. Plenum Press,New York, USA.Aller, R.C., and J.Y. Aller. 1998. The effect <strong>of</strong> bioirrigation intensityand solute exchange on diagenetic reaction rates in marinesediments. Journal <strong>of</strong> Marine Research 56:905–936.Berg P., and K.J. McGla<strong>the</strong>ry. 2001. A high-resolution pore watersampler for sandy sediments. Limnology and Oceanography46:203-10.Bradley, P.M., and J.T. Morris. 1990. Influence <strong>of</strong> oxygen andsulfide concentration on nitrogen uptake kinetics in Spartina alterniflora.Ecology 71:282-287.Britto, D.T., and H.J. Kronzucker. 2002. Journal <strong>of</strong> Plant Physiology159:567-584.Christensen, B., A. Vedel, and E. Kristensen. 2000. Carbon andnitrogen fluxes in sediment inhabited by suspension-feeding (Nereisdiversicolor) and non-suspension-feeding (N. virens) polychaetes.Marine Ecological Progress Series 192:203-217.Cline, J.D. 1969. Spectrophotometric determination <strong>of</strong> hydrogensulfide in natural waters. Limnology and Oceanography 14:454-8.Dai, T., and R.G. Wiegert. 1997. A field study <strong>of</strong> photosyn<strong>the</strong>ticcapacity and its response to nitrogen fertilization in Spartina alterniflora.Estuarine, Coastal, and Shelf Science 45:273-83.DeLaune, R.D., C.J. Smith, and W.H. Patrick, Jr. 1983. Relationship<strong>of</strong> marsh elevation, redox potential, and sulfide to Spartinaalterniflora productivity. Soil Science Society <strong>of</strong> America Journal47:930-5.Drake, J.A., H.A. Mooney, F. DiCastri, R.H. Groves, F.J. Kruger,M. Rejmanek, and M. Williamson, eds. 1989. Biological Invasions:A Global Perspective. John Wiley & Sons, Chichester,UK.Francois, F., M. Gerino, G. Stora, J. Durbec, and J. Poggiale. 2002.Functional approach to sediment reworking by gallery-formingmacrobenthic organisms: modeling and application with <strong>the</strong>polychaete Nereis diversicolor. Marine Ecological Progress Series229:127-136.Gallagher, J.L. 1975. Effect <strong>of</strong> an ammonium pulse on <strong>the</strong> growthand elemental composition <strong>of</strong> natural stands <strong>of</strong> Spartina alternifloraandJuncus roemerianus. American Journal <strong>of</strong> Botany62:644-648.Hansen, K., G. King, and E. Kristensen. 1996. Impact <strong>of</strong> <strong>the</strong> s<strong>of</strong>tshellclam Mya arenaria on sulfate reduction in an intertidalsediment. Aquatic Microbial Ecology 10:181 -194.King, G.M., M.J. Klug, R.G. Wiegert, and A.G. Chalmers. 1982.Relation <strong>of</strong> soil water movement and sulfide concentration toSpartina alterniflora production in a Georgia salt marsh. Science218:61-63.Levin, L.A., C. Neira, and E.D. Grosholz. 2006. Invasive cordgrassmodifies wetland trophic function. Ecology. 87:419-432.Marinelli, R.L., and B.P. Boudreau. 1996. An experimental andmodeling study <strong>of</strong> pH and related solutes in irrigated anoxiccoastal sediment. Journal <strong>of</strong> Marine Research 54:939-966.Michaud, E., G. Desrosiers, F. Mermillod-Blondin, B. Sundby, andG. Stora. 2006. The functional group approach to bioturbation:<strong>the</strong> effects <strong>of</strong> <strong>the</strong> Macoma balthica community on fluxes <strong>of</strong> nutrientsand dissolved organic carbon across <strong>the</strong> sediment-waterinterface. Journal <strong>of</strong> Experimental Marine Biology and Ecology.337:178-189.Morris, J.T., and J.W.H. Dacey. 1984. Effects <strong>of</strong> O 2 on ammoniumuptake and root respiration by Spartina alterniflora. AmericanJournal <strong>of</strong> Botany. 71:979-985.Neira, C., L.A. Levin, and E.D. Grosholz. 2005. Benthic macr<strong>of</strong>aunalcommunities <strong>of</strong> three sites in San Francisco Bay invaded byhybrid Spartina with comparison to uninvaded habitats. MarineEcological Progress Series 292:111-126.Neira, C., E.D. Grosholz, L.A. Levin, and R. Blake. 2006. Mechanismsgenerating modificaiton <strong>of</strong> benthos following tidal flat invasionby a Spartina hybrid. Ecological Applications 16:1391-1404.- 168 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaPelegri, S.P., and T.H. Blackburn. 1995. Effect <strong>of</strong> bioturbation byNereis sp., Mya arenaria and Cerastoderma sp. on nitrificationand denitrification in estuarine sediments. Ophelia 42:289-99.Peterson, B.J., and K.L. Heck. 1999. The potential for suspensionfeeding bivalves to increase seagrass productivity. Journal <strong>of</strong>Marine Biology and Ecology 240:37-52.Rysgaard, S., P.B. Christensen, M.V. Sorensen, P. Funch, and P.Berg. 2000. Marine mei<strong>of</strong>auna, carbon and nitrogen mineralizationin sandy and s<strong>of</strong>t sediments <strong>of</strong> Disko Bay, West Greenland.Aquatic Microbial Ecology 21:59-71.Solorzano, L. 1969. Determination <strong>of</strong> ammonia in natural waters by<strong>the</strong> phenolhypochlorite method. Limnology and Oceanography14:799-801.Tyler, A.C., J.G. Lambrinos, and E.D. Grosholz. 2007. Nitrogeninputs promote <strong>the</strong> spread <strong>of</strong> an invasive marsh grass. EcologicalApplications 17:1886-1898.Tyler, A.C., T.A. Mastronicola, and K.J. McGla<strong>the</strong>ry. 2003. Nitrogenfixation and nitrogen limitation <strong>of</strong> primary production alonga natural marsh chronosequence. Oecologia 136:431-438.Tylovia, E., L. Steinbachova, O. Votrubova, B. Lorenzen, and H.Brix. 2008. Different sensitivity <strong>of</strong> Phragmites australis andGlycera maxima to high availability <strong>of</strong> ammonium-N. AquaticBotany 88:93-98.Vitousek, P.M., H.A. Mooney, J. Lubchenco, and J.M. Melillo.1997. Human domination <strong>of</strong> Earth's ecosystems. Science277:494-499.Wheatcr<strong>of</strong>t, R.A., I. Olmez, and F.X. Pink. 1994. Particle bioturbationin Massachusetts Bay: preliminary results using a new deliberatetracer technique. Journal <strong>of</strong> Marine Research 52:1129-1150.- 169 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaQUANTIFYING THE POTENTIAL IMPACT OF THE SPARTINA INVASION ON INVERTEBRATEFOOD RESOURCES FOR FORAGING SHOREBIRDS IN SAN FRANCISCO BAYN. CHRISTIANSEN 1 , E.D. GROSHOLZ 2 AND P. ROSSODepartment <strong>of</strong> Environmental Science and Policy, University <strong>of</strong> California, Davis, One Shields Ave., Davis, CA 956161 n.a.christiansen@gmail.com; 2 tedgrosholz@ucdavis.eduThe extensive mudflats <strong>of</strong> <strong>the</strong> San Francisco Bay estuary are critical habitats and foraging groundsfor migratory shorebirds. However, San Francisco Bay has experienced invasion by hundreds <strong>of</strong>non-native introduced species, and its productive mudflats are now threatened by one <strong>of</strong> <strong>the</strong>se —hybrid Spartina alterniflora. To assess <strong>the</strong> potential effects <strong>of</strong> <strong>the</strong> invasive cordgrass on <strong>the</strong>shorebird foraging habitat, we quantified invertebrate abundance and biomass along transectsextending from <strong>the</strong> edge <strong>of</strong> <strong>the</strong> Spartina meadow to <strong>the</strong> lowest tide line at two locations in <strong>the</strong> bay,Robert’s Landing and Elsie Roemer. We also estimated <strong>the</strong> size <strong>of</strong> areas potentially invaded bySpartina using tidal elevation data based on LIDAR images and tidal inundation data taken fromcontinuous tidal height monitoring stations. We found that invertebrate biomass was greatest athigher tidal elevations nearer to <strong>the</strong> marsh, while abundance was greatest at intermediate distancesfrom <strong>the</strong> marsh. Current estimates <strong>of</strong> tolerance to tidal inundation indicate that <strong>the</strong> area colonized bySpartina could increase by as much as four to eight times its present size, depending on <strong>the</strong>individual site characteristics. These data suggest that <strong>the</strong>re is a high potential for losing significantamounts <strong>of</strong> valuable foraging habitat for shorebirds if Spartina extends its distribution to predictedtidal elevations in San Francisco Bay.Keywords: shorebirds, San Francisco Bay, Hybrid Spartina, infaunaINTRODUCTIONSan Francisco Bay is now home to more than 250species <strong>of</strong> non-native plants and animals (Cohen and Carlton1998). One <strong>of</strong> <strong>the</strong> most serious recent invasions has been <strong>the</strong>smooth cordgrass Spartina alterniflora, which is native to<strong>the</strong> eastern coast <strong>of</strong> <strong>the</strong> North America. Since itsintroduction, S. alterniflora has hybridized with <strong>the</strong> nativecordgrass S. foliosa creating a hybrid that has rapidlycolonized many areas <strong>of</strong> central and south San FranciscoBay (Daelher and Strong 1997, Ayres et al. 2004). Thehybrid can colonize open mudflats as well as out-competenative vegetation at higher tidal heights (Ayres et al. 2004).The ecological repercussions <strong>of</strong> this invasion are farreaching with widespread impacts on community structureand ecosystem function (Neira et al. 2005; 2006, Levin et al.2006).One <strong>of</strong> <strong>the</strong> most serious consequences <strong>of</strong> <strong>the</strong> hybrid Spartinainvasion is <strong>the</strong> loss <strong>of</strong> open mudflat habitat for shorebirds.San Francisco Bay, with its extensive mudflats, containscritical habitat for more than one million migratory shorebirdsand is one <strong>of</strong> <strong>the</strong> most important estuaries along <strong>the</strong>Pacific Flyway (Page et al. 1999). This important habitat isnow threatened by <strong>the</strong> smooth cordgrass. Shorebirds requireunvegetated mudflat habitats for foraging (Goss-Custard andMoser 1988), and invasive hybrid Spartina threatens tocolonize critical upper levels <strong>of</strong> this important habitat. As aresult <strong>of</strong> <strong>the</strong> rapid spread <strong>of</strong> hybrid Spartina, which increasedby 317% at sampled sites between 2001 and 2003(Ayres et al. 2004), by <strong>the</strong> end <strong>of</strong> 2003 it was estimated tooccupy nearly 800 hectares <strong>of</strong> bay habitat, up from less than200 acres in 2001 (Zaremba and McGowen 2004). Thisrapid spread suggests that <strong>the</strong> risk for shorebird habitat lossmay be substantial.To better estimate <strong>the</strong> potential value and extent <strong>of</strong> lostforaging habitat, we measured <strong>the</strong> biomass and abundance <strong>of</strong>invertebrates at different tidal elevations at two sites in SanFrancisco Bay. To accurately measure tidal elevation acrossbroad areas <strong>of</strong> <strong>the</strong> two study sites, we used LIDAR (LightDetection and Ranging), a flight-based radar system that canaccurately measure elevation with high spatial resolution.We used <strong>the</strong>se elevation data toge<strong>the</strong>r with tidal height datafrom continuous recording stations to estimate <strong>the</strong> arealikely to be invaded by Spartina based on current estimates<strong>of</strong> its inundation tolerance. Therefore, we were able toestimate <strong>the</strong> value and extent <strong>of</strong> habitat lost to shorebirdsunder different scenarios <strong>of</strong> Spartina spread. Thisinformation, when combined with <strong>the</strong> data on patterns <strong>of</strong>shorebird usage generated by PRBO Conservation Science(Stralberg et al. 2010) can help guide future eradicationefforts by highlighting areas <strong>of</strong> particular importance forshorebirds.- 171 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaMETHODSTo determine <strong>the</strong> distribution and abundance <strong>of</strong> keyinvertebrate taxa <strong>of</strong> importance for foraging shorebirds, wesampled infauna at two sites in central San Francisco Baywith open mudflats that are being invaded by hybridSpartina: Elsie Roemer Bird Sanctuary on Alameda Island(37º45’35’’N; 122º28’48’’W) and Roberts Landing in SanLorenzo (37º40’13’’N; 122º28’48’’W). At both sites, weestablished three transects spanning <strong>the</strong> entire tidal range <strong>of</strong><strong>the</strong> mudflat (several hundred meters) from <strong>the</strong> edge <strong>of</strong> <strong>the</strong>Spartina meadow to <strong>the</strong> channel. We collected sedimentcore samples from <strong>the</strong> following distances from <strong>the</strong> seawardedge <strong>of</strong> <strong>the</strong> hybrid Spartina meadow: 1, 10, 50, 100, and 200meters (m) and <strong>the</strong>reafter at approximately 100 m intervalsto within 50 m <strong>of</strong> <strong>the</strong> channel edge. At Elsie Roemersampling extended up to 350 m, whereas at Robert’sLanding, sampling extended up to 850 m from <strong>the</strong> meadow.At each sampling location, one infauna sample was collectedwith a 5 centimeters (cm) diameter core taken to 5 cm depth.The location <strong>of</strong> each sample site was recorded with ashoulder mounted GPS (Global Positioning System, TrimblePro-XR). In <strong>the</strong> lab, samples were sieved at 500 micrometers(μm) to collect organisms <strong>of</strong> importance for shorebirds. Allorganisms were fixed in 10% buffered formalin for at least24 hours, <strong>the</strong>n stored in 70% ethanol stained with rosebengal.We estimated <strong>the</strong> biomass <strong>of</strong> invertebrates for <strong>the</strong> threemost common phyla (>95% <strong>of</strong> <strong>the</strong> biomass): Annelida,Arthropoda (Crustacea), and Mollusca. We regressedbiomass for all taxa and for each <strong>of</strong> <strong>the</strong> three common phylaon distance from <strong>the</strong> edge <strong>of</strong> <strong>the</strong> Spartina meadow definedas <strong>the</strong> lower limit <strong>of</strong> contiguous aboveground vegetation.We used LIDAR images that resulted from a moreextensive flyover <strong>of</strong> San Francisco Bay, but includeddetailed views <strong>of</strong> <strong>the</strong> two study sites. Images permittedestimations <strong>of</strong> tidal elevations to within 0.1 m atapproximately 1 m spatial resolution in <strong>the</strong> x-y plane. Thesedata were used to develop images <strong>of</strong> <strong>the</strong> study site withwhich we could estimate area covered by Spartina assumingit could spread downward in tidal elevation to a particulartidal level. We used <strong>the</strong>se images toge<strong>the</strong>r withgeoreferenced sampling sites to determine <strong>the</strong> abundanceand distribution <strong>of</strong> invertebrates at particular tidal elevations.We calculated inundation as a function <strong>of</strong> tidal elevations,using continuous water level data available from <strong>the</strong> NationalOceanic and Atmospheric Administration (NOAA) usingstation # 9414750 Alameda, CA for both sites. Tidal data weremeasured in meters above mean low low water (MLLW) basedon <strong>the</strong> updated National Tidal Datum Epoch (NTDE). We usedhourly tidal data to estimate <strong>the</strong> percentage time for <strong>the</strong> year2004 that a given elevation determined from LIDAR imageswould be inundated.RESULTSInvertebrate biomass at higher tidal elevations wasapproximately triple <strong>the</strong> biomass at lower elevations whenboth sites were analyzed toge<strong>the</strong>r (Fig. 1) based on anexponential regression curve:y = 18.593 e -0.0032x , R² = 0.36 (1)However, <strong>the</strong> abundance <strong>of</strong> invertebrates was greatest atintermediate tidal elevations (Fig. 2), and <strong>the</strong> least at <strong>the</strong>lowest elevations based on a polynomial regression curve:Total Biomass (gm)60y = 18.593e -0.0032x50R 2 = 0.36344030201000 200 400 600 800 1000Distance from Marsh Edge (m)Fig. 1. Total invertebrate biomass as a function <strong>of</strong> <strong>the</strong> distance from <strong>the</strong>marsh edge at <strong>the</strong> current lower tidal limit <strong>of</strong> hybrid Spartina.Fig. 2. LIDAR image <strong>of</strong> <strong>the</strong> Elsie Roemer site showing lower extent <strong>of</strong>hybrid Spartina distribution (grey) and invertebrate sampling points alongtransects (dots).- 172 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaInundation Time vs. Tidal Height0.80Inundation Time (%)0.750.700.650.600.550.500.450.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2Tidal Height (m)Fig. 4. Plot <strong>of</strong> inundation time (%) calculated from continuous tidalheight data relative to estimated elevation from LIDAR data.Fig. 3. LIDAR image <strong>of</strong> Robert’s Landing showing lower extent <strong>of</strong> hybridSpartina distribution (grey) and invertebrate sampling points along transects(dots).y = -0.0006x² + 0.3889x + 56.787, R² = 0.1713 (2)The high abundances at intermediate tidal elevationswere due to high numbers <strong>of</strong> small juvenile clams (> 1 mm)that did not contribute very significant biomass. When weanalyzed <strong>the</strong> data by phyla, we found that annelids, mollusksand crustaceans generally showed similar patterns. Annelidsalso showed a substantial decline in biomass with tidalelevation:y = -0.0165x + 13.624, R 2 =0.12 (3)Molluscs and crustaceans showed a similar but morevariable decline (R 2

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinainvaded by hybrid Spartina could see a potential eightfoldincrease at Elsie Roemer and a fourfold expansion atRobert’s Landing. We emphasize that this is likely to be anupper estimate <strong>of</strong> <strong>the</strong> area invaded, given that it assumes thathybrid Spartina could withstand being inundatedapproximately 75% <strong>of</strong> <strong>the</strong> time, which is beyond currentestimates <strong>of</strong> tidal inundation tolerance. However, <strong>the</strong>seestimates are based in part on present distributions, whichare subject to change. These estimates also do not considerthat this tolerance may be under strong selection and mayevolve significantly over time. In any case, our data suggestthat much <strong>of</strong> <strong>the</strong> most important foraging habitat formigratory shorebirds in San Francisco Bay that lies between0.6-1.1 m MLLW may be lost over <strong>the</strong> next several years.ACKNOWLEDGMENTSWe would like to thank R. Blake, C. Love, U. Mahl, and C.Tyler for <strong>the</strong>ir help and labor.REFERENCESAyres, D.R., D.L. Smith, K. Zaremba, S. Klohr, and D.R. Strong.2004. Spread <strong>of</strong> exotic cordgrasses and hybrids (Spartina sp.) in<strong>the</strong> tidal marshes <strong>of</strong> San Francisco Bay, California, USA.Biological Invasions 6:221-231.Cohen, A.N. and J.T. Carlton. 1998. Accelerating invasion rate in ahighly invaded estuary. Science 279:555-558.Daelher, C.C., and D.R. Strong. 1997. Hybridization betweenintroduced smooth cordgrass (Spartina alterniflora; Poaceae)and native California cordgrass (S. foliosa) in San Francisco Bay,California, USA. American Journal <strong>of</strong> Botany 84:607-611.Goss-Custard, J.D., and M.E. Moser. 1988. Rates <strong>of</strong> change in <strong>the</strong>numbers <strong>of</strong> Dunlin Calidris-alpina wintering in British estuariesin relation to <strong>the</strong> spread <strong>of</strong> Spartina anglica. Journal <strong>of</strong> AppliedEcology 25:95-110.Levin, L.A., C. Neira and E.D. Grosholz. 2006. Invasive cordgrassmodifies wetland trophic function. Ecology 87:419-432.Neira, C., L.A. Levin, and E.D. Grosholz. 2005. Benthicmacr<strong>of</strong>aunal communities <strong>of</strong> three Spartina-hybrid invaded sitesin San Francisco Bay, with comparison to uninvaded habitats.Marine Ecology Progress Series 292:111-126.Neira, C., E.D. Grosholz, L.A. Levin, and R. Blake. 2006.Mechanisms generating modification <strong>of</strong> benthos following tidalflat invasion by a Spartina (alterniflora x foliosa) hybrid.Ecological Applications 16:1391-1404.Page, G.W., L.E. Stenzel, and J.E. Kjelmyr. 1999. Overview <strong>of</strong>shorebird abundance and distribution in wetlands <strong>of</strong> <strong>the</strong> PacificCoast <strong>of</strong> <strong>the</strong> contiguous United States. Condor 101:461-471.Stralberg, D., V. Toniolo, G.W. Page and L.E.Stenzel. 2010. Potentialimpacts <strong>of</strong> Spartina spread on shorebird populations in SouthSan Francisco Bay. In: Ayres, D.R., D.W. Kerr, S.D. Ericson andP.R. Ol<strong>of</strong>son, eds. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong>on Invasive Spartina, 2004 Nov 8-10, San Francisco,CA, USA. San Francisco Estuary Invasive Spartina Project <strong>of</strong> <strong>the</strong>California State Coastal Conservancy: Oakland, CA (this volume).Zaremba, K. and M.F. McGowan. 2004. San Francisco EstuaryInvasive Spartina Project Monitoring Report for 2003. SanFrancisco Estuary Invasive Spartina Project. (available fromhttp://www.spartina.org/project.htm).- 174 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaPOTENTIAL IMPACTS OF SPARTINA SPREAD ON SHOREBIRD POPULATIONS IN SOUTH SANFRANCISCO BAYD. STRALBERG 1 , V. TONIOLO 1,2 , G.W. PAGE 1 , AND L.E. STENZEL 11 PRBO Conservation Science, 3820 Cypress Drive #11, Petaluma, CA 94954, dstralberg@prbo.org2 Current address: Ocean Biogeochemistry Lab, Dept. <strong>of</strong> Environmental Earth System Science, Mitchell Bldg, A67, StanfordUniversity Stanford, CA 94305-2215San Francisco Bay holds 70% <strong>of</strong> California’s mudflats and provides habitat to more wintering andmigratory shorebirds than any o<strong>the</strong>r wetland along <strong>the</strong> Pacific coast <strong>of</strong> <strong>the</strong> contiguous U.S. The bay’smudflats are currently threatened by <strong>the</strong> spread <strong>of</strong> a non-native cordgrass, Spartina alterniflora, andassociated hybrids, which grow at lower elevations than <strong>the</strong> native S. foliosa and can render largemudflat areas effectively unavailable to shorebirds for foraging. Using shorebird survey data, tidestation data, and GIS-based habitat data, we analyzed <strong>the</strong> potential effect <strong>of</strong> S. alterniflora x foliosahybrids on shorebird habitat in <strong>the</strong> South Bay by creating grid-based spatial models <strong>of</strong> shorebirdhabitat value and potential Spartina spread. We developed three scenarios <strong>of</strong> potential habitat valueloss for shorebirds based on assumptions about <strong>the</strong> inundation tolerance <strong>of</strong> S. alterniflora and temporalavailability <strong>of</strong> mudflat resources. Predictions <strong>of</strong> habitat value loss across seasons ranged from27% to 80%. We identified <strong>the</strong> upper mudflats, due to <strong>the</strong>ir greater exposure time, and <strong>the</strong> east andsouth shore mudflats, due to <strong>the</strong> high numbers <strong>of</strong> birds detected <strong>the</strong>re, as <strong>the</strong> areas <strong>of</strong> highest valueto shorebirds in <strong>the</strong> South Bay. These areas also coincide with <strong>the</strong> areas <strong>of</strong> greatest Spartina invasionpotential.Keywords: Spartina alterniflora, mudflats, inundation tolerance, Charadrii, GISINTRODUCTIONThe San Francisco Bay estuary holds 70% <strong>of</strong> <strong>the</strong> mudflatsin California (Josselyn et al. 1990), providing habitatto over 350,000 migrating shorebirds (Charadrii) in <strong>the</strong>fall, over 325,000 in winter, and over 900,000 in <strong>the</strong> spring(based on single-day counts) (Stenzel et al. 2002). Along <strong>the</strong>Pacific coast <strong>of</strong> <strong>the</strong> contiguous United States alone (excludingAlaska), San Francisco Bay holds more shorebirds thanany o<strong>the</strong>r wetland in all seasons (Page et al. 1999). Surveydata suggest that San Francisco Bay holds over 50% <strong>of</strong> <strong>the</strong>Pacific coast populations <strong>of</strong> several shorebird species—11<strong>of</strong> <strong>the</strong> 12 most abundant species in fall, six <strong>of</strong> 13 in winter,and seven <strong>of</strong> 13 in spring (Page et al. 1999). Within <strong>the</strong> Bay,mudflats are <strong>the</strong> most important habitats for <strong>the</strong> majority <strong>of</strong><strong>the</strong>se species (Stenzel et al. 2002).San Francisco Bay’s mudflats are now threatened by<strong>the</strong> spread <strong>of</strong> cordgrass hybrids (Spartina alterniflora xfoliosa) (Ayres et al. 2008). The non-native S. alterniflorawas originally introduced to <strong>the</strong> South San Francisco Bay(South Bay) in <strong>the</strong> 1970s (Callaway and Josselyn 1992) andhybridized with native S. foliosa (Ayres et al. 1999). Spartinaalterniflora x foliosa hybrids exhibit higher tolerance totidal submersion and salinity, as well as higher growth andgermination rates than <strong>the</strong> native S. foliosa (Callaway andJosselyn 1992; Daehler and Strong 1997; Anttila et al. 1998;Collins 2002). Projections from sampling sites suggested thatnearly 800 hectares (ha) <strong>of</strong> South Bay marshes, channels,and mudflats had been invaded by 2003, an increase <strong>of</strong> morethan 300% since 2001 (Zaremba et al., this volume). Fur<strong>the</strong>rspread <strong>of</strong> hybrid Spartina on <strong>the</strong> Bay’s mudflats poses a greatthreat to shorebirds, which cannot forage in areas <strong>of</strong> densegrowth (Josselyn 1983, Evans 1986, Goss-Custard and Moser1988, White 1995).In Willapa Bay, Washington, which was initiallyinvaded in <strong>the</strong> late 19th century, S. alterniflora recentlyexperienced explosive growth, tripling its areal extentbetween 1994 and 2002 and converting many mudflats toSpartina marshes (Buchanan 2003; Civille 2005). Surveys<strong>of</strong> Willapa Bay birds suggested a reduction in shorebirdnumbers by as much as 67% and foraging time by as muchas 50% (Jaques 2002). Unlike San Francisco Bay, WillapaBay is outside <strong>the</strong> range <strong>of</strong> native Spartina and <strong>the</strong>reforehas no hybrid Spartina, which has been found to exhibitmuch more rapid rates <strong>of</strong> lateral expansion than its parentalspecies (Ayres et al. 2008). Thus <strong>the</strong> invasion potential inSan Francisco Bay is thought to be substantially greaterthan in Willapa Bay.While San Francisco Bay and Willapa Bay also differin <strong>the</strong>ir bathymetry and tidal inundation regimes, relationshipsthat have been identified between tidal inundationparameters and S. alterniflora growth tolerance (McKeeand Patrick 1988; Collins 2002) may be used to estimate <strong>the</strong>potential for hybrid Spartina spread in San Francisco Bay.Here we present a preliminary GIS-based analysis <strong>of</strong> <strong>the</strong>potential effects <strong>of</strong> hybrid Spartina on shorebird habitat inSouth San Francisco Bay (<strong>the</strong> South Bay), using grid-basedspatial models <strong>of</strong> (a) shorebird habitat value and (b) potentialSpartina spread.- 175 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaMETHODSMany studies have demonstrated that shorebird use <strong>of</strong>mudflat habitats is spatially and temporally variable, and thatthis variation is closely tied to cycles <strong>of</strong> tidal inundation, aswell as prey availability across <strong>the</strong> intertidal zone (Burger etal. 1977; Goss-Custard et al. 1977; Puttick 1977; Page et al.1979; Colwell and Landrum 1993; Yates et al. 1993). While wedid not have data on prey availability in San Francisco Bay,our quantification <strong>of</strong> shorebird habitat value did incorporatevariation within mudflats based on tidal inundation cycles, aswell as variation among mudflats based on shorebird use datafrom PRBO’s Pacific Flyway surveys (1988-1993, Page et al.1999). For <strong>the</strong> purpose <strong>of</strong> this exercise, we assumed that SouthBay mudflats were at carrying capacity (i.e., <strong>the</strong> maximumnumber <strong>of</strong> birds that can be supported by a finite food supply)at <strong>the</strong> time <strong>of</strong> <strong>the</strong> surveys. By extension, we assumed thatloss <strong>of</strong> habitat in one area would not be compensated for byincreased use <strong>of</strong> o<strong>the</strong>r areas.The spread potential <strong>of</strong> S. alterniflora and associatedhybrids was based on percentiles <strong>of</strong> cumulative monthly tidalinundation across <strong>the</strong> mudflats. The cumulative monthlyduration <strong>of</strong> inundation at a particular site is a function <strong>of</strong>mudflat elevation and tidal range, with a greater tidal rangeresulting in a longer duration <strong>of</strong> inundation. According toCollins’ (2002) analyses <strong>of</strong> non-native Spartina locationsin San Francisco Bay, <strong>the</strong> lower limit <strong>of</strong> Spartina growthappears to correspond with cumulative monthly inundation,and existing non-native Spartina locations suggest that <strong>the</strong>maximum cumulative duration <strong>of</strong> inundation tolerated during<strong>the</strong> month <strong>of</strong> June is approximately 70%, regardless <strong>of</strong> meantidal range. 1 This means that <strong>the</strong> smaller <strong>the</strong> tidal range, <strong>the</strong>lower <strong>the</strong> elevation at which non-native Spartina would bepredicted to grow. Due to uncertainty about <strong>the</strong> behavior <strong>of</strong>S. alterniflora hybrids, and because <strong>the</strong>se plants are known tochange <strong>the</strong>ir environment over time (Ranwell 1964; Daehlerand Strong 1996), accreting sediment at rates <strong>of</strong> 1-2 cm/yearin Willapa Bay (Sayce 1988) and up to 4 cm/year in Australia(Bascand 1970), we evaluated a range <strong>of</strong> inundation tolerancesbetween 60% and 80%. We assumed that mudflat areascovered by S. alterniflora and associated hybrids would beeffectively lost to shorebirds.Our GIS-based analysis was restricted to mudflatsmapped by <strong>the</strong> San Francisco Estuary Institute’s EcoAtlas(v. 1.50b, SFEI 1998) south <strong>of</strong> <strong>the</strong> San Francisco Bay Bridge.Using EcoAtlas map layers, PRBO shorebird surveys (Stenzelet al. 2002), and tide level data from <strong>the</strong> National Oceanicand Atmospheric Administration’s (NOAA) tide stations, wedeveloped a set <strong>of</strong> grid-based data layers (ArcInfo format)that were combined to generate predictions about <strong>the</strong> potentialloss <strong>of</strong> mudflat habitat and shorebird numbers.To generate <strong>the</strong> GIS grid layers for this analysis, wecompleted <strong>the</strong> following steps using <strong>the</strong> ArcInfo 8.3 GRIDmodule (ESRI 2002).1 Initial estimates <strong>of</strong> 40% presented in Collins (2002) have since been revised.Fig. 1. Tide station locations, station IDs, and allocation <strong>of</strong> tide stationswith continuous water level data to mudflat areas. Tide station informationwas obtained from NOAA/NOS (http://tidesandcurrents.noaa.gov/).Elevation/BathymetryWhen this analysis was performed, bathymetry datalayers <strong>of</strong> a fine enough resolution to capture mudflat topographywere not available for <strong>the</strong> South Bay. Thus, to createa spread model for S. alterniflora and associated hybrids,we elected to estimate mudflat elevation at a 3x3 m 2 (3-m)pixel resolution, creating a digital elevation model (DEM)based on mapped mudflat boundaries, tide level data, and anassumed linear mudflat slope.First, mean tide level (MTL) and mean lower low water(MLLW) contours were estimated from EcoAtlas (SFEI1998) and were defined based on <strong>the</strong> boundaries betweenmudflat and tidal marsh and between open water and mudflat,respectively. Actual elevations along <strong>the</strong> MTL contour werenot assumed to be constant, but were assigned based on MTLelevation at <strong>the</strong> closest tide station. MTL elevations wereobtained from NOAA’s National Oceanic Service (NOS)published benchmark sheets for seven South Bay locationsthat have been referenced to <strong>the</strong> new National Tidal DatumEpoch (NTDE; 1983-2001) (Fig. 1). For each mudflat area weassumed that local MTL was <strong>the</strong> same as that <strong>of</strong> <strong>the</strong> nearestNTDE-referenced tide station.Next we used MTL and MLLW contours to determine<strong>the</strong> width <strong>of</strong> <strong>the</strong> mudflat for each 3-m pixel. We calculated <strong>the</strong>distance from each pixel to <strong>the</strong> MTL line and to <strong>the</strong> MLLWline, to obtain two separate distance grids, which were <strong>the</strong>nadded to obtain a single grid representing mudflat width. For- 176 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartinaeach mudflat section we estimated <strong>the</strong> slope (assumed linear)by dividing <strong>the</strong> total change in elevation across <strong>the</strong> mudflat(MTL) by mudflat width (slope = rise/run). Then we createdtwo 3-m elevation grids based on <strong>the</strong> following equations,where each 3- m pixel value was equal to <strong>the</strong> elevation atthat point:(1) elevation 1 = slope * distance to MLLW; and(2) elevation 2 = MTL – (slope * distance to MTL).We averaged <strong>the</strong>se two grids to obtain <strong>the</strong> final 3-mmudflat elevation grid (DEM). Values greater than localMTL were redefined to be equal to local MTL.Tidal Inundation and Shorebird Habitat ValuePublished verified continuous (six-minute interval)water level data was available from NOS for only three <strong>of</strong> <strong>the</strong>seven tide station locations: Alameda (station ID 9414750,year 2001), Dumbarton Bridge (station ID 9414509, year1996), and Redwood City (station ID 9414523, year 2002).Assuming that <strong>the</strong> tidal inundation regime for each mudflatarea was most similar to <strong>the</strong> nearest tide station with availablecontinous water level data, we created a tide station gridlayer by allocating each 3-m pixel to a tide station (Fig. 1).Because our shorebird data collection efforts were centeredaround April and September, we generated monthly inundationcurves for <strong>the</strong>se months using cumulative water level datafrom each tide station using Stata 8.0 (2003) (Fig. 2).Water level values were in meters above MLLW. Topredict <strong>the</strong> monthly tidal inundation percent <strong>of</strong> each 3-mmudflat pixel, we performed separate polynomial regressionanalyses (Stata 8.0, 2003) for each tide station and eachmonth (April and September) using <strong>the</strong> appropriate NOS6 minute water level data. Resulting regression equations(Table 1) were used to calculate grids representing April andSeptember inundation. Separate equations were developedfor each tide station so grids for each <strong>of</strong> <strong>the</strong> three tide stationTable 1. Coefficients for regression equations used to predict inundationfrom elevation. Equations were based on inundation curves obtained from6-minute water level data from NOS tide stations and took <strong>the</strong> form y =ax + bx 2 + cx 3 + d, where y = percent <strong>of</strong> time inundated and x = elevationabove MLLW.Tide station Month a b c d R2Alameda April -27.9 -24.9 7.85 91.3 0.9974Alameda Sept -21.6 -28.4 7.92 96.4 0.9997DumbartonBridgeApril -19.7 -13.7 2.93 94.6 0.9990DumbartonBridgeSept -20.4 -12.1 2.36 98.2 0.9984RedwoodCityApril -20.4 -16.8 4.01 93.9 0.9974RedwoodCitySept -15.7 -18.2 3.79 97.9 0.9994areas could be calculated separately and <strong>the</strong>n merged to createone seamless 3-m inundation grid for each season. We <strong>the</strong>ngenerated grids representing April and September mudflatexposure (100 - monthly inundation percent), as indices <strong>of</strong>overall shorebird habitat value (Fig. 3).Potential Spartina SpreadCumulative water level data (m above MLLW) from eachtide station were used to generate monthly June inundationcurves using Stata 8.0 (2003). For each tide station, June dura-Fig. 2. Cumulative duration <strong>of</strong> tidal inundation curves for April based onwater level data from Alameda, Redwood City, and Dumbarton Bridgetide stations. Elevations corresponding to 60% inundation were 0.762 m,1.018 m, and 1.083 m, respectively.Fig. 3. April mudflat habitat availability, based on monthly mudflat exposureexpressed as 100% - monthly mudflat inundation. Color shadings representmudflat habitat quantiles, where 20% <strong>of</strong> <strong>the</strong> total area is contained in eachshading category and darker colors have higher value.- 177 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaFig. 4. Cumulative duration <strong>of</strong> inundation curves for June based on waterlevel data from Alameda, Redwood City, and Dumbarton Bridge tide stations.Elevations corresponding to 60% inundation were 0.864 m, 1.101m, and 1.113 m, respectively. Elevations corresponding to 70% inundationwere 0.674 m, 0.882 m, and 0.874 m, respectively. Elevations correspondingto 80% inundation were 0.413 m, 0.575 m, and 0.646 m, respectively.Table 2. Projected percent <strong>of</strong> mudflat area and shorebird habitat value lost,based on a range <strong>of</strong> Spartina spread scenarios (1-3). All predictions assumethat mudflats are at carrying capacity, and that mudflat habitat value isinversely proportional to tidal inundation.Fall1 (60% inundationtolerance)2 (70% inundationtolerance)3 (80% inundationtolerance)Spring1 (60% inundationtolerance)2 (70% inundationtolerance)3 (80% inundationtolerance)Percent <strong>of</strong> mudflatinvaded bySpartinaPercent<strong>of</strong> habitatvalue lost toshorebirds14% 29%33% 57%54% 80%14% 27%33% 54%54% 76%Fig. 5. Shorebird biomass density (kg/ha) based on overall means fromcomprehensive spring shorebird survey data (1988-1993). Species-specificbiomass estimates were taken from <strong>the</strong> shorebird literature.tion <strong>of</strong> inundation curves were used to identify <strong>the</strong> elevationsat which 60%, 70%, and 80% cumulative monthly inundationwere achieved (Fig. 4). Each elevation was considered apotential threshold below which S. alterniflora and associatedhybrids would not grow (i.e., <strong>the</strong>ir inundation tolerance). We<strong>the</strong>n calculated potential Spartina spread for each inundationtolerance, selecting all mudflat pixels with modeled elevationsabove that particular elevation threshold.Shorebird NumbersWe used PRBO’s shorebird survey data to estimate falland spring shorebird numbers and overall biomass in kilograms(kg) for each <strong>of</strong> six South Bay mudflat census tracts(Stenzel et al. 2002). This resulted in fall and spring shorebirddensity grids, with densities uniformly distributed over eachcensus tract (Fig. 5).Potential Effects <strong>of</strong> Spartina Spread on Shorebird NumbersUsing <strong>the</strong> GIS grid layers described above, we estimated<strong>the</strong> potential effects <strong>of</strong> non-native Spartina spread on shorebirdnumbers. For each <strong>of</strong> <strong>the</strong> six South Bay census tractsand for each Spartina spread scenario (60%, 70%, and80% inundation tolerance), we first calculated <strong>the</strong> percent<strong>of</strong> mudflat area that would be lost. Then we calculated <strong>the</strong>proportion <strong>of</strong> habitat value that would be lost to shorebirds ifmudflat value were inversely related to <strong>the</strong> percent <strong>of</strong> mudflatinundation time.For each <strong>of</strong> <strong>the</strong> three Spartina spread scenarios, we calculateda predicted loss <strong>of</strong> spring and fall shorebird biomassby multiplying <strong>the</strong> proportion <strong>of</strong> mudflat value lost in eachcensus tract with <strong>the</strong> total estimated shorebird biomass supportedby that census tract. The same was done for individualspecies’ numbers.RESULTSOur Spartina spread model predicted that between14% and 54% <strong>of</strong> <strong>the</strong> total South Bay mudflat area could beencroached upon by S. alterniflora and associated hybrids(Fig. 6, Table 2). The areas <strong>of</strong> greater Spartina spread poten-- 178 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaFig. 6. Predicted extent <strong>of</strong> Spartina spread based on monthly inundationtolerances ranging from 60% to 80% (from Collins 2002). Sharp breaks inpredictions are due to breaks in nearest tide station locations.*tial were <strong>the</strong> upper mudflats, due to <strong>the</strong>ir higher elevation andlower tidal inundation frequency (Fig. 6).Weighting <strong>the</strong> areas <strong>of</strong> potential Spartina spread byrelative shorebird value, <strong>the</strong> predicted loss to shorebirdsranged from 27% to 80% (Table 2). The upper mudflats had<strong>the</strong> highest relative value due to <strong>the</strong>ir lower tidal inundationfrequency (Figs. 7, 8). The east and south shore mudflats had<strong>the</strong> highest value during <strong>the</strong> fall (Fig. 7), and <strong>the</strong> east shoremudflats had <strong>the</strong> highest value during <strong>the</strong> spring (Fig. 8),based on shorebird survey numbers.Multiplying <strong>the</strong> predicted mudflat habitat loss in eachcensus tract by estimated current shorebird numbers yieldedloss projections <strong>of</strong> 48,615 to 94,175 birds in <strong>the</strong> fall (Table 3)and 104,793 to 212,813 birds in <strong>the</strong> spring (Table 4).Species that concentrated in <strong>the</strong> South Bay were BlackbelliedPlover, Willet, Marbled Godwit, small sandpipers, anddowitchers (Stenzel et al. 2002). Overall shorebird numberswere higher in spring than in fall, driven primarily by <strong>the</strong>large number <strong>of</strong> Western Sandpipers that use San FranciscoBay as a staging area during spring migration. BecauseWestern Sandpipers are small-bodied shorebirds, <strong>the</strong> differencebetween fall and spring biomass was much smaller than<strong>the</strong> difference between fall and spring numbers.Because our models did not incorporate any species-specificdifferences in shorebird foraging habits, <strong>the</strong> predictedproportional change in numbers was <strong>the</strong> same for all groups.Due to migration timing and overall numbers detected on* Current extent <strong>of</strong> invasive Spartina is based on pre-control mapping effortsFig. 7. Predicted extent <strong>of</strong> Spartina spread, based on a 70% inundationtolerance, and overlap with tidal mudflats, classified according to <strong>the</strong>irpotential fall season value for shorebirds. Shorebird habitat value wasbased on: (a) PRBO Pacific Flyway shorebird survey data (1988-1993) and(b) length <strong>of</strong> mudflat inundation during September.*Fig. 8. Predicted extent <strong>of</strong> Spartina spread, based on a 70% inundationtolerance, and overlap with tidal mudflats, classified according to <strong>the</strong>irpotential spring season value for shorebirds. Shorebird habitat value wasbased on: (a) PRBO Pacific Flyway shorebird survey data (1988-1993) and(b) length <strong>of</strong> mudflat inundation during April.*- 179 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaTable 3. Predicted loss <strong>of</strong> fall shorebird numbers by species, based on arange <strong>of</strong> Spartina spread scenarios. All predictions assume that mudflatsare at carrying capacity, and that mudflat habitat value is proportional totidal inundation.Table 4. Predicted loss <strong>of</strong> spring shorebird numbers by species, based on arange <strong>of</strong> Spartina spread scenarios. All predictions assume that mudflatsare at carrying capacity, and that mudflat habitat value is proportional totidal inundation.CurrentScenario 1:60%inundationtoleranceScenario 2:70%inundationtoleranceScenario 3:80%inundationtoleranceCurrentScenario 1:60%inundationtoleranceScenario 2:70%inundationtoleranceScenario 3:80%inundationtoleranceMudflatHectares6,062 -904 -1,988 -3,269MudflatHectares6,062 -904 -1,988 -3,269Bird Numbers:AmericanAvocetBlack-belliedPlover5,023 -1,701 -3,034 -4,1018,138 -2,756 -4,916 -6,645Dowitcher 13,377 -4,530 -8,081 -10,923Long-billedCurlewMarbledGodwit371 -126 -224 -30314,251 -4,826 -8,609 -11,636Red Knot 1,678 -568 -1,014 -1,370SemipalmatedPlover1,501 -508 -907 -1,225Willet 15,612 -5,286 -9,431 -12,747Westernand LeastSandpiper,160,374 -54,305 -96,880 -130,948DunlinTotal 220,325 -48,615 -70,055 -94,175Bird Numbers:AmericanAvocetBlack-belliedPlover844 -263 -476 -6564,595 -1,432 -2,590 -3,570Dowitchers 33,008 -10,289 -18,608 -25,644Long-billedCurlewMarbledGodwit218 -68 -123 -16913,437 -4,188 -7,575 -10,439Red Knot 503 -157 -284 -391SemipalmatedPlover725 -226 -409 -563Willet 2,112 -658 -1,191 -1,641Westernand LeastSandpiper,Dunlin450,817 -140,528 -254,137 -350,241Total 506,259 -104,793 -156,097 -212,813shorebird surveys, Spartina spread would have <strong>the</strong> biggestnumerical impact on small shorebirds, dowitchers, andMarbled Godwits (Tables 3, 4). Willets and Black-belliedPlovers would be most affected during <strong>the</strong> fall, when <strong>the</strong>irnumbers are highest. In terms <strong>of</strong> total bird biomass (seeStenzel et al. 2002), <strong>the</strong> largest predicted losses were in <strong>the</strong>spring, due to higher overall biomass densities (Table 5).DISCUSSIONThe results presented herein are based on several basicassumptions, all <strong>of</strong> which should be examined in fur<strong>the</strong>rdetail in order to restrict <strong>the</strong> wide range <strong>of</strong> predicted shorebirdlosses. Our most fundamental assumption was that <strong>the</strong>mudflat habitats were at carrying capacity during <strong>the</strong> falland spring survey periods. However, it is possible that onlypreferred areas are functioning at carrying capacity (Goss-Custard 1979). If individuals could switch to lower-qualitymudflat areas without significantly affecting <strong>the</strong>ir overallfitness, <strong>the</strong>n <strong>the</strong> potential loss <strong>of</strong> birds may have been overestimated(Goss-Custard 2003). Anecdotal evidence suggeststhat San Francisco Bay mudflats may reach carrying capacityduring <strong>the</strong> winter, when storm-related flooding may promptsome species to move inland to <strong>the</strong> Central Valley (Warnocket al. 1995; Takekawa et al. 2002) but we do not know if mudflatsand neighboring tidal and salt pond habitats are at carryingcapacity during migration. Currently, we lack <strong>the</strong> data onmudflat food resources, shorebird energetics, and individualforaging behavior (especially prey preference) that would benecessary to obtain an estimate <strong>of</strong> carrying capacity.A potential bias in <strong>the</strong> o<strong>the</strong>r direction, however, was thatour models examined only <strong>the</strong> lower limits <strong>of</strong> Spartina spreadand <strong>the</strong> subsequent impacts on shorebird habitat value. Inreality, upward Spartina spread may pose an equally seriousthreat to shorebirds as mudflats along tidal channels andopen areas within <strong>the</strong> marsh plain are colonized by invasiveSpartina and become unavailable to foraging shorebirds.Table 5. Predicted loss <strong>of</strong> fall and spring shorebird biomass, based on a range<strong>of</strong> Spartina spread scenarios (1-3). All predictions assume that muflats areat carrying capacity, and that mudflat habitat value is proportional to tidalinundation.Fallbiomass(kg)Springbiomass(kg)CurrentScenario 1:60%inundationtoleranceScenario 2:70%inundationtoleranceScenario 3:80%inundationtolerance21,416 -6,245 -12,263 -17,06725,289 -6,884 -13,560 -19,202- 180 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaAno<strong>the</strong>r assumption <strong>of</strong> our habitat value models wasthat exposed mudflat areas are used evenly by shorebirds,and thus in direct proportion to <strong>the</strong>ir temporal availability.However, we know that individuals <strong>of</strong> many species tendto forage along <strong>the</strong> rising or receding tide line (Colwell andLandrum 1993; Durell et al. 1997), as invertebrates are moreabundant and more accessible in wet substrates and tendto burrow deeper as <strong>the</strong> tide recedes and <strong>the</strong> mud dries out(Goss-Custard 1984; White 1995). For shorebirds that onlyforaged along <strong>the</strong> tide line, any given full tidal range mudflatshould support a constant number <strong>of</strong> birds as long as someminimum mudflat area was exposed, and shorebird densities(with respect to exposed mudflat) should increase as <strong>the</strong> tiderises. In reality, it is likely that some intermediate conditionexists and mudflat use patterns vary by species. In SanFrancisco Bay we have observed that Semipalmated Plover,Least Sandpiper, and Black-bellied Plover tend to foragehigher along <strong>the</strong> tidal gradient, whereas Dunlin, dowitchers,Marbled Godwit, and o<strong>the</strong>r species forage closer to <strong>the</strong> tideline (PRBO unpubl. data; USGS unpubl. data).Shorebird densities are also known to vary accordingto <strong>the</strong> uneven distribution <strong>of</strong> sediments, prey densities, andprey availability across <strong>the</strong> intertidal zone (Burger et al.1977; Goss-Custard et al. 1977; Puttick 1977; Page et al.1979; Quammen 1982, Colwell and Landrum 1993; Yates etal. 1993). This highlights <strong>the</strong> need to study <strong>the</strong> spatial distribution<strong>of</strong> shorebirds and <strong>the</strong>ir invertebrate prey over SanFrancisco Bay mudflats, especially given <strong>the</strong> dynamic nature<strong>of</strong> <strong>the</strong> largely non-native invertebrate community (Nichols etal. 1986; Cohen and Carlton 1998).With respect to our estimates <strong>of</strong> mudflat elevation andtidal inundation, we used a simple approach that involvedassuming an unrealistic linear mudflat slope. While highresolutiontopographic/ bathymetric data were not availablefor <strong>the</strong> entire South Bay mudflat region at <strong>the</strong> time <strong>of</strong> thisstudy, we hope that recently-obtained Light Detection andRanging (LiDAR) data (Foxgrover and Jaffe 2005) may beused to improve our models in <strong>the</strong> future.Finally, our model predictions were based on a staticpicture <strong>of</strong> mudflat spatial extent and elevation. In reality,<strong>the</strong>re are several factors in addition to hybrid Spartina spreadthat may affect mudflat extent and quality, including sea levelrise (Donnelly and Bertness 2001; Galbraith et al. 2002),tidal marsh restoration, and natural geomorphic processes(Foxgrover et al. 2004). Thus it would be useful to apply ourmodels to future mudflat predictions developed by coastalgeomorphologists.CONCLUSIONThe above-described models <strong>of</strong> Spartina inundationtolerance and shorebird habitat value allowed us to generatepreliminary predictions about <strong>the</strong> potential impact <strong>of</strong>future Spartina spread on South Bay shorebird populations.Because mudflat habitat availability and Spartina spreadpotential are both determined by tidal inundation frequency,<strong>the</strong> areas <strong>of</strong> greatest value to shorebirds coincide with <strong>the</strong>areas <strong>of</strong> greatest Spartina invasion potential. In addition,<strong>the</strong> eastern shore mudflats, which had <strong>the</strong> highest recordedshorebird densities, are adjacent to one <strong>of</strong> <strong>the</strong> sites <strong>of</strong> initialSpartina invasion at Coyote Hills Slough, increasing <strong>the</strong>irsusceptibility to Spartina encroachment.Due to various sources <strong>of</strong> uncertainty, our preliminaryanalysis resulted in a wide range <strong>of</strong> predictions: a 14% to 54%loss in mudflat area, and a 27% to 80% loss in habitat valueacross <strong>the</strong> three Spartina spread scenarios examined. At <strong>the</strong>low end, this represents a significant loss <strong>of</strong> habitat for shorebirds,which may or may not result in population declines.At <strong>the</strong> high end, our predictions suggest an extreme changein habitat availability that would almost certainly reduce <strong>the</strong>number <strong>of</strong> foraging shorebirds in <strong>the</strong> South Bay.At <strong>the</strong> same time, o<strong>the</strong>r major changes are occurring in<strong>the</strong> South Bay, including <strong>the</strong> restoration <strong>of</strong> up to 5,000 hectares<strong>of</strong> commercial salt ponds to tidal, muted, and managedmarsh. The loss <strong>of</strong> <strong>the</strong>se salt ponds, which currently serveas valuable foraging and roosting habitat, may also reduceshorebird numbers (Stralberg et al. 2006). Thus we suspectthat South Bay shorebird populations may suffer from multiplenegative impacts within a relatively short timeframe ifSpartina spread is not arrested, reducing <strong>the</strong> value <strong>of</strong> SanFrancisco Bay as a major migratory stopover and winteringsite. Additional research and modeling will be needed toassess <strong>the</strong> cumulative effects <strong>of</strong> habitat change on shorebirds,while large-scale, integrated habitat conservation planningmay help to mitigate <strong>the</strong>se negative effects.AUTHORS’ NOTE:Due to extensive Spartina control efforts since <strong>the</strong>writing <strong>of</strong> this paper, Spartina spread now poses much less<strong>of</strong> an immediate threat to shorebirds in San Francisco Bay.ACKNOWLEDGMENTSThis project was requested by <strong>the</strong> Invasive SpartinaProject (ISP), and funded by <strong>the</strong> California CoastalConservancy, <strong>the</strong> State Resources Agency, and <strong>the</strong> CALFEDProgram, under contract #02-212. We would like to thankKaty Zaremba and Peggy Ol<strong>of</strong>son <strong>of</strong> <strong>the</strong> ISP, Debra Ayres,Don Strong, Janie Civille and Ted Grosholz <strong>of</strong> UC Davis,and Bruce Jaffe and John Takekawa <strong>of</strong> USGS for <strong>the</strong>ir valuableinput and support. We are also grateful to Josh Collinsand Stuart Siegel for advice on methods, to Hildie Spautz forassistance with <strong>the</strong> literature review, and to an anonymousreviewer for feedback on an earlier draft <strong>of</strong> this manuscript.Finally, we are very grateful for <strong>the</strong> assistance <strong>of</strong> hundreds <strong>of</strong>volunteers who made <strong>the</strong> shorebird surveys possible. This isPRBO contribution number 1221.REFERENCESAnttila, C.K., C.C. Daehler, N.E. Rank, and D.R. Strong. 1998.Greater male fitness <strong>of</strong> a rare invader (Spartina alterniflora,Poaceae) threatens a common native (Spartina foliosa) withhybridization. Am. J. Bot. 85:1597–1601.Ayres, D.R., D. Garcia-Rossi, H.G. Davis, and D.R. Strong. 1999.Extent and degree <strong>of</strong> hybridization between exotic (Spartina- 181 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaalterniflora) and native (S. foliosa) cordgrass (Poaceae) inCalifornia, USA determined by random amplified polymorphicDNA (RAPDs). Mol. Ecol. 8:1179-1186.Ayres, D.R., Zaremba, K., Sloop, C.M., Strong, D.R., 2008. Sexualreproduction <strong>of</strong> cordgrass hybrids (Spartina foliosa x al-terniflora)invading tidal marshes in San Francisco Bay. Divers.Distrib. 14:187-195.Bascand, L.D. 1970. The roles <strong>of</strong> Spartina species in New Zealand.Proc. N.Z. Ecol. Soc. 17:33-40.Buchanan, J.B. 2003. Spartina invasion <strong>of</strong> Pacific coast estuariesin <strong>the</strong> United States: implications for shorebird conservation.Wader Study Group Bulletin 100:47-49.Burger, J., M.A. Howe, D.C. Hahn, and J. Chase. 1977. Effects<strong>of</strong> tide cycles on habitat selection and habitat partitioning bymigrating shorebirds. Auk 9:743-758.Callaway, J.C., and M.N. Josselyn. 1992. The introduction andspread <strong>of</strong> smooth cordgrass (Spartina alterniflora) in South SanFrancisco Bay. Estuaries 15:218-226.Civille, J.C., K. Sayce, S.D. Smith, and D.R. Strong. 2005.Reconstructing a century <strong>of</strong> Spartina alterniflora invasion withhistorical records and contemporary remote sensing. Ecoscience12:330-338.Cohen, A.N. and J.T. Carlton. 1998. Accelerating invasion rate in ahighly invaded estuary. Science 279:555-558.Collins, J.N. 2002. Invasion <strong>of</strong> San Francisco Bay by smooth cordgrass,Spartina alterniflora: a forecast <strong>of</strong> geomorphic effects on<strong>the</strong> intertidal zone. Unpublished report <strong>of</strong> San Francisco EstuaryInstitute, Oakland, CA.Colwell, M.A. and S.L. Landrum. 1993. Nonrandom shorebirddistribution and fine-scale variation in prey abundance. Condor95:94-103.Daehler, C.C. and D.R. Strong. 1996. Status, prediction and prevention<strong>of</strong> introduced cordgrass Spartina spp. invasions in Pacificestuaries, USA. Biological Conservation 78:51-58.Daehler, C.C. and D.R. Strong. 1997. Hybridization betweenintroduced Smooth Cordgrass (Spartina alterniflora) and nativeCalifornia cordgrass (S. foliosa) in San Francisco Bay, California,USA. American Journal <strong>of</strong> Botany 84:607-611.Donnelly, J.P. and M.D. Bertness. 2001. Rapid shoreward encroachment<strong>of</strong> salt marsh cordgrass in response to accelerated sea-levelrise. PNAS 98:14218-14223.Durell, S.E., A. LeDit, J.D. Goss-Custard, and R.T. Clarke. 1997.Differential response <strong>of</strong> migratory subpopulations to winterhabitat loss. J. Appl. Ecol. 34:1155-1164.ESRI. 2002. ArcInfo 8.3. Environmental Systems Research Institute,Redlands, CA.Evans, P.R. 1986. Use <strong>of</strong> <strong>the</strong> herbicide ‘Dalapon’ for control <strong>of</strong>Spartina encroaching on intertidal mudflats: beneficial effectson shorebirds. Colonial Waterbirds 9:171-175.Foxgrover, A.C., S.A. Higgins, M.K. Ingraca, B.E. Jaffe, and R.E.Smith. 2004. Deposition, erosion, and bathymetric change inSouth San Francisco Bay: 1858-1983: U.S. Geological SurveyOpen-File Report 2004-1192.Foxgrover, A.C., B.E. Jaffe. 2005. South San Francisco Bay 2004topographic lidar survey: Data overview and preliminaryquality assess-ment, U.S. Geological Survey Open-File ReportOFR-2005-1284, version 1.0. http://pubs.usgs.gov/<strong>of</strong>/2005/1284/.Galbraith, H., R. Jones, R. Park, J. Clough, S. Herrod-Julius, B.Harrington, and G. Page. 2002. Global climate change and sealevel rise: potential losses <strong>of</strong> intertidal habitat for shorebirds.Waterbirds 25:173-183.Goss-Custard, J.D. 1979. Effect <strong>of</strong> habitat loss on <strong>the</strong> numbers <strong>of</strong>overwintering shorebirds. Studies in Avian Biology 2:167-177.Goss-Custard, J. D. 1984. Intake rates and food supply in migratingand wintering shorebirds. In: Burger, J. and B.L. Olla, eds.Shorebirds: Migration and Foraging Behaviour, pp. 233-270.Plenum Press, London and New York.Goss-Custard, J.D. 2003. Fitness, demographic rates and managing<strong>the</strong> coast for wader populations. Wader Study Group Bulletin100:183-191.Goss-Custard, J.D., R.E. Jones, and P.E. Newbery. 1977. Theecology <strong>of</strong> <strong>the</strong> Wash. I. Distribution and diet <strong>of</strong> wading birds(Charadrii). J. Appl. Ecol. 14:681-700.Goss-Custard, J. D. and M. E. Moser. 1988. Rates <strong>of</strong> change in <strong>the</strong>numbers <strong>of</strong> Dunlin, Calidris alpina, wintering in British estuariesin relation to <strong>the</strong> spread <strong>of</strong> Spartina Anglica. J. Appl. Ecol.25:95-109.Jaques, D. 2002. Shorebird status and effects <strong>of</strong> Spartina alternifloraat Willapa NWR Unpublished report. US Fish and WildlifeService, Willapa Bay National Wildlife Refuge.Josselyn, M. 1983. The ecology <strong>of</strong> San Francisco Bay tidal marshes:a community pr<strong>of</strong>ile. U.S. Fish and Wildlife Service, Div. <strong>of</strong>Biological Services, FWS/OBS-83/23. Washington D.C.Josselyn, M., J. Zedler, and T. Griswold. 1990. Wetland mitigationalong <strong>the</strong> Pacific Coast <strong>of</strong> <strong>the</strong> United States. In: Kentula, J,A, andM.E. Kusler, eds. Wetland Creation and Restoration: <strong>the</strong> Status<strong>of</strong> <strong>the</strong> Science. Island Press, Washington D.C.Nichols, F.H., J.E. Cloern, S.N. Luoma, and D.H. Peterson. 1986.The modification <strong>of</strong> an estuary. Science 231:567-573.Page, G.W., L.E. Stenzel, and J.E. Kjelmyr. 1999. Overview <strong>of</strong>shorebird abundance and distribution in wetlands <strong>of</strong> <strong>the</strong> PacificCoast <strong>of</strong> <strong>the</strong> contiguous United States. Condor 101:461-471.Page, G.W., L.E. Stenzel, and C.M. Wolfe. 1979. Aspects <strong>of</strong> <strong>the</strong>occurrence <strong>of</strong> shorebirds on a central California estuary. Studiesin Avian Biology 2:15-32.Puttick, G.M. 1977. Spatial and temporal variations in intertidalanimal distribution at Langebaan Lagoon, South Africa. RoyalSociety <strong>of</strong> South Africa. Trans. 42:403-440.Quammen, M.L. 1982. Influence <strong>of</strong> subtle substrate differenceson feeding by shorebirds on intertidal mudflats. Mar. Biol.71:339-343.Ranwell, D.S. 1964. Spartina salt marshes in sou<strong>the</strong>rn England:II. rate and seasonal pattern <strong>of</strong> sediment accretion. J. Ecol.52:79-94.Sayce, K. 1988. Introduced cordgrass, Spartina alterniflora Loiselin saltmarshes and tidelands <strong>of</strong> Willapa Bay, Washington. Reportto US Fish and Wildlife Service, Willapa NWR.SFEI. 1998. EcoAtlas beta release, version 1.5b4. San FranciscoEstuary Institute, Oakland, CA.Stata. 2003. Intercooled Stata 8.0 for Windows. Stata Corporation,College Station, TX.Stenzel, L.E., C.M. Hickey, J.E. Kjelmyr, and G.W. Page. 2002.Abundance and distribution <strong>of</strong> shorebirds in <strong>the</strong> San FranciscoBay area. Western Birds 33:69-98.Stralberg, D., M. Herzog, N. Warnock, N. Nur, and S. Valdez. 2006.Habitat-based modeling <strong>of</strong> wetland bird communities: an evaluation<strong>of</strong> potential restoration alternatives for South San FranciscoBay. PRBO unpubl. report to California Coastal Conservancy.http://www.prbo.org/wetlands/hcm.Takekawa, J.Y., N. Warnock, G.M. Martinelli, A.K. Miles, andD. C. Tsao. 2002. Waterbird use <strong>of</strong> bayland wetlands in <strong>the</strong> San- 182 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaFrancisco Bay Estuary: movements <strong>of</strong> Long-billed Dowitchersduring winter. Waterbirds 25 (Special Publication 2):93-105.Warnock, N., and M.A. Bishop. 1998. Spring stopover ecology <strong>of</strong>migrant Western Sandpipers. Condor 100:456-467.Warnock, N., G.W. Page, and L. E. Stenzel. 1995. Non-migratorymovements <strong>of</strong> Dunlin on <strong>the</strong>ir California wintering grounds.Wilson Bulletin 107:131-139.White, B. C. 1995. The shorebird foraging response to <strong>the</strong> eradication<strong>of</strong> <strong>the</strong> introduced cordgrass, Spartina alterniflora. M.A.Thesis. San Francisco State University, San Francisco, CA.Yates, M.G., J.D. Goss-Custard, S. McGrorty, K.H. Lakhani, S.Durell, R.T. Clarke, W.E. Rispin, I. Moy, T. Yates, R.A. Plant,and J. Frost. 1993. Sediment characteristics, invertebrate densitiesand shorebird densities on <strong>the</strong> inner banks <strong>of</strong> <strong>the</strong> Wash. J.Appl. Ecol. 30:599-614.Zaremba, K., M. McGowan and D.R. Ayres. Spread <strong>of</strong> invasiveSpartina in <strong>the</strong> San Francisco estuary. In: Ayres, D.R., D.W.Kerr, S.D. Ericson and P.R. Ol<strong>of</strong>son, eds. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong><strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartina, 2004 Nov8-10, San Francisco, CA, USA. San Francisco Estuary InvasiveSpartina Project <strong>of</strong> <strong>the</strong> California State Coastal Conservancy:Oakland, CA. (this volume).- 183 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaNON-NATIVE CORDGRASS AND THE CALIFORNIA CLAPPER RAIL:BIOGEOGRAPHICALOVERLAP BETWEEN AN INVASIVE PLANT AND AN ENDANGERED BIRDJ. EVENS 1 ,K.ZAREMBA 2 , AND J. ALBERTSON 31Avocet Research Associates, P.O. Box 839, Point Reyes Station, CA 94956; email: jevens@svn.net2San Francisco Estuary Invasive Spartina Project, 2612-A 8 th St., Berkeley, CA 947102Current address: 971 Village Dr. Bowen Island, BC, V0N 1G0 Canada; katyzaremba@yahoo.ca3 Don Edwards San Francisco Bay National Wildlife Refuge, 9500 Thornton Ave.,Newark, CA 94560-3300The federally endangered California clapper rail (Rallus longirostris obsoletus) is a tidal marshdependentbird whose distribution is restricted entirely to <strong>the</strong> San Francisco Estuary. The nativecordgrass, Spartina foliosa, has long been recognized as a critical component <strong>of</strong> clapper rail habitatwithin <strong>the</strong> estuary. The distribution and abundance <strong>of</strong> <strong>the</strong> clapper rail has been monitoredintermittently since <strong>the</strong> mid-1970s and that effort has increased since <strong>the</strong> early 1990s. Theconcurrent invasion <strong>of</strong> <strong>the</strong> bay’s tidal marshes by non-native Spartina alterniflora and its hybridswith S. foliosa has impacted clapper rail abundance and distribution in some areas <strong>of</strong> <strong>the</strong> estuary. Inthis paper we discuss: (1) habitat affinities and density estimates <strong>of</strong> rail in invaded and non-invadedmarshes; (2) <strong>the</strong> distribution <strong>of</strong> non-native Spartina relative to that <strong>of</strong> <strong>the</strong> clapper rails over <strong>the</strong> last15 years; and, (3) potential impacts <strong>of</strong> changing marsh ecology on rail distribution and abundance inboth <strong>the</strong> near term and <strong>the</strong> long term.Keywords: California clapper rail, Rallus longirostris obsoletus, Spartina, cordgrass, impacts,habitat characteristics, distribution, abundanceCLAPPER RAIL DISTRIBUTION, HABITAT AFFINITIES ANDABUNDANCEFormerly more widespread along California’s outercoast (e.g. Morro Bay, Elkhorn Slough, Tomales Bay), <strong>the</strong>breeding distribution <strong>of</strong> <strong>the</strong> California clapper rail is nowrestricted entirely to <strong>the</strong> San Francisco Estuary (Albertsonand Evens 2000). Within <strong>the</strong> estuary, <strong>the</strong> clapper rail ispatchily distributed through tidally influenced marshes, withpopulation centers fairly evenly dispersed among <strong>the</strong> SouthBay and <strong>the</strong> North Bay (aka “San Pablo Bay”) marshes. TheCentral Bay also hosts significant populations, but, like <strong>the</strong>habitat, <strong>the</strong>se tend to be relatively discrete and locallyclustered (e.g. Arrowhead Marsh, San Bruno Marsh, CorteMadera marshes). In Suisun Bay, distribution is very spotty,densities are apparently low, and occurrence may besporadic, especially in <strong>the</strong> nor<strong>the</strong>rn reaches <strong>of</strong> Suisun Marsh(Collins et al. 1994; Albertson and Evens 2000; Estrella2007). Indeed, no clapper rails were detected in <strong>the</strong> Suisunsystem in 2005 or 2007 (Herzog et al. 2005; Estrella 2007).The most recent population estimates suggest thatapproximately 1500 California clapper rails remain in <strong>the</strong>San Francisco Estuary with approximately one-third <strong>of</strong> <strong>the</strong>population in San Pablo Bay and two-thirds in <strong>the</strong> Centraland South bays, combined (Albertson and Evens 2000;Avocet Research Associates 2004; USFWS unpubl. data).This compares with estimates <strong>of</strong> 4200-6000 rails in <strong>the</strong> mid-1970s (Gill 1979). However <strong>the</strong>se numbers require a caveat:Population estimates <strong>of</strong> clapper rails are fraught withuncertainty, survey coverage is sporadic, and numbers mayvary widely from year-to-year.CALIFORNIA CLAPPER RAIL HABITATCHARACTERISTICS: PRE-INVASIONHabitat availability and <strong>the</strong> characteristics <strong>of</strong> tidalmarshes differ between South, Central, and North Baymarshlands. In general, marshlands in <strong>the</strong> nor<strong>the</strong>rn reachesare more extensive and less modified by human activity thanthose in <strong>the</strong> central and sou<strong>the</strong>rn portions <strong>of</strong> <strong>the</strong> estuary.Additionally, invasion <strong>of</strong> <strong>the</strong> estuary by non-nativeSpartina—confined largely to <strong>the</strong> sou<strong>the</strong>rn half <strong>of</strong> <strong>the</strong>estuary (http://www.spartina.org/maps.htm)—has amplifiedthis disparity.Synoptic surveys <strong>of</strong> North Bay marshes (north <strong>of</strong> PointsSan Pedro and San Pablo) in <strong>the</strong> early 1990s describedhabitat use in <strong>the</strong> nor<strong>the</strong>rn reaches <strong>of</strong> <strong>the</strong> estuary (Evens andCollins 1992; Collins et al. 1994). We assume that <strong>the</strong>seecological parameters most closely resemble habitatpreferences <strong>of</strong> California clapper rail’s prior to <strong>the</strong> extensivemodification <strong>of</strong> <strong>the</strong> estuary that began in <strong>the</strong> mid-1800s(Conomos 1979; Goals Project 1999). Additionally, earlierstudies throughout <strong>the</strong> estuary described distribution,abundance, and habitat affinities <strong>of</strong> obsoletus prior toinvasion by non-native Spartina (Grinnell and Miller 1944;Gill 1979; Avocet Research 1992; Evens and Collins 1992;Collins et al. 1994; Albertson 1995). Generalized habitat- 185 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaaffinities <strong>of</strong> clapper rails in non-invaded marshes aresummarized as follows:• Distribution limited to fully tidally, saline to brackishmarshlands and <strong>the</strong>ir foreshores within <strong>the</strong> estuary. Clapperrails do not occur upstream in <strong>the</strong> Sacramento-San JoaquinDelta and fresher portions <strong>of</strong> <strong>the</strong> system.• Well-developed channel and slough systems that extendthrough or into patches <strong>of</strong> tall monocot vegetation.Channels function as important areas for foraging and asmovement corridors.• Monocot vegetation is used for nesting material;nests are built at or about Mean Higher High Water(MHHW), <strong>of</strong>ten at <strong>the</strong> headward reach <strong>of</strong> a tidal channel(ARA 1992). Non-native Spartina extends upward to nearlyMHHW (Collins 2002.)• Habitat patches typically comprise some mature andsome youthful marsh. 1 The highest densities in <strong>the</strong> NorthBay marshes are generally associated with <strong>the</strong> largestcontiguous areas <strong>of</strong> youthful, saline marshland adjoining atleast moderate amounts <strong>of</strong> historical, mature marshland(Evens and Collins 1992; Collins et al. 1994).• Broad marshes on <strong>the</strong> bayshore, or near <strong>the</strong> mouths<strong>of</strong> tidal tributaries are favored; narrow, linear, strip marshessupport much lower densities, but serve as important corridorsbetween habitat patches and as foraging areas.• Densities <strong>of</strong> rails are highest where patches <strong>of</strong> habitatare at least 100 hectares (ha) in size (Collins et al. 1994).• Densities <strong>of</strong> rails tend to increase with channel density(length <strong>of</strong> channel per unit <strong>of</strong> marshland.)• Density is positively correlated (R 2 = 0.74) to arealextent <strong>of</strong> contiguous marshland.• Small parcels <strong>of</strong> marsh along <strong>the</strong> immediate margin<strong>of</strong> a tributary are more likely to support California clapperrails than small parcels that are more isolated.• Refugial vegetation, or even man-made structures(docks, duck blinds, etc.) may provide some protection forrails, particularly during high tides and flood events. Populations<strong>of</strong> mesopredators (raccoons, skunks, feral cats, etc.),and especially <strong>the</strong> non-native red fox (Vulpes vulpes),threaten rail survival; aerial predators also pose a threat.• Boardwalks, fence lines, towers, and stakes may increasepredation pressure by providing perch sites for avianpredators (Barn Owl, Great Horned Owl, Nor<strong>the</strong>rn Harrier, etc.)Numerous protocol-level population surveys have beenconducted in <strong>the</strong> nor<strong>the</strong>rn reaches <strong>of</strong> <strong>the</strong> estuary since <strong>the</strong>early 1980s (e.g. Evens and Page 1987; Collins et al. 1994;ARA 2004; Herzog et al. 2005).At San Pablo Bay sites where rails have been detected,densities typically varied between 0.1 and 2.8 birds perhectare (Collins et al. 1992; ARA 2004; Herzog et al. 2005)with highest densities at sites where youthful marsh hasrecently developed, e.g. Bahia Lagoon with 1.7 to 2.8birds/ha (ARA 2004; Herzog et al. 2005).In large maturemarshes, with well-developed channel systems, but wherenative cordgrass is limited to a narrow fringing edge, densityestimates are more typically on <strong>the</strong> order <strong>of</strong> about 0.5birds/ha but, in <strong>the</strong> largest marshes may range up as high as1.8 birds/ha (e.g Gallinas Creek mouth—ARA 2004; Herzoget al. 2005).NON-NATIVE SPARTINA DISTRIBUTION AND ABUNDANCEAND EXTENT OF OVERLAP WITH RAILSWithin <strong>the</strong> estuary, non-native cordgrass (Spartinaalterniflora and hybrids) is limited primarily to <strong>the</strong>marshlands and tidal flats south <strong>of</strong> Point San Pedro andPoint San Pablo (Hogle 2006). Therefore, <strong>the</strong> San Pablo Bayrail subpopulations exist essentially independent <strong>of</strong> invadedhabitat while <strong>the</strong> Central and South Bay rail subpopulationsoverlap extensively with actively colonizing non-nativecordgrass (Broom 2008).Protocol-level surveys over <strong>the</strong> three years 2005-2007that covered 60 sites and were limited to marshes invaded bynon-native cordgrass found mean baywide densities in <strong>the</strong>range <strong>of</strong> 0.58 rails/ha to 0.68 rails/ha (Broom 2008). Thesevalues are not highly disparate from those reported in <strong>the</strong>nor<strong>the</strong>rn reaches <strong>of</strong> <strong>the</strong> estuary (Collins et al. 1994; ARA2004; Herzog et al. 2005) Ano<strong>the</strong>r study <strong>of</strong> 45 Central andSouth bay sites conducted in 2005-2006 derived an overallmean abundance 0f 0.84 (±0.163) rails/ha and found nostatistically significant differences among categories (0 to>50%) <strong>of</strong> Spartina hybrid cover (Spautz et al. 2006).However, at several locations where Spartina clones haveaggresively invaded <strong>the</strong> foreshore, clapper rail densitiesgreatly exceeded those reported in San Pablo Bay, in <strong>the</strong>Central and South bays prior to <strong>the</strong> invasion by non-nativecordgrass, and <strong>the</strong> mean density reported in <strong>the</strong> twoaforementioned surveys. The most obvious examples:San Bruno Marsh (Colma Creek mouth): 2.4 to 3.8birds/ha (J. Evens, unpublished data., Spautz et al. 2006)Arrowhead Marsh: 3.8 to 4.2 birds/ha (Spautz et al.2006; J. Didonato, EBRPD, pers. comm.).This positive response to youthful, recently established,intertidal marsh vegetation by obsoletus is not unexpected.As North Bay studies in <strong>the</strong> early 1990s reported: “. . .highest densities <strong>of</strong> clapper rails were generally associatedwith <strong>the</strong> largest contiguous areas <strong>of</strong> youthful, salinemarshland adjoining at least moderate areas <strong>of</strong> historical,mature marshland . . .” (Collins et al. 1994). The colonization<strong>of</strong> marsh edge and foreshore by non-native cordgrass isapparently mimicking conditions in “native marshes” thatprograde as ecological conditions change (e.g. depositionalerosionpatterns). However, <strong>the</strong> Central and South Baymarshes invaded by non-native cordgrass differ frommarshlands <strong>of</strong> San Pablo Bay, in that few adjoin largemature marshlands, few support a heterogeneity <strong>of</strong> marshage classes, few exhibit a natural transitional grade from <strong>the</strong>low marsh to <strong>the</strong> high marsh plain, and vegetative cover at<strong>the</strong> marsh/upland ecotone is <strong>of</strong>ten sparse or absent. North- 186 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaBay marshes are also larger than marshes in <strong>the</strong> sou<strong>the</strong>rnreach, higher in elevation, have a lower rate <strong>of</strong> subsidencerelative to mean tidal level, and have more emergent marshvegetation (Atwater et al. 1979; Josselyn 1983; GoalsProject 1999; Baye et al. 2000).ONGOING THREATS TO THE CLAPPER RAIL POPULATIONExtensive conversion <strong>of</strong> tidal lands resulting fromhistoric and ongoing pressures <strong>of</strong> agricultural production,urbanization, and salt production has drastically reducedCalifornia clapper rail habitat in <strong>the</strong> San Francisco Estuary(Goals Project 1999; Albertson and Evens 2000). Theremnant tidal marshlands <strong>of</strong> <strong>the</strong> estuary, <strong>the</strong> largest and lastrefuge <strong>of</strong> R.l. obsoletus, occupy only about 12-15% <strong>of</strong> <strong>the</strong>irhistoric extent, yet even in such diminished capacitycomprise more than 90% <strong>of</strong> all remaining California tidalmarshlands (Dedrick 1989; Goals Project 1999). In additionto, and exacerbated by, habitat loss and fragmentation, <strong>the</strong>rail population is vulnerable to a redundancy <strong>of</strong> threats—predation by introduced red fox (Vulpes vulpes); depredation<strong>of</strong> nests and eggs by non-native rodents (Rattus spp.) andinflated populations <strong>of</strong> native mammals (e.g. striped skunk,raccoon); contamination (Lonzarich et al. 1992;Schwarzbach et al. 2006); diminution <strong>of</strong> peripheral refugialvegetation; and sea-level rise. The ecological alteration <strong>of</strong><strong>the</strong> estuary’s tidal marshes by non-native vegetation,especially Spartina alterniflora and its hybrids, and possiblyLepidium (Spautz and Nur 2004), poses yet ano<strong>the</strong>r potentialthreat to an already beleaguered species.Ecologists have found that non-native hybrid cordgrassexceeds <strong>the</strong> narrower niche <strong>of</strong> <strong>the</strong> native tidal marsh plantand predicted that as <strong>the</strong> invasion progresses, hybridSpartina will likely spread to higher and lower elevations(Cohen 2001; Callaway and Josselyn 1992; Daehler andStrong 1996; Collins 2002; Baye 2004). As cordgrass bedsexpand, <strong>the</strong>y are expected to “constrain <strong>the</strong> tidal network”(Collins 2002) and, as <strong>the</strong> culms reach mature densities, <strong>the</strong>stands are apt to become effective sediment traps, promotinginfilling (Baye 2004). Additionally, <strong>the</strong> lower elevationlimits <strong>of</strong> non-native cordgrass “correspond to <strong>the</strong> slumpblocks and lower banks <strong>of</strong> large channels” (Collins 2002),prime clapper rail foraging areas.One <strong>of</strong> <strong>the</strong> forecasts <strong>of</strong> a study on <strong>the</strong> geomorphiceffects <strong>of</strong> non-native cordgrass invasion was that it will tend“to isolate <strong>the</strong> headward reaches <strong>of</strong> first order channelsfrom <strong>the</strong>ir networks,” [in effect] “shortening and simplifyingintertidal channel networks and <strong>the</strong> shoreline <strong>of</strong> <strong>the</strong> Estuaryas a whole” (Collins 2002). This narrow elevational nicheoverlaps almost precisely with habitat that is critical for <strong>the</strong>California clapper rail (Collins et al. 1994; Albertson andEvens 2000). Given <strong>the</strong> importance <strong>of</strong> channel complexity toclapper rails, a reduction in channel density and complexitypredicted for marshes invaded by non-native Spartina(Collins 2002; Ayres et al. 2003; Baye 2004) is likely toreduce <strong>the</strong> availability <strong>of</strong> habitat for clapper rails.RAIL DENSITY AND HABITAT QUALITYThe population estimates reported above suggest thatclapper rails may occur in dramatically higher densities insome invaded marshes than in non-invaded marshes. Thisapparent selection <strong>of</strong> non-native over native habitats by railspredicates a question: Do increased densities <strong>of</strong> rails meanincreased viability <strong>of</strong> rail populations?Some have interpreted <strong>the</strong> density <strong>of</strong> rails detectedwithin non-native Spartina patches as a positive trend forthis highly endangered taxon. However, <strong>the</strong> literaturecautions that density is a misleading indicator <strong>of</strong> habitatquality or reproductive success (VanHorne 1983; Vickery etal. 1992; Pulliam 1996). Given <strong>the</strong> biogeography <strong>of</strong> tidalmarshlands within <strong>the</strong> estuary, and <strong>the</strong> disparity in habitattypes that exists between marshes in its nor<strong>the</strong>rn andsou<strong>the</strong>rn reaches (Josselyn 1983; Goals Project 1999; Bayeet al. 2000) land managers and biologists should consider<strong>the</strong> possibility that non-native Spartina pastures might beecological traps, or perhaps “attractive sinks,” phenomenathat occur most commonly in anthropogenically modifiedhabitats (Delibes et al. 2001; Schlaepfer et al. 2002; Battin2004; Robertson and Hutto 2006). Indeed, San FranciscoEstuary and <strong>the</strong> California clapper rail fulfill most if not all<strong>of</strong> <strong>the</strong> criteria that increase vulnerability <strong>of</strong> animalpopulations to ecological traps summarized by Battin(2004). It is also worth considering <strong>the</strong> possibility that <strong>the</strong>increasing trends in abundance associated with invadedmarshes (east bayshore) and coincident negative trends inless invaded sites (west bayshore) reported in Broom (2008),suggest <strong>the</strong> possibility that rails are being lured away fromhigher quality habitat (larger patch size, age heterogeneity,native Spartina beds ) to occupy falsely attractive habitat.However, <strong>the</strong> supposition that ecological traps account for<strong>the</strong>se apparent trends is open to o<strong>the</strong>r interpretations (Battin2004; Robertson and Hutto 2006; Gilroy and Su<strong>the</strong>rland2007).To determine <strong>the</strong> impact <strong>of</strong> non-native Spartina beds toclapper rail long-term population viability, studiesdetermining reproductive success, survivorship, andpredation rates in both habitat types will be needed. Futurestudies aimed at determining reproductive success ininvaded and non-invaded habitat may help address thisimportant question <strong>of</strong> conservation biology.SUMMARY AND CONCLUSIONSNon-native cordgrass has invaded intertidal marshlandsin <strong>the</strong> San Francisco Estuary, crucial habitat for <strong>the</strong> endemicand endangered California clapper rail. The ecologicaltransformation <strong>of</strong> <strong>the</strong> intertidal marsh zone is apparentlyunderway and future changes are anticipated. Densities <strong>of</strong>rails in invaded habitat appear to be somewhat higher than at- 187 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinanon-invaded habitat, but <strong>the</strong> ultimate effects on <strong>the</strong>reproductive success and recovery <strong>of</strong> <strong>the</strong> California clapperrail poplation are unknown.The South and Central Bay subpopulations <strong>of</strong> railsoccupy tidal marshes that are undergoing an aggressiveinvasion by non-native cordgrass. The North Bay clapperrail populations occupy marshes that are mostly free <strong>of</strong> nonnativecordgrass. The North Bay is also an area whereextensive marsh restoration efforts are underway or in <strong>the</strong>planning phases (Siegel 2002). These differing situations<strong>of</strong>fer opportunities for:• Non-native cordgrass control in <strong>the</strong> South and Centralbays employing “best management practices” to reduceor minimize impacts to rails, currently in progress by <strong>the</strong> SanFrnacisco Bay Invasive Spartina Project.• Monitoring <strong>of</strong> North Bay marshes for <strong>the</strong> presence <strong>of</strong>non-native cordgrass, particularly in marsh restoration sites.• Control <strong>of</strong> non-native cordgrass in <strong>the</strong> North Bayprior to alteration <strong>of</strong> <strong>the</strong> structure and function <strong>of</strong> <strong>the</strong> habitat.• Research focused on population dynamics <strong>of</strong> rails atdifferent sites (invaded and “pristine”) to determine reproductivesuccess and survivorship.REFERENCESAlbertson, J. and J. Evens. 2000. California Clapper Rail. In:Ol<strong>of</strong>son, P.R., ed. Baylands Ecosystem Species and CommunityPr<strong>of</strong>iles: life histories and environmental requirements <strong>of</strong> keyplants, animals, and wildlife, pp. 332-341. San Francisco BayRegional Water Quality Control Board, Oakland, California.Atwater, B.F., S.G. Conard, J.N. Dowden, C.W. Hedel, R.L.MacDonald, and W. Savage. 1979. History, landforms, and vegetation<strong>of</strong> <strong>the</strong> estuary’s tidal marshes. In: Conomos, T.J., ed. SanFrancisco Bay: <strong>the</strong> urbanized estuary, pp. 348-385. Pacific Division,American Association for <strong>the</strong> Advancement <strong>of</strong> Science. SanFrancisco, California.Avocet Research Associates. 1992. Evaluation <strong>of</strong> impacts <strong>of</strong> NavalRiverine Forces Training Operations on nesting habitat <strong>of</strong> <strong>the</strong>California Clapper Rail at Napa River, California. Report toU.S. Department <strong>of</strong> <strong>the</strong> Navy, Western Division, San Bruno,California.Avocet Research Associates. 2004. California Clapper Rail (Ralluslongirostris obsoletus): breeding season surveys, San Pablo Bayand tributaries—2004. Final report to Marin Audubon Societyfrom Avocet Research associates.Ayres, D.R., D.R. Strong, and P. Baye. 2003. Spartina foliosa(Poaceae)—A common species on <strong>the</strong> road to rarity? Madrono50:209-213.Battin, J. 2004. When good animals love bad habitats: Ecologicaltraps and <strong>the</strong> conservation <strong>of</strong> animal populations. ConservationBiology 18: 1482-1491.Baye, P., P.M. Faber, and B. Grewell. 2000. Tidal marsh plants <strong>of</strong><strong>the</strong> San Francisco Estuary. In: Ol<strong>of</strong>son, P.R., ed. Baylands EcosystemSpecies and Community Pr<strong>of</strong>iles: life histories and environmentalrequirements <strong>of</strong> key plants, animals, and wildlife, pp.9-67. San Francisco Bay Regional Water Quality Control Board,Oakland, California.Baye, P. 2004. A Review and Assessment <strong>of</strong> Potential Long-termEcological Consequences <strong>of</strong> <strong>the</strong> Introduced Cordgrass Spartinaalternifora in <strong>the</strong> San Francisco Estuary. Draft ms.Broom, J. 2008. Regional Trends in California Clapper RailAbundance at Non-native Spartina-invaded Sites in <strong>the</strong> SanFrancisco Estuary from 2005 to 2007. Report prepared by Ol<strong>of</strong>sonEnvironmental, Inc. for <strong>the</strong> State Coastal Conservancy.http://www.spartina.org/project_documents/clapper_rails/CLRA_Trends_2005-07(web).pdf.Callaway, J.C., and M.N. Josselyn. 1992. The introduction andspread <strong>of</strong> smooth cordgrass (Spartina alterniflora) in South SanFrancisco Bay. Estuaries 15(2):218-226.Cohen, A.N. 2001. Conservation management <strong>of</strong> exotic cordgrassin <strong>the</strong> Pacific coast salt marshes: preserving clapper rails vs.preserving <strong>the</strong> clapper rail. Abstract for Second Symposium onMarine Conservation Biology. San Francisco state University,June 21-26, 2001.Collins, J. 2002. Invasion <strong>of</strong> San Francisco Bay by SmoothCordgrass, Spartina alterniflora: a forecast <strong>of</strong> geomorphic effectson <strong>the</strong> intertidal zone. Report prepared for U.S. EPA Region9.Collins, J.N., J.G. Evens, and B. Grewell. 1994. A synoptic survey<strong>of</strong> <strong>the</strong> distribution and abundance <strong>of</strong> <strong>the</strong> California Clapper Rail,Rallus longirostris obsoletus, in <strong>the</strong> nor<strong>the</strong>rn reaches <strong>of</strong> <strong>the</strong> SanFrancisco Estuary during <strong>the</strong> 1992 and 1993 breeding seasons.Draft Technical Report to California Department <strong>of</strong> Fish andGame (8/1/94).Conomos, T.J., ed. 1979. San Francisco Bay: The UrbanizedEstuary. http://www.estuaryarchive.org/archive/conomos_1979.Daehler, C.C. and D.R. Strong. 1996. Status, prediction, andprevention <strong>of</strong> introduced cordgrass Spartina spp. invasions inPacific estuaries, USA. Biological Conservation 78:51-58.Dedrick, K. 1989. San Francisco bay tidal marshland acreages:recent and historic values. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> 6 th Symposium inCoastal and Ocean Management (Coastal Zone 1989). Am. Soc.Of Engineers, pp. 383-398.Delibes, M., P. Gaona, and P. Ferreras. 2001. Effects <strong>of</strong> an attractivesink leading into maladaptive habitat selection. AmericanNaturalist 158: 277-285.Estrella, S. California Clapper Rail and California Black Rail:Suisun Marsh Survey 2007. California Department <strong>of</strong> Fish andGame Bay-Delta Region. Report to California Department <strong>of</strong>Water Resources. Contract #4600004374.Evens, J.G. and G.W. Page. 1987. Clapper rail populations atCorte Madera Creek, Muzzi Marsh, San Clemente Creek, andTriangle Marsh, Marin County, California. Report to MarinAudubon Society from Point Reyes Bird Observatory.Evens, J.G. and J.N. Collins. 1992. Distribution, abundance, andhabitat affinities <strong>of</strong> <strong>the</strong> California Clapper Rail, Rallus longirostrisobsoletus, in <strong>the</strong> nor<strong>the</strong>rn reaches <strong>of</strong> <strong>the</strong> San FranciscoEstuary during <strong>the</strong> 1992 breeding season. Report by AvocetResearch Associates to California Department <strong>of</strong> Fish and Game.Gill, R. 1979. Status and distribution <strong>of</strong> <strong>the</strong> California Clapper Rail(Rallus longirostris obsoletus). California Fish and Game 65:36-49.Gilroy, J.J. and W.J. Su<strong>the</strong>rland. 2007. Beyond ecological traps:perceptual errors and undervalued resources. Trends in Ecologyand Evolution 22(7):351-6.Goals Project. 1999. Baylands ecosystem habitat goals. A report <strong>of</strong>habitat recommendations prepared by <strong>the</strong> San Francisco BayArea Wetlands Ecosystem Goals Project. USEPA, San Francisco/SFBRWQCB,Oakland, CA.Grinnell, J. and A. Miller. 1944. Distribution <strong>of</strong> <strong>the</strong> Birds <strong>of</strong>California. Pacific Coast Avifauna No. 27.Herzog, M., L. Liu, J. Evens, N. Nur, and N. Warnock. 2005.Temporal and spatial patterns in population trends in California- 188 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaClapper Rail (Rallus longirostris obsoletus). 2005 ProgressReport to California Department <strong>of</strong> Fish and Game and U.S. Fishand Wildlife Service. PRBO Conservation Science.Hogle, I. 2008. San Francisco Estuary Invasive Spartina ProjectMonitoring Report for 2006. Appendix 3. Report prepared byOl<strong>of</strong>son Environmental, Inc. for <strong>the</strong> State Coastal Conservancy.http://www.spartina.org/project_documents/2006_MonRept_App3_Maps_Web.pdfJosselyn, M. 1983. The ecology <strong>of</strong> <strong>the</strong> San Francisco Bay tidalmarsh: a community pr<strong>of</strong>ile. U.S. Fish and Wildlife Service,Division <strong>of</strong> Biological Services, Washington, D.C. FWS/OBS-83/23.Lonzarich, D.G., T.E. Harvey, and J.E. Takekawa. 1992. Traceelement and organochlorine concentrations in California ClapperRail (Rallus longirostris obsoletus) eggs. Archives <strong>of</strong> EnvironmentalContamination and Toxicology 23:147-153.Pulliam, H.R. 1996. Sources and sinks: Empirical evidence andpopulation consequences. Rhodes, O.E., R.K. Chesser, and M.H.Smith, eds. In: Population dynamics in ecological space andtime, pp 45-69. Univ. Chicago Press.Robertson, B.A. and R.L. Hutto. 2006. A framework for understandingecological traps and an evaluation <strong>of</strong> existing evidence.Ecology 87: 1075-1085Schwarzbach, S.E., J.D. Albertson, and C.M. Thomas. 2006,Effects <strong>of</strong> predation, flooding, and contamination on <strong>the</strong> reproductivesuccess <strong>of</strong> California Clapper Rails (Rallus longirostrisobsoletus) in San Francisco Bay: Auk, v. 123, no. 1, p. 45–60.Siegel, S. 2002. Wetland Restoration & Enhancement Projects:Completed and Planned in <strong>the</strong> North Bay, San Francisco Estuary,California (January 2002). Produced by Wetland WaterResources for CALFED Bay-Delta Program.Schlaepfer, M.A., M.C. Runge, and P.W. Sherman, 2002. Ecologicaland evolutionary traps. Trends in Ecology and Evolution 17:474-480.Spautz, H. and N. Nur. 2004. Impacts <strong>of</strong> Non-native PerennialPepperweed (Lepidium latifolium) on Abundance, Distributionand Reproductive Success <strong>of</strong> San Francisco Bay Tidal MarshBirds. Point Reyes Bird Observatory Conservation Science.http://www.prbo.org/cms/docs/wetlands/lepidium04.pdfSpautz, H., J. McBroom, K. Zaremba, J. Evens, J. Albertson, S.Bobzien, and M. Herzog. 2006. California Clapper Rails in SanFrancisco Bay: Modeling habitat relationships at multiple scalesto guide habitat restoration and eradication on non-nativeSpartina. Poster paper presented at South Bay Salt Pond Symposium.June 6, 2006. San Jose State University and at CALFEDU.S. Fish and Wildlife Service. 2001. Draft Survey Protocol:California Clapper Rail. (1/21/2001).VanHorne, B. 1983. Density as a misleading indicator <strong>of</strong> habitatquality. J. Wildlife Management 47:893-901.Vickery, P.D., M.L. Hunter, and J.V. Wells. 1992. Is density anindicator <strong>of</strong> breeding success? Auk 109:706-710.- 189 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaTHE EFFECTS OF GRAZING GEESE ON HYBRID AND NATIVE SPARTINA IN SANFRANCISCO BAYE.D. GROSHOLZ 1 AND R.E. BLAKE 2Department <strong>of</strong> Environmental Science and Policy, University <strong>of</strong> California, Davis, One Shields Ave., Davis, CA 95616tedgrosholz@ucdavis.edu 1 ; reblake@vims.edu 2The invasion <strong>of</strong> Spartina alterniflora has become one <strong>of</strong> <strong>the</strong> most significant invasions in <strong>the</strong>already heavily invaded San Francisco Bay. The Spartina invasion has resulted in many changes tocommunity and ecosystem processes at lower trophic levels, however, we have so far failed todocument equivalent changes at higher trophic levels that are likely occurring as well. In this study,we report <strong>the</strong> consequences <strong>of</strong> <strong>the</strong> hybrid invasion with respect to an important vertebrate grazer,western Canada geese (Branta canadensis m<strong>of</strong>fitti). Canada geese regularly nest during wintermonths in <strong>the</strong> vicinity <strong>of</strong> <strong>the</strong> invaded central and sou<strong>the</strong>rn areas <strong>of</strong> San Francisco Bay. We foundthat Canada geese intensively graze native Spartina foliosa at several study sites in <strong>the</strong> central bayregion removing up to 90% <strong>of</strong> <strong>the</strong> aboveground vegetation. However, geese completely ignorehybrid Spartina with virtually no evidence <strong>of</strong> grazing in hybrid areas adjacent to heavily grazednative Spartina areas. Experiments with captive geese demonstrated that geese repeatedly preferrednative Spartina when presented intact clones <strong>of</strong> hybrid and native clones in side-by-side preferencetrials. However, when cut stems were presented in similar trials, geese showed no preference for <strong>the</strong>native over <strong>the</strong> hybrid. We conclude that <strong>the</strong> preference demonstrated in <strong>the</strong> first experiment wasnot <strong>the</strong> result <strong>of</strong> plant defensive chemistry, but <strong>the</strong> result <strong>of</strong> physical differences between intactclones and cut stems that geese could assess. Field exclosure experiments conducted for two years inareas where hybrid Spartina was overgrowing native Spartina showed that grazing <strong>of</strong> nativeSpartina by geese resulted in 25% greater rates <strong>of</strong> lateral spread <strong>of</strong> <strong>the</strong> hybrid into native areas. Thissuggests that grazing by geese may be accelerating <strong>the</strong> rate <strong>of</strong> invasion and ultimate replacement <strong>of</strong>native Spartina by hybrid Spartina.Keywords: Canada geese, San Francisco Bay, hybrid Spartina, grazing, invasion rateINTRODUCTIONThe introduction <strong>of</strong> <strong>the</strong> smooth cordgrass (Spartinaalterniflora) has been among <strong>the</strong> most significant <strong>of</strong> <strong>the</strong>nearly 250 non-native species invasions recorded in SanFrancisco Bay (Cohen and Carlton 1998). Spartinaalterniflora became established in 1976 as <strong>the</strong> result <strong>of</strong> anintentional introduction by <strong>the</strong> Army Corp <strong>of</strong> Engineers formitigation purposes (Faber 2000). Following establishment,S. alterniflora hybridized with <strong>the</strong> native cordgrass S.foliosa. Hybrid Spartina has rapidly colonized many areas<strong>of</strong> central and south San Francisco Bay (Daehler and Strong1997; Ayres et al. 2004). Hybrid Spartina can colonize openmudflats as well as out-compete native vegetation at highertidal heights (Ayres et al. 2004).Recent studies have shown that <strong>the</strong> ecologicalrepercussions <strong>of</strong> <strong>the</strong> hybrid Spartina invasion are farreaching with widespread impacts on community structureand ecosystem function (Neira et al. 2005, 2006, 2007;Levin et al. 2006). However, work to date has focused onlyon changes at lower trophic levels, despite <strong>the</strong> fact thatvertebrate herbivores are known to consume S. alterniflorain its native region (Buchsbaum et al. 1981). Little is knownregarding <strong>the</strong> consequences <strong>of</strong> hybridization forsusceptibility to herbivores <strong>of</strong> any kind (Daehler and Strong1997).Among <strong>the</strong> most common and potentially importantvertebrate herbivores <strong>of</strong> Spartina are western Canada geese,Branta canadensis m<strong>of</strong>fitti (Banks et al. 2004). The westernsubspecies is <strong>the</strong> only one regularly found in San FranciscoBay (Mowbray et al. 2002) and <strong>the</strong>y typically nest fromNovember to April. While nesting, <strong>the</strong>ir foraging isgenerally restricted to nearby areas that typically includemudflats with native Spartina. Our initial observationssuggested that geese were grazing intensively on areasoccupied by native Spartina. One <strong>of</strong> <strong>the</strong> key questions <strong>of</strong>this study was to determine whe<strong>the</strong>r grazing by westernCanada geese could influence <strong>the</strong> rate <strong>of</strong> invasion <strong>of</strong> hybridSpartina in San Francisco Bay. Therefore, our goal was tomeasure grazing by Canada geese on native Spartina andcompare this with grazing on hybrid Spartina. If we foundevidence <strong>of</strong> selective grazing, <strong>the</strong> next goal would be todetermine whe<strong>the</strong>r selective grazing was due to ameasurable preference for one plant over <strong>the</strong> o<strong>the</strong>r. Wewould also determine <strong>the</strong> basis for any preference that mightexist. Finally, we would quantify any influence that geese- 191 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinagrazing might have on <strong>the</strong> rate <strong>of</strong> invasion <strong>of</strong> hybridSpartina into areas occupied by native Spartina.METHODSGrazing Intensity in San Francisco BayTo estimate <strong>the</strong> intensity <strong>of</strong> grazing by western Canadageese, we measured grazing at three sites in <strong>the</strong> centralportion <strong>of</strong> San Francisco Bay: Point Isabel (37º53’30’’N;122º20’55’’W), Robert’s Landing (37º40’13’’N;122º28’48’’W), and Oro Loma (37º37’45’’N;122º09’08’’W). These sites were distributed along <strong>the</strong>eastern margin <strong>of</strong> San Francisco Bay and each containedareas <strong>of</strong> native S. foliosa with varying degrees <strong>of</strong> cover byhybrid Spartina. At each site, we laid 30 meter (m) transectsalong which 10 0.25 m x 0.25 m quadrats were established atrandomly chosen points. Plant samples to verify geneticidentification (hybrid vs. native) were taken within severalquadrats and kept refrigerated until genetic assays usingrandom amplified polymorphic DNA (RAPD) markers couldbe conducted at <strong>the</strong> Spartina genetics lab at UC Davis(Ayres, unpublished results), typically less than a week aftercollection. Genetic determination followed previouslypublished protocols (Ayres et al. 1999). In some cases, <strong>the</strong>area covered by native S. foliosa was smaller so shortertransects with fewer quadrats were used. Within eachquadrat, we measured <strong>the</strong> number <strong>of</strong> stems that had beengrazed or not by geese and <strong>the</strong> heights <strong>of</strong> 15 stems to <strong>the</strong>nearest one centimeter (cm). Grazing produces characteristicdamage that is easily identified. At sites where we also haddata for hybrid Spartina, <strong>the</strong> same measurements were madealong similar transects. We tested <strong>the</strong> proportion <strong>of</strong> stemsgrazed by geese between native and hybrid transects.Field Grazing ExclosuresTo experimentally measure <strong>the</strong> impact <strong>of</strong> grazingCanada geese on both native and hybrid Spartina in <strong>the</strong>field, we conducted exclosure experiments at <strong>the</strong> RobertsLanding site. At this site, we established goose exclosuresin three areas (six replicates per treatment). The first areawas in a continuous meadow <strong>of</strong> native S. foliosa. Thesecond was in a continuous meadow <strong>of</strong> hybrid Spartina. Thethird was located where hybrid Spartina was overgrowingnative Spartina. In single species stands we constructed a 1m x 1 m exclosure cage <strong>of</strong> plastic mesh (4 cm x 4 cmopenings) attached by electrical cable ties to polyvinylchloride (PVC) posts. At border areas with both native andhybrid Spartina, we established exclosure cages 1 m x 4 mwith <strong>the</strong> long axis <strong>of</strong> <strong>the</strong> exclosure perpendicular to <strong>the</strong>hybrid-native border. The center <strong>of</strong> <strong>the</strong> exclosure (at 2 m)was positioned to be over <strong>the</strong> approximate border edge at <strong>the</strong>time <strong>of</strong> construction. Adjacent control areas for both singlespecies exclosures and border exclosures were establishedwithin 2 m <strong>of</strong> <strong>the</strong> exclosure cages. The control areas weredelineated with PVC posts at <strong>the</strong> corners. We estimated <strong>the</strong>linear distance in spread for control and exclosure treatmentsby measuring <strong>the</strong> distance (to nearest 0.1 m) from <strong>the</strong> hybridborder at <strong>the</strong> start <strong>of</strong> <strong>the</strong> experiment to <strong>the</strong> leading edge <strong>of</strong><strong>the</strong> spread after two years. We tested differences withAnalysis <strong>of</strong> Variance (ANOVA) using log transformedvalues for distance and treatment (exclosure vs. control) asfixed factors using SAS version 9.1 (SAS Institute, Cary,North Carolina).Grazing Trials with Intact ClonesIn order to determine <strong>the</strong> extent that observed grazingwas <strong>the</strong> result <strong>of</strong> preference for native Spartina ra<strong>the</strong>r thanavailability or o<strong>the</strong>r factors, we conducted grazing trialsduring February-March 2003 using a small group (six) <strong>of</strong>captive western Canada geese maintained by <strong>the</strong> WildlifeDepartment at Humboldt State University (HSU), Arcata,California. This group <strong>of</strong> geese had been maintainedtoge<strong>the</strong>r for several years on processed diets. Theexperiment involved using intact segments (0.5 m x 0.5 m x0.5 m) <strong>of</strong> two clones <strong>of</strong> ei<strong>the</strong>r hybrid or native Spartina<strong>of</strong>fered in brown plastic bins. Each plant segment had beenexcavated from field sites in San Francisco Bay (hybridplants from Elsie Roemer Bird Sanctuary, City <strong>of</strong> Alamedaand S. foliosa from Bothin Marsh, City <strong>of</strong> Mill Valley) anddriven to <strong>the</strong> Humboldt aviary <strong>the</strong> next day. All stems andleaves were counted for each clone prior to <strong>the</strong> start <strong>of</strong> <strong>the</strong>experiment. At <strong>the</strong> beginning <strong>of</strong> <strong>the</strong> experiment, <strong>the</strong> tubswith <strong>the</strong> clone segments were established in a two by twoarray approximately 2 m apart. The geese were <strong>the</strong>n allowedto enter <strong>the</strong> aviary and encounter <strong>the</strong> clones. We made focalobservations on each clone at one minute intervals noting <strong>the</strong>number <strong>of</strong> geese grazing on each. After two hours, wecollected <strong>the</strong> clones and measured <strong>the</strong> number <strong>of</strong> stems andleaves that had been grazed by geese. The entire experimentwas repeated three times over <strong>the</strong> following three weeks.We estimated <strong>the</strong> intensity <strong>of</strong> grazing by counting <strong>the</strong>percentage <strong>of</strong> leaves and stems that had been grazed during<strong>the</strong> trial for both hybrid and native clones. We testeddifferences in grazing with ANOVA using arc-sintranformed percentage <strong>of</strong> leaves or stems with treatment as afixed factor (as above).Grazing Trials with Clipped StemsTo determine whe<strong>the</strong>r <strong>the</strong> preferences <strong>of</strong> experimentalgeese were determined by differences in <strong>the</strong> physicalproperties <strong>of</strong> <strong>the</strong> plants versus differences in plant chemistry,we used <strong>the</strong> same group <strong>of</strong> geese during experiments fromFebruary-March 2004. We collected stems <strong>of</strong> both nativeand hybrid Spartina at <strong>the</strong> same sites used for collecting <strong>the</strong>clones <strong>the</strong> previous year. We cut <strong>the</strong> stems at <strong>the</strong> base <strong>of</strong> <strong>the</strong>plant and wrapped <strong>the</strong>m in large plastic bags for shipment.Stems were kept refrigerated from <strong>the</strong> time <strong>of</strong> clipping andsent to <strong>the</strong> aviary at HSU within 24 hours and were usedimmediately <strong>the</strong>reafter in feeding trials. Geese showed <strong>the</strong>- 192 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartinasame interest in <strong>the</strong> clipped stems as <strong>the</strong>y did in <strong>the</strong> intactclones, <strong>the</strong>refore we assume that <strong>the</strong> condition <strong>of</strong> <strong>the</strong> stemsand leaves were similar to that in <strong>the</strong> previous experiment.In this second experiment, we <strong>of</strong>fered <strong>the</strong> geese a choice<strong>of</strong> cut stems <strong>of</strong> ei<strong>the</strong>r hybrid Spartina or native S. foliosausing <strong>the</strong> same bins <strong>the</strong> geese are normally <strong>of</strong>fered <strong>the</strong>irpelleted feed. We arranged three bins <strong>of</strong> each <strong>of</strong> <strong>the</strong> twoplant types in a 2 x 3 array with each bin approximately 2 mfrom <strong>the</strong> o<strong>the</strong>rs. After setting up <strong>the</strong> feeding array, <strong>the</strong> geesewere allowed to graze for at least one hour after which all<strong>the</strong> plants remaining in <strong>the</strong> bins were collected. Also, plantsthat had been removed from <strong>the</strong> bins and discarded werealso collected and stored separately. We repeated thisexperiment two times over <strong>the</strong> following eight weeks. Weestimated <strong>the</strong> intensity <strong>of</strong> grazing by weighing <strong>the</strong> dryweight <strong>of</strong> <strong>the</strong> plants in each bin remaining at <strong>the</strong> end <strong>of</strong> <strong>the</strong>experiment as well as <strong>the</strong> discarded plants adjacent to thatbin. We tested differences in grazing with ANOVA usinglog-transformed weights for ei<strong>the</strong>r bins or discards withtreatment as a fixed factor (as above).RESULTSGrazing Intensity in San Francisco BayOur data from several sites in San Francisco Bay showintense grazing by western Canada geese on native S.foliosa. Our most extensive data from Robert’s Landing(Fig. 1) show greater than 90% loss <strong>of</strong> aboveground biomasswith nearly all stems grazed to within a few centimeters <strong>of</strong><strong>the</strong> ground on all three transects (Fig. 1). By contrast, geesevirtually ignored hybrid Spartina with no measurablegrazing on any plants along <strong>the</strong> transect. Similarly intensegrazing was measured at Oro Loma (greater than 85% <strong>of</strong>stems grazed) and Pt. Isabel (greater than 95% <strong>of</strong> stemsgrazed).Field Grazing ExclosuresHybrid Spartina grew laterally into <strong>the</strong> native Spartinazones in nearly all <strong>of</strong> <strong>the</strong> experimental areas independent <strong>of</strong>treatment. However, <strong>the</strong> rate <strong>of</strong> lateral spread <strong>of</strong>aboveground plant biomass was significantly greater(p

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaRemaining Plants (%)100806040200StemsLeavesHybridS. foliosaFig. 3. Results <strong>of</strong> feeding preference experiments showing means (+ 1s.e.) <strong>of</strong> remaining stems and leaves <strong>of</strong> hybrid (black bars) and nativeSpartina (white bars) after being grazed by captive geese.Grazing Trials with Clipped StemsThe results <strong>of</strong> <strong>the</strong> second set <strong>of</strong> feeding trials withclipped stems produced very different results than <strong>the</strong> trialsinvolving intact clones. The captive geese showed nosignificant preference for clipped stems <strong>of</strong> native Spartinacompared with stems <strong>of</strong> hybrid Spartina. Geese grazing wasvigorous overall and geese spent similar amounts <strong>of</strong> timegrazing as <strong>the</strong>y did in <strong>the</strong> trials with <strong>the</strong> clones (p>0.05).However, nei<strong>the</strong>r <strong>the</strong> amount <strong>of</strong> biomass consumed bygrazing geese nor <strong>the</strong> amount discarded was significantlydifferent for hybrid stems relative to native stems (p>0.25for all).DISCUSSIONOur findings from <strong>the</strong> field surveys suggest that grazingby western Canada geese on native S. foliosa is widespreadthroughout <strong>the</strong> central portion <strong>of</strong> San Francisco Bay. Datafrom at least five sites showed a large reduction in <strong>the</strong>aboveground biomass <strong>of</strong> native Spartina. By contrast,adjacent areas where hybrid Spartina was present showedlittle sign <strong>of</strong> grazing. Geese appeared to ignore hybridSpartina entirely when it was present in our surveys. Thereis <strong>the</strong> possibility that geese may consume a higherproportion <strong>of</strong> hybrid genotypes that are morphologicallymore similar to native S. foliosa, however, we have not yetseen evidence for this.Our results from <strong>the</strong> goose exclusion studies indicatethat lateral spread <strong>of</strong> <strong>the</strong> hybrid into areas occupied by nativeS. foliosa is more rapid in areas where goose grazing hasbeen excluded for two years. The rate <strong>of</strong> lateral spread is25% greater than in adjacent control areas where goosegrazing continued. This 25% increase in lateral spread mayhave significant consequences for <strong>the</strong> overall rate <strong>of</strong>invasion <strong>of</strong> hybrid Spartina in San Francisco Bay. Although<strong>the</strong> rate <strong>of</strong> sexual reproduction <strong>of</strong> hybrid Spartina isoverwhelming native reproduction due to higher seed set andpollen swamping, plant clonal growth is also important formaintaining its presence in <strong>the</strong> bay. The degree to whichhybrid plants can overgrow and outcompete native plantswill have a significant effect on <strong>the</strong> speed that hybridSpartina may be able to drive S. foliosa locally extinct incentral San Francisco Bay (Ayres et al. 2004).The eventual replacement <strong>of</strong> native with hybrid mayalso have some negative consequences for <strong>the</strong> nesting geese.Native Spartina is not likely to be <strong>the</strong> only food source fornesting geese in an urbanized estuary like San FranciscoBay, where lawns and golf courses <strong>of</strong>fer easy forage.However, <strong>the</strong> high levels <strong>of</strong> grazing on native Spartina at ourstudy sites suggest that this is still an important source <strong>of</strong>forage for <strong>the</strong>se birds. O<strong>the</strong>r sources <strong>of</strong> forage are not likelyas proximate to nesting sites as native Spartina and,<strong>the</strong>refore may be costly in terms <strong>of</strong> foraging time and timespent away from <strong>the</strong> nest. This ultimately may affect nestingsuccess. Unfortunately, <strong>the</strong>re are no data with which to test<strong>the</strong>se possibilities.The results from our experiments with captive geesestrongly support our observations in <strong>the</strong> field. The data from<strong>the</strong> first set <strong>of</strong> aviary trials show a strong ability <strong>of</strong> geese todiscriminate between native S. foliosa and hybrid Spartina inside-by-side choice trials. This supports <strong>the</strong> idea that geeseare also discriminating among <strong>the</strong>se same plant types in <strong>the</strong>field. It is possible that geese may also discriminate amongdifferent hybrids, possibly preferring hybrids that moreclosely resemble <strong>the</strong> native S. folioa. However, we cannotdetermine that from <strong>the</strong> current study.Our data from <strong>the</strong> second set <strong>of</strong> aviary trials isconsistent with <strong>the</strong> idea that geese are able to distinguishhybrid from native Spartina based on physicalcharacteristics ra<strong>the</strong>r than chemical characteristics. Incontrast to <strong>the</strong> first trials (plants upright and intact), resultsfrom <strong>the</strong>se trials (cut stems laid sideways in trays) showedthat geese exhibited no ability to distinguish hybrid fromnative.Physical characteristics <strong>of</strong> intact plants may allow geeseto assess palatability <strong>of</strong> potential forage by tugging on <strong>the</strong>plant and gauging its resistance (Lang and Black 2001).Physical differences between native and hybrid Spartina areimmediately obvious to any human investigator as hybridstems and leaves are much thicker, tougher and morefibrous. Results from <strong>the</strong> second experiment, designed toeliminate physical differences experienced by geese tuggingon leaves and stems, showed that geese failed todiscriminate between native and hybrid Spartina. Thegeneral conclusion from studies <strong>of</strong> secondary defensive plantcompounds is that if by removing physical differences, <strong>the</strong>reis no subsequent preference shown by <strong>the</strong> herbivore, <strong>the</strong>nplant chemistry can generally be ruled out as important. If- 194 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartinageese preference in <strong>the</strong> first experiment was based ondifferences in defensive chemistry between native S. foliosaand hybrid Spartina, <strong>the</strong>y would presumably still be able touse those cues in <strong>the</strong> second trial. We also measureddifferences in <strong>the</strong> amount <strong>of</strong> material that geese discarded toexplore <strong>the</strong> idea that geese might discard a greaterproportion <strong>of</strong> <strong>the</strong> hybrid if defensive chemistry was a cue.However, we measured no significant differences indiscarded material between native and hybrid Spartina.Our results parallel o<strong>the</strong>r studies <strong>of</strong> factors influencinggoose grazing. Earlier work by Buchsbaum et al. (1981)with eastern Canada geese (Branta canadensis canadensis)suggested that geese can detect defensive compounds suchas ferulic acid in lawn grass. Their results suggest thatphenolics at relatively high concentrations could beimportant in grazing preferences. O<strong>the</strong>r studies havegenerally found less support for <strong>the</strong> role <strong>of</strong> phenolicsinfluencing goose grazing. A study with greater snow geese(Anser caerulescens atlantica) tested <strong>the</strong> influence <strong>of</strong> water,fiber, phenolic and protein content among grass species andfound that only water content was important (Gauthier andBedard 1991). Ano<strong>the</strong>r study using barnacle geese (Brantaleucopsis) found that feeding preferences <strong>of</strong> geese were bestcorrelated with water and nitrogen content, although <strong>the</strong>relationship with water was stronger (Owen et al. 1977).Coleman and Boag (1987) found that for western Canadageese non-structural carbohydrate and not fiber or proteincontent was <strong>the</strong> best predictor <strong>of</strong> preferences among foragespecies. Structural compounds such as cellulose have alsobeen found to be poorly digested by geese relative toproteins and soluble carbohydrates (Buchsbaum et al. 1986).The conclusions from our study are generally consistent with<strong>the</strong>se o<strong>the</strong>r studies that found little evidence for a strong rolefor defensive chemicals such as phenolics in determining <strong>the</strong>observed preferences and stronger influences <strong>of</strong> structuralcompounds, nitrogen and water content. Interestingly,native S. foliosa has a lower carbon to nitrogen (C:N) ratio(more protein, less cellulose) relative to hybrid Spartina(Tyler et al. 2007), and also has a less coarse physicalstructure and presumably higher water content.In summary, our results suggest that Canada geese maybe influencing <strong>the</strong> rate <strong>of</strong> spread <strong>of</strong> hybrid Spartina in <strong>the</strong>central portion <strong>of</strong> San Francisco Bay. The eventual loss <strong>of</strong>native S. foliosa due to <strong>the</strong> hybrid invasion, at least in someportion <strong>of</strong> San Francisco Bay, could also have negativeconsequences for foraging by removing a potentiallyimportant food source. The intense and highly selectivegrazing on native Spartina and almost total avoidance <strong>of</strong> <strong>the</strong>hybrid results from a strong preference for <strong>the</strong> native that isapparently based on structural ra<strong>the</strong>r than chemicaldifferences.ACKNOWLEDGMENTSOur greatest thanks to J. Black, Humboldt StateUniversity, for many informative discussions <strong>of</strong> geese andgrazing, his very generous assistance and collaboration with<strong>the</strong> HSU captive goose facility and his help with permitting,logistics, and procedures for behavioral studies. We alsothank HSU students E. Bjerre and K. Spragens for <strong>the</strong>irgenerous assistance with followup experiments, aviarylogistics and Spartina care. We also thank S. Norton, N.Rayl, U. Mahl, N. Christiansen and C. Love for help withfield grazing studies, plant collections, sample processingand data entry.REFERENCESAyres, D.R., D. Garcia-Rossi, H.G. Davis, and D.R. Strong1999. Extent and degree <strong>of</strong> hybridization between exotic(Spartina alterniflora) and native (S. foliosa) cordgrass(Poaceae) in California, USA determined by random amplifiedpolymorphic DNA (RAPDs). Molecular Ecology 8:1179-1186.Ayres D.R., D.L. Smith, K. Zaremba, S. Klohr, and D.R.Strong. 2004. Spread <strong>of</strong> exotic cordgrasses and hybrids(Spartina sp.) in <strong>the</strong> tidal marshes <strong>of</strong> San Francisco Bay,California, USA. Biological Invasions 6: 221-231.Banks, R.C., C. Cicero, J.L. Dunn, A.W. Kratter, P.C. Rasmussen,J.V. Remsen, J.D. Rising, and D.F. Stotz. 2004.Forty-fifth supplement to <strong>the</strong> American Ornithologists'Union Check-list <strong>of</strong> North American Birds. Auk 121:985-995.Buchsbaum, R., I. Valiela, and J. Teal. 1981. Grazing byCanada geese and related aspects <strong>of</strong> <strong>the</strong> chemistry <strong>of</strong> saltmarsh vegetation. Colonial Waterbirds 4: 126-131.Buchsbaum, R., J. Wilson, and I.Valiela. 1986. Digestibility<strong>of</strong> plant constituents by Canada geese and Atlantic Brant.Ecology 67: 386-393.Coleman, T.S., and D.A. Boag. 1987. Canada goose foods:<strong>the</strong>ir significance to weight gain. Wildfowl 38: 82-88.Cohen, A.N., and J.T. Carlton. 1998. Accelerating invasionrate in a highly invaded estuary. Science 279: 555-558.Daehler, C.C., and D.R. Strong. 1997. Reduced herbivoreresistance in introduced smooth cordgrass (Spartina alterniflora)after a century <strong>of</strong> herbivore-free growth.Oecologia 110: 99-108.Faber, R.M. 2000. Good intentions gone awry. CaliforniaCoast and Ocean 16: 14-17.Gauthier, G., and J. Bedard. 1991. Experimental tests <strong>of</strong> <strong>the</strong>palatability <strong>of</strong> forage plants in greater snow geese. Journal<strong>of</strong> Applied Ecology 28: 491-500.Lang, A., and J.M. Black. 2001. Foraging efficiency in BarnacleGeese Branta leucopsis: a functional response tosward height and an analysis <strong>of</strong> sources <strong>of</strong> individualvariation. Wildfowl 52: 7-20.Levin, L.A., C. Neira, and E.D. Grosholz. 2006. Invasivecordgrass modifies wetland trophic function. Ecology 87:419-432.- 195 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaNeira, C., L.A. Levin, and E.D. Grosholz. 2005. Benthicmacr<strong>of</strong>aunal communities <strong>of</strong> three Spartina-hybrid invadedsites in San Francisco Bay, with comparison to uninvadedhabitats. Marine Ecology Progress Series292:111-126.Neira, C., E.D. Grosholz, L.A. Levin, and R. Blake. 2006.Mechanisms generating modification <strong>of</strong> benthos followingtidal flat invasion by a Spartina (alterniflora X foliosa)hybrid. Ecological Applications 16:1391-1404.Neira, C., L.A. Levin, E.D. Grosholz, and G. Mendoza.2007. The influence <strong>of</strong> invasive Spartina growth phaseson associated macr<strong>of</strong>aunal communities. Biological Invasions9:975-993.Owen, M., M. Nugent, and N. Davies. 1977. Discriminationbetween grass species and nitrogen-fertilized vegetationby young Barnacle Geese. Wildfowl 28:21-26.Poole, A. (Editor). 2005. The birds <strong>of</strong> North America online:http://bna.birds.cornell.edu/BNA/. Cornell Laboratory <strong>of</strong>Ornithology, Ithaca, NY.Tyler, A.C., J.G. Lambrinos, and E.D. Grosholz. 2007. Nitrogeninputs promote <strong>the</strong> spread <strong>of</strong> an invasive marshgrass. Ecological Applications 17:1886-1898.- 196 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive SpartinaIMPACT OF INVASIVE HYBRID CORDGRASS (SPARTINA ALTERNIFLORA X FOLIOSA) ONSONG SPARROW AND MARSH WREN POPULATIONS IN SAN FRANCISCO BAY SALTMARSHESJ.C. NORDBY 1,4 , J.T. MCBROOM 2 ,A.N.COHEN 3 AND S.R. BEISSINGER 11Department <strong>of</strong> Environmental Science, Policy and Management, University <strong>of</strong> California, Berkeley, 151 Hilgard Hall #3110Berkeley, CA, 947202San Francisco Estuary Invasive Spartina Project, 2612-A 8 th Street, Berkeley, CA 947103 San Francisco Estuary Institute, 7770 Pardee Lane, 2 nd Floor, Oakland, CA 946214Current address: Institute <strong>of</strong> <strong>the</strong> Environment and Department <strong>of</strong> Ecology and Evolutionary Biology, University <strong>of</strong>California, Los Angeles, Los Angeles, CA 90095; nordby@ucla.eduExotic hybrid cordgrass, Spartina alterniflora x foliosa, is altering <strong>the</strong> vegetative structure andcomposition <strong>of</strong> <strong>the</strong> San Francisco Bay tidal marsh ecosystem, which has multiple impacts on <strong>the</strong>Alameda song sparrow (Melospiza melodia pusillula), a resident passerine species that is aCalifornia Species <strong>of</strong> Special Concern. These sparrows are affected not only by <strong>the</strong> altered habitat,but also by <strong>the</strong> occupation <strong>of</strong> this habitat by a potential competitor, <strong>the</strong> marsh wren (Cistothoruspalustris). To assess <strong>the</strong> impact <strong>of</strong> <strong>the</strong> S. alterniflora invasion on song sparrow and marsh wrenpopulations we: 1) located song sparrow nests to observe nesting habitat preferences and nestsuccess and 2) used focal observations <strong>of</strong> color-banded birds to assess <strong>the</strong> vegetation composition <strong>of</strong>each territory and <strong>the</strong> amount <strong>of</strong> territory overlap between <strong>the</strong> two species. Our findings suggest that<strong>the</strong> changes in salt marsh habitat associated with <strong>the</strong> invasion <strong>of</strong> cordgrass hybrids may favor marshwrens over song sparrows and could eventually result in a decrease in salt marsh song sparrowpopulations.Keywords: Spartina alterniflora, song sparrow, marsh wren, nest success, territory, competition,San Francisco BayINTRODUCTIONNative San Francisco Bay tidal salt marshes arecharacterized by broad expanses <strong>of</strong> open tidal mudflats, anarrow mid-marsh zone where Spartina foliosa occurs, and ahigh-marsh zone composed mainly <strong>of</strong> low-growingSarcocornia spp. (formerly Salicornia spp.) with narrowareas <strong>of</strong> Grindelia that line <strong>the</strong> meandering tidal channels.Spartina alterniflora, a cordgrass native to <strong>the</strong> Atlantic andGulf coasts <strong>of</strong> North America, was introduced to SanFrancisco Bay in <strong>the</strong> early 1970s (Ayres et al. 2004). Theexotic cordgrass subsequently hybridized with <strong>the</strong> nativecordgrass, S. foliosa, and <strong>the</strong>se hybrids have spread and nowcover more than 600 hectares (ha) <strong>of</strong> tidal flat and tidalmarsh habitat (Zaremba et al., this volume).The tall, dense exotic Spartina (S. alterniflora and/or<strong>the</strong> hybrid S. alterniflora x foliosa) can grow fur<strong>the</strong>r down<strong>the</strong> tidal gradient than any native tidal marsh plant speciesand so is able to colonize open tidal flats. This exoticcordgrass can also grow much fur<strong>the</strong>r up <strong>the</strong> tidal gradientthan native S. foliosa and thus displaces o<strong>the</strong>r native plantspecies in <strong>the</strong> mid- to high-marsh zones as well (Ayres et al.1999; Nordby pers. obs.).The pr<strong>of</strong>ound changes in habitat structure andcomposition that accompany <strong>the</strong> exotic Spartina invasion(Callaway and Josselyn 1992) will likely have <strong>the</strong> greatestimpact on species, such as birds, that are wholly dependenton <strong>the</strong> tidal salt marsh system. The Alameda song sparrowFig. 1. Alameda song sparrow (Melospiza melodia pusillula). Photo byJen McBroom.- 197 -

Chapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartina<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaCriterion) revealed that overall nesting success wasestimated to be 30% lower in areas <strong>of</strong> exotic Spartina than inareas <strong>of</strong> native vegetation. We also found strong evidence <strong>of</strong>marsh wren destruction <strong>of</strong> sparrrow eggs, particularly inareas <strong>of</strong> high marsh wren density (Nordby et al. 2009).Fig. 2. Marsh wren (Cistothorus palustris). Photo by Jen McBroom.(Melospiza melodia pusillula) (Fig. 1), a California Species<strong>of</strong> Special Concern, resides entirely within salt marshes inSouth San Francisco Bay. In a native marsh, this sparrow is<strong>the</strong> main resident passerine species and occupies territoriesand nests in Sarcocornia and Grindelia near tidal channels.Ano<strong>the</strong>r passerine species, <strong>the</strong> marsh wren (Cistothoruspalustris) (Fig. 2), that normally occurs in fresh or brackishwater marshes on <strong>the</strong> Pacific Coast, has started to occupy <strong>the</strong>newly available exotic Spartina habitat (Nordby and Cohen,pers. obs.). Marsh wrens are very aggressive and will defend<strong>the</strong>ir territories against o<strong>the</strong>r birds, even o<strong>the</strong>r species, bybreaking <strong>the</strong> eggs in nests that are close to <strong>the</strong>ir ownterritories (Picman 1977).To assess <strong>the</strong> impact <strong>of</strong> <strong>the</strong> exotic Spartina invasion onsong sparrow and marsh wren populations in San FranciscoBay we 1) studied sparrow nesting habitat preferences andnest success and also looked for evidence <strong>of</strong> <strong>the</strong> destruction<strong>of</strong> song sparrow eggs by marsh wrens, and 2) assessed <strong>the</strong>vegetation composition <strong>of</strong> sparrow and wren territories aswell as <strong>the</strong> amount <strong>of</strong> territory overlap between <strong>the</strong> twospecies.SONG SPARROW NEST SUCCESSDuring <strong>the</strong> 2002 and 2003 breeding seasons wefollowed and observed <strong>the</strong> fates <strong>of</strong> 351 nests in 45+territories across three study sites (Newark, San Leandro andAlameda). Once <strong>the</strong> fate <strong>of</strong> a nest had been determined(failed due to predation, failed due to tidal flooding, failedfor o<strong>the</strong>r reasons, or successful), we recorded <strong>the</strong> locationand vegetation composition <strong>of</strong> <strong>the</strong> nest site.We found that song sparrows did use exotic Spartina asnesting habitat, but <strong>the</strong>se nests were much more likely to faildue to tidal flooding than nests placed in native vegetation.As a result, our model <strong>of</strong> daily nest survival (a generalizedlinear modeling approach using Akaike’s InformationSONG SPARROW AND MARSH WREN TERRITORIESDuring <strong>the</strong> 2003 breeding season (from March toAugust), we conducted two to four focal observations oneach color-banded male to map <strong>the</strong> territory boundaries for<strong>the</strong> 32 song sparrows and 16 marsh wrens in <strong>the</strong> twoSpartina-invaded study sites (San Leandro and Alameda).Observers used binoculars, a compass, and a range-finder tomap bird locations during each one-hour observation period.Observation locations were marked using a GPS unit, birdlocations were <strong>the</strong>n calculated and <strong>the</strong>se data were added toan ArcView GIS database. The points were used to construct100% minimum convex polygons <strong>of</strong> each territory. Usingcolor-infrared aerial photographs, we identified areas <strong>of</strong>invasive Spartina throughout <strong>the</strong> sites. In ArcView, wecombined <strong>the</strong> layer <strong>of</strong> Spartina vegetation with <strong>the</strong> layers <strong>of</strong>marsh wren and song sparrow territory polygons. We <strong>the</strong>ndetermined <strong>the</strong> percentage <strong>of</strong> each territory that wascomposed <strong>of</strong> Spartina habitat, as well as <strong>the</strong> amount <strong>of</strong>overlap between <strong>the</strong> two species.Although song sparrows did include some exoticSpartina habitat in <strong>the</strong>ir territories, all but one song sparrowterritory included some areas <strong>of</strong> native salt marsh habitat.The one territory that was determined from aerialphotographs to be entirely covered by Spartina, actually hada large portion <strong>of</strong> native pickleweed vegetation underlying<strong>the</strong> Spartina stand. In contrast, marsh wren territories weremore highly correlated with exotic Spartina habitat andmany territories were exclusively composed <strong>of</strong> <strong>the</strong> exoticcordgrass. We also found that <strong>the</strong>re was little overlapbetween <strong>the</strong> territories <strong>of</strong> <strong>the</strong> two species (Nordby et al., inprep).DISCUSSIONThese results suggest that <strong>the</strong> changes in salt marshhabitat associated with <strong>the</strong> invasion <strong>of</strong> exotic Spartina mayfavor marsh wrens over song sparrows. While song sparrowsare occupying and nesting in <strong>the</strong> exotic Spartina, those thatdo so may be at a disadvantage. It is possible that songsparrows are being drawn to nesting sites in exotic Spartinathat are inappropriate because <strong>the</strong>y are too low in elevationrelative to <strong>the</strong> tides. This increase in nest failure due t<strong>of</strong>looding coupled with an apparent increase in interferencecompetition from marsh wrens may serve to negativelyimpact salt marsh song sparrow populations in SanFrancisco Bay.However, we do not yet know whe<strong>the</strong>r exotic Spartinais acting as an ‘ecological trap’ for song sparrows, where- 198 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 3: Ecosystem Effects <strong>of</strong> Invasive Spartinaoverall reproductive success is reduced. It is also not yetknown whe<strong>the</strong>r marsh wrens are excluding song sparrowsfrom <strong>the</strong> Spartina habitat or if song sparrows are selectingagainst those areas for o<strong>the</strong>r reasons (e.g., nesting habitat orfood resources are limited).ACKNOWLEDGMENTSWe thank Letitia Grenier, Jules Evens, Joy Albertson,Katy Zaremba and Hildie Spautz for assistance withinitiating this research. We also thank Lisa Eigner, BeckyDucore and April Robinson for assistance in <strong>the</strong> field. TheCalifornia Coastal Conservancy’s Invasive Spartina Projectand Pablo Rosso graciously provided <strong>the</strong> aerial photo GISfiles. We thank U.S. Fish and Wildlife Service, CADepartment <strong>of</strong> Fish and Game, East Bay Regional ParkDistrict and <strong>the</strong> City <strong>of</strong> San Leandro for research permitsand access to marshes. Funding was provided by a David H.Smith Conservation Research Fellowship from The NatureConservancy, <strong>the</strong> National Science FoundationBiocomplexity grant DEB 0083583, and <strong>the</strong> San FranciscoEstuary Institute.REFERENCESAyres, D.R., D.L. Smith, K. Zaremba, S. Klohr, and D.R. Strong.2004. Spread <strong>of</strong> exotic cordgrasses and hybrids (Spartina sp.) in<strong>the</strong> tidal marshes <strong>of</strong> San Francisco Bay, CA, USA. Biological Invasions6: 221-231.Ayres, D.R., D. Garcia-Rossi, H.G. Davis, and D.R. Strong. 1999.Extent and degree <strong>of</strong> hybridization between exotic (Spartina alterniflora)and native (S. foliosa) cordgrass (Poaceae) in California,USA determined by random amplified polymorphic DNA(RAPDs). Molecular Ecology 8: 1179-1186.Callaway, J.C., and M.N. Josselyn. 1992. The introduction andspread <strong>of</strong> smooth cordgrass (Spartina alterniflora) in South San-Francisco Bay. Estuaries 15: 218-226.Nordby, J.C., A.C. Cohen and S.R. Beissinger. 2009. Effects <strong>of</strong> ahabitat-altering invader on nesting sparrows: An ecological trap?Biological Invasions 11:565-575.Picman, J. 1977. Destruction <strong>of</strong> eggs by <strong>the</strong> long-billed marshwren. Canadian Journal <strong>of</strong> Zoology 55: 1914-1920.Zaremba, K., M. McGowan, and D.R. Ayres. 2010. Spread <strong>of</strong> invasiveSpartina in <strong>the</strong> San Francisco estuary. In: Ayres, D.R.,D.W. Kerr, S.D. Ericson and P.R. Ol<strong>of</strong>son, eds. <strong>Proceedings</strong> <strong>of</strong><strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartina, 2004Nov 8-10, San Francisco, CA, USA. San Francisco Estuary InvasiveSpartina Project <strong>of</strong> <strong>the</strong> California State Coastal Conservancy:Oakland, CA. (this volume).- 199 -

CHAPTER FOURSpartina Control and Management

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementTAKING ADVANTAGE OF SPARTINA’S SPATIAL PATTERN FOR EFFICIENT CONTROLF.S. GREVSTADOlympic Natural Resources Center, University <strong>of</strong> Washington, 2907 Pioneer Road, Long Beach, WA 98631;grevstad@u.washington.eduThe invasion <strong>of</strong> open mudflats by Spartina alterniflora takes on a distinctive spatial pattern. Thispattern <strong>of</strong> spread <strong>of</strong>fers opportunity for strategic placement <strong>of</strong> control efforts. Spartina seedlingsestablish on open mud and <strong>the</strong>n spread vegetatively to form expanding circular patches, which dot<strong>the</strong> mudflats and eventually coalesce into a contiguous monospecific meadow. The invasiontypically begins in <strong>the</strong> upper tide zone and <strong>the</strong>n moves down <strong>the</strong> tidal gradient. Using a spatiallyexplicit model, I simulated <strong>the</strong> spread <strong>of</strong> S. alterniflora and compared various strategies for controlin a situation where only a fraction <strong>of</strong> <strong>the</strong> total infestation could be controlled each year. A strategy<strong>of</strong> killing outlying patches first and <strong>the</strong>n attacking <strong>the</strong> dense meadows (moving up <strong>the</strong> tidal gradient)led to eradication in up to 44% less time and effort than a strategy <strong>of</strong> targeting <strong>the</strong> dense meadowsfirst and outlying patches second (moving down <strong>the</strong> tidal gradient). In <strong>the</strong> control <strong>of</strong> contiguousmeadows located adjacent to <strong>the</strong> shoreline, <strong>the</strong> best strategy was to approach one end <strong>of</strong> <strong>the</strong>infestation, moving across <strong>the</strong> meadow to <strong>the</strong> o<strong>the</strong>r end. Suppression <strong>of</strong> seeds was not an effectivecontrol strategy by itself. In general, effective control strategies were those that first eliminate <strong>the</strong>plant in areas where current or future vegetative growth is greatest. Field application <strong>of</strong> <strong>the</strong>se resultsfor S. alterniflora and similar invasive plants could greatly reduce <strong>the</strong> costs <strong>of</strong> control work andimprove <strong>the</strong> likelihood <strong>of</strong> local or complete eradication.Keywords: Spartina alterniflora, Willapa Bay, spatial pattern, control strategy, control efficiencyINTRODUCTIONWhen resources for control work are limited, <strong>of</strong>ten onlya fraction <strong>of</strong> a weed invasion can be controlled in any givenyear. Under <strong>the</strong>se conditions, <strong>the</strong> spatial pattern <strong>of</strong> a weedinvasion can provide an opportunity for strategic placement<strong>of</strong> control efforts to achieve <strong>the</strong>ir greatest effect.The invasion <strong>of</strong> mudflats by Spartina spp. (cordgrasses)takes on a characteristic pattern. Seedlings establish in openmud and <strong>the</strong>n spread vegetatively to form expanding circularpatches. These initially dot <strong>the</strong> mudflats and can eventuallycoalesce into a contiguous meadow. The invasion typicallybegins in <strong>the</strong> upper intertidal zone and <strong>the</strong>n moves down <strong>the</strong>tidal gradient. A distinct boundary is formed by <strong>the</strong> highernative marsh, which Spartina rarely invades. This study usesa simulation to compare control approaches for this pattern<strong>of</strong> spread. A more detailed version <strong>of</strong> it appears Grevstead2005.METHODSA stochastic grid-based model was developed usingMatlab® s<strong>of</strong>tware to simulate <strong>the</strong> spread <strong>of</strong> Spartinaalterniflora on a mudflat and to compare different controlstrategies. The simulations used a grid dimension <strong>of</strong> 120 by120 cells where each cell was 1 square meter (m 2 ). One axis<strong>of</strong> <strong>the</strong> model space follows a tidal elevation gradient, with<strong>the</strong> upper edge representing <strong>the</strong> native marsh boundary and<strong>the</strong> bottom edge representing <strong>the</strong> lower extent <strong>of</strong> S.alterniflora growth. Cells are considered ei<strong>the</strong>r empty oroccupied by S. alterniflora. They become occupied throughvegetative spread from neighboring cells or by establishment<strong>of</strong> seedlings dispersed from an occupied cell. Parameterestimates were obtained from field data for S. alterniflora inWillapa Bay. Assumptions and parameters <strong>of</strong> <strong>the</strong> modelwere as follows:• Vegetative spread = 0.77 m radial increase per year(measured from aerial photos)• Seeds disperse from occupied sites according to aGaussian distribution ( = 50)• Seedling recruitment declines linearly with tidalelevation (based on Feist 1999)• Seedling recruitment produces a 17% increase in areaper year (based on Murphy 2003)• Once established, clonal patches do not die until treatedOutlying patches first vs. meadows firstThe time and total effort needed to eradicate apopulation was compared for a meadows-first vs. anoutliers-first treatment strategy. In each case a fixed amount<strong>of</strong> Spartina was removed each year starting with a 20-yearoldinvasion (Fig.1). Paired trials were replicated 10 timesfor each <strong>of</strong> three levels <strong>of</strong> yearly effort.Approach direction for meadowsFor a case where <strong>the</strong>re are no outliers but only anoblong meadow adjacent to <strong>the</strong> native marsh boundary, threestrategies were compared: (1) Approaching from <strong>the</strong> lowertide zone (mudflat) and moving toward <strong>the</strong> upper tide zone- 203 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaA. Meadow first B. Outlying clones firstTime to eradication (years)403530252015105A.Yearly effort5001000150001 2Meadow firstOutliers firstFig. 1 Meadow-first (left) and outlier-first (right) approaches to controllingSpartina. Shown is <strong>the</strong> remaining Spartina at times 22, 25 and 30years. Removal <strong>of</strong> 500 occupied squares each year began on year 20.Fig. 1. Meadow-first (left) and outlier-first (right) approaches to controllingSpartina. Shown is <strong>the</strong> remaining Spartina (black areas) at times 2, 5,and 10 years. Removal <strong>of</strong> 500 occupied squares each year began on year20.(native marsh), (2) approaching from <strong>the</strong> upper tide zone andmoving toward <strong>the</strong> lower tide zone, and (3) approachingfrom one end <strong>of</strong> <strong>the</strong> meadow and moving parallel to shore to<strong>the</strong> o<strong>the</strong>r end. Each year a fixed amount <strong>of</strong> Spartina wasremoved.Is seed suppression effective?Seed suppression, through mowing or low concentrationherbicide spray, can be carried out at a much lower cost thancompletely killing <strong>the</strong> plants. To test <strong>the</strong> effectiveness <strong>of</strong>seed suppression, three start conditions were used (a 15-year-old invasion, a 20-year-old invasion, and a 15-year-oldmeadow without outliers). Seeds were suppressed at threelevels: 0%, 50%, or 100%. In each case, seeds weresuppressed every year, while established plants wereallowed to spread vegetatively.RESULTSThe outliers-first strategy achieved eradication substantiallysooner and with less total effort than <strong>the</strong> meadow firstTotal cost (m 2 controlled)20000180001600014000120001000080006000strategy. The difference was especially great when <strong>the</strong> availableyearly effort was low (Fig. 2).For oblong meadows, Spartina was eliminated fastest,and with <strong>the</strong> least total effort, when control work was appliedfirst to one end <strong>of</strong> <strong>the</strong> meadow and <strong>the</strong>n moved across<strong>the</strong> meadow in subsequent years. Slightly less effective was<strong>the</strong> approach <strong>of</strong> moving from lower to upper intertidal zone.The least effective strategy was to control from <strong>the</strong> uppertide zone, because this opened up an additional growingedge. Again, <strong>the</strong> largest difference in strategy effectivenesswas found when <strong>the</strong> yearly control effort was low.B.Yearly effort500100015001 2Meadow firstOutliers firstFig. 2. Comparison <strong>of</strong> <strong>the</strong> amount <strong>of</strong> time (A) and amount <strong>of</strong> effort (B)needed to eradicate a Spartina population using a meadow-first or outliersfirstapproach. Plots show mean and standard error <strong>of</strong> 10 replicate trials foreach strategy and yearly effort combination.- 204 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementTime to eradication (years)30252015105*A.Yearly effort200300600Area occupied1400012000100008000600040002000Seeds suppressed after year 150% seed suppression50% seed suppression100% seed suppression0Upper to lower Lower to upper End onApproach direction00 5 10 15 20 25 30 35 40 45Time (years)14000Relative cost <strong>of</strong> eradication(m 2 controlled)50004000300020001000*B.Yearly effort200300600Area occupied12000100008000600040002000Seeds suppressed after year 200% seed suppression50% seed suppression100% seed suppression00 5 10 15 20 25 30 35 40 45Time (years)0Upper to lower Lower to upper End onApproach directionFig. 3. Comparison <strong>of</strong> <strong>the</strong> amount <strong>of</strong> time (A) and amount <strong>of</strong> effort (B)needed to eradicate a Spartina population in <strong>the</strong> case <strong>of</strong> an oblong meadowadjacent to <strong>the</strong> native marsh using three approach directions. Plots showmean and standard error <strong>of</strong> 10 replicate trials for each strategy and yearlyeffort combination.When <strong>the</strong> mudflat was already dotted with patches <strong>of</strong>Spartina, seed suppression only slightly reduced <strong>the</strong> rate <strong>of</strong>invasion and it took many years to see an effect. Seed suppressionwas somewhat more effective in <strong>the</strong> meadow-onlycase. However, in all cases, <strong>the</strong> area occupied by Spartinacontinued to increase.CONCLUSIONSIn a situation where only a fraction <strong>of</strong> <strong>the</strong> invasion canbe controlled each year, strategic placement <strong>of</strong> controltreatments can greatly reduce <strong>the</strong> time and effort needed toeradicate Spartina. Control is more effective if it first targetsoutlying clones in <strong>the</strong> lower tide zone and <strong>the</strong>n moves up <strong>the</strong>tidal gradient to target <strong>the</strong> larger meadow. These results areconsistent with expectations based on <strong>the</strong> more generalizedstudy by Moody and Mack (1988); <strong>the</strong>y also agree with aArea occupied12000100008000600040002000Meadow only infestationSeeds suppressed after year 150% seed suppression50% seed suppression100% seed suppression00 5 10 15 20 25 30 35 40 45Time (years)Fig. 4. Growth <strong>of</strong> simulated S. alterniflora populations for which seedswere suppressed at levels <strong>of</strong> 0%, 50% and 100% every year for a 20 yearperiod. Seed suppression treatment was applied to <strong>the</strong> invading populationafter year 15 in (A) and after year 20 in (B). In (C) seed suppression wasapplied to a 15-year-old meadow surrounded by an uninfested mudflat.model <strong>of</strong> Spartina control by Taylor et al. (this volume). Theoutliers-first approach eliminates Spartina in <strong>the</strong> areas wherevegetative spread is greatest. In addition, it avoids openingup a new growing edge in <strong>the</strong> upper intertidal zone and also- 205 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaavoids opening up <strong>the</strong> habitat that is most favorable forseedlings.For <strong>the</strong> control <strong>of</strong> solid meadows without outlyingpatches (i.e. after outliers have been controlled), treatmentshould begin with ei<strong>the</strong>r <strong>the</strong> lower edge or, slightly better,one end <strong>of</strong> <strong>the</strong> meadow. Again, this more rapidly cuts down<strong>the</strong> amount <strong>of</strong> <strong>the</strong> growing edge and avoids opening a newedge in <strong>the</strong> upper zone.An assumption behind <strong>the</strong>se results is that <strong>the</strong> cost <strong>of</strong>control per unit area is similar in different parts <strong>of</strong> <strong>the</strong> invasion.This may or may not be true depending on what toolsare available to <strong>the</strong> control crew. It is anticipated that landmanagers will consider <strong>the</strong> results <strong>of</strong> this study in combinationwith <strong>the</strong>ir own costs <strong>of</strong> control for different sections <strong>of</strong><strong>the</strong> S. alterniflora invasion.Seed suppression alone is not likely to be effective, especiallywhen <strong>the</strong> invasion is already widespread. This isbecause <strong>the</strong> established plants will continue to spread vegetatively.Seed suppression may be warranted if used in combinationwith control <strong>of</strong> vegetative spread and when it can bedone completely and economically.ACKNOWLEDGEMENTSThis work was inspired by discussions with state andfederal agency Spartina control coordinators in Willapa Bay,including Todd Brownlee, Blaine Reeves, Kyle Murphy,Charlie Stenvall, Brady Scott, Wendy Brown, Lester Holcum,Dave Heimer, and Kim Patten. Photographs used inparameterizing <strong>the</strong> model were made available by WendyBrown and Janie Civille <strong>of</strong> <strong>the</strong> Washington Department <strong>of</strong>Natural Resources. Financial support was provided by <strong>the</strong>National Sea Grant Program and <strong>the</strong> U.S. Department <strong>of</strong>Fish and Wildlife.REFERENCESFeist, B.E. 1999. Spatio-temporal analysis <strong>of</strong> <strong>the</strong> environmentaland climatic factors controlling <strong>the</strong> expansion <strong>of</strong> Spartina alterniflorain Willapa Bay, Washington. PhD <strong>the</strong>sis. University <strong>of</strong>Washington.Grevstad, F.S. 2005. Strategies for controlling a spatially structuredplant invasion: Spartina alterniflora in Pacific Coast estuaries.Biological Invasions 7: 665-667.Moody, M.E. and R.N. Mack. 1988. Controlling <strong>the</strong> spread <strong>of</strong> plantinvasions: <strong>the</strong> importance <strong>of</strong> nascent foci. J Appl Ecol 25:1009-1021.Murphy K.C. 2003. Report to <strong>the</strong> Legislature: Progress <strong>of</strong> <strong>the</strong>Spartina Eradication and Control Programs. Olympia, WA:Washington State Department <strong>of</strong> Agriculture.Taylor, C.M., A. Hastings, H.G. Davis, J.C. Civille, and F.S.Grevstad. Modeling <strong>the</strong> spread and control <strong>of</strong> Spartina alterniflorain Willapa Bay. In: Ayres, D.R., D.W. Kerr, S.D. Ericsonand P.R. Ol<strong>of</strong>son, eds. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong><strong>Conference</strong> on Invasive Spartina, 2004 Nov 8-10, SanFrancisco, CA, USA. San Francisco Estuary Invasive SpartinaProject <strong>of</strong> <strong>the</strong> California State Coastal Conservancy: Oakland,CA. (this volume).- 206 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementSHOREBIRD USE OF SPARTINA-AFFECTED TIDELANDS –CAN WE ACHIEVE FUNCTIONALHABITAT POST-CONTROL?K. PATTEN AND C. O’CASEYWashington State University, Long Beach Research and Extension Unit.2907 Pioneer Road, Long Beach, WA 98631; pattenk@wsu.edu; ocasey@wsu.eduOne <strong>of</strong> <strong>the</strong> major threats <strong>of</strong> invasive Spartina is <strong>the</strong> loss <strong>of</strong> shorebird foraging habitat. The AudubonSociety lists <strong>the</strong> invasion <strong>of</strong> Willapa Bay (WB) by Spartina as <strong>the</strong> second most critical threat toshorebird habitat in <strong>the</strong> nation. Studies were conducted on how large-scale mechanical and chemicalcontrol efforts affect shorebird and waterfowl usage <strong>of</strong> Spartina meadows in WB. Food abundanceand accessibility, shorebird, waterfowl and bird <strong>of</strong> prey density, and bird behaviour were evaluatedon treated meadows and compared to untreated meadows and bare mudflats. Based on long-termpoint counts and remote video monitoring, <strong>the</strong>re was no bird usage (<strong>of</strong> any species) in Spartinameadows. Highest food abundance and accessibility was found in mudflats. Waterfowl and birds <strong>of</strong>prey preferred herbicide-treated sites over <strong>the</strong> tilled and mudflat sites. Shorebirds preferred mudflatsfollowed by tilling over that <strong>of</strong> herbicide-treated sites. Bird behaviour (feeding or resting) wasvariable and dependent on species, time <strong>of</strong> year and treatment. Although tilling appears to beinitially effective in expediting restoration for shorebirds, it is too costly to implement on a largescale. The most significant long-term concern for shorebird usage is <strong>the</strong> Spartina-induced increasein tidal elevations on <strong>the</strong>se meadows (>35 cm). Less than 20% <strong>of</strong> <strong>the</strong> gain in elevation wasattributable to sediment accretion; <strong>the</strong> rest was root biomass. Due to <strong>the</strong> change in bathymetry, onceSpartina was controlled at <strong>the</strong>se sites, native salt marsh plants (Salicornia, Triglochin and Spergula)immediately invaded more than 400 meters out into what were previously intertidal mudflats. Thispotentially permanent large scale conversion <strong>of</strong> mudflat to salt marsh will have pr<strong>of</strong>oundimplications for shorebird habitat. Some potential remedies will be suggested.Keywords: restoration, Dunlin, Western Sandpiper, birds <strong>of</strong> prey, native salt marsh, Willapa Bay,SpartinaINTRODUCTIONSpartina has colonized and eliminated much <strong>of</strong> <strong>the</strong>upper portion <strong>of</strong> <strong>the</strong> wide expansive intertidal mudflats <strong>of</strong>Willapa Bay. Species most threatened by Spartina are likelyto be <strong>the</strong> thirty species <strong>of</strong> shorebirds that rely upon WillapaBay’s 47,000 acres <strong>of</strong> tideland for food and shelter duringannual migrations to and from <strong>the</strong> Arctic. Much <strong>of</strong> <strong>the</strong> mostpreferredshorebird habitat <strong>of</strong> Willapa Bay, sheltered uppertidal mudflats in <strong>the</strong> sou<strong>the</strong>rn bay, has been displaced bySpartina. Peak winter and spring shorebird usage in sections<strong>of</strong> <strong>the</strong> bay has declined over 60 percent in <strong>the</strong> past decade asSpartina meadows have replaced <strong>the</strong> tidal mudflats (Jaques2002). Census studies on shorebird abundance in WillapaBay in 1991-1995, prior to <strong>the</strong> major increase in Spartinagrowth, found that 44 percent <strong>of</strong> <strong>the</strong> total bird usage waswithin two areas, <strong>the</strong> Bear River/Lewis Unit – South WillapaBay region and <strong>the</strong> Willapa River area (Buchanan andEvenson 1997). These two areas have become almostcontiguous Spartina meadows. Because <strong>of</strong> <strong>the</strong> loss <strong>of</strong> habitatcaused by Spartina, <strong>the</strong> Audubon Society has listed WillapaBay as <strong>the</strong> second most endangered shorebird habitat in <strong>the</strong>United States (Audubon 2004).The ongoing chemical and mechanical control effort is<strong>the</strong> first step in recovering that habitat. The ultimate goal <strong>of</strong>a control effort should not be limited to control, but alsoneeds to consider restoration <strong>of</strong> <strong>the</strong> affected habitat formaximal ecological value. Little information exists to dateon how <strong>the</strong> numerous chemical and mechanical controlmethods being used to manage Spartina have expeditedhabitat restoration. The long-term ecological impact <strong>of</strong>invasive Spartina on shorebirds in England has beenrecently reviewed by Lacambra et al. 2004. They concludethat a return <strong>of</strong> shorebirds to English estuaries followingSpartina removal is not axiomatic. In Washington, where<strong>the</strong>re have been long-term control efforts by variousagencies, use <strong>of</strong> <strong>the</strong> affected tideflats by shorebirds andwaterfowl increases dramatically within several years <strong>of</strong>removal <strong>of</strong> invasive Spartina from mudflats (Patten andO’Casey 2007). The objective <strong>of</strong> this study was to assess <strong>the</strong>likelihood and <strong>the</strong> limiting factors involved in achievingfunctional habitat at Spartina-affected mudflats after control.METHODSSite information:Direct and indirect assessments were made <strong>of</strong> shorebird,waterfowl and birds <strong>of</strong> prey usage <strong>of</strong> Spartina meadows(treated and untreated) in comparison to bare mudflats.These assessments were made for five sites: bare mudflat,- 207 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaShorebirds/m 2 /hr86420Native Tilled Sprayedmudflat meadow /mowedmeadowSprayedmeadowUntreatedSpartinameadowFig. 1. A summary <strong>of</strong> <strong>the</strong> comparative use <strong>of</strong> Spartina affected tideflats byshorebirds during <strong>the</strong> winter/spring migration in 2003 based on foragingflux density data from remote sensing cameras as a funciton <strong>of</strong> Spartinacontrol method. Bars = Std. Err.Bird number / ha / 15 minutes1500 Western SandPiperDunlin1000Untreated Spartina50001500100050001500100050001500Sprayed SpartinaTilled Spartinatilled Spartina meadow, sprayed Spartina meadow, spraymowedSpartina meadow and an untreated Spartinameadow. Data collected included beak probe density,footprint density, fecal dropping density, visual countsduring peak migration in spring 2003 and winter <strong>of</strong>2003/2004, and remote monitoring with video cameras inwinter/spring 2003. The study site was on Willapa NationalWildlife Refuge property at <strong>the</strong> south end <strong>of</strong> Willapa Bay.The Spartina infestation was 10 to 14 years old and coveredover 1000 hectares (ha). Treatment sites were adjacent toeach o<strong>the</strong>r and large enough to be considered ecologicallysignificant units (>60 ha). This part <strong>of</strong> <strong>the</strong> bay supported anabundant bird population prior to infestation by Spartina(Jacques 2002). Although <strong>the</strong> sites had similar bathymetryprior to Spartina infestation, <strong>the</strong>ir current elevations weremeasured to be greater than 35 centimeters (cm) above <strong>the</strong>adjacent mudflats. The bare mudflat site is and has beenSpartina-free. The tilled site has been treated since2000/2001 with mowing, tilling, and spraying for cleanup. Ithas been relatively free <strong>of</strong> Spartina since 2002. The sprayedand spray-mowed sites were treated with 2 gallons/acre <strong>of</strong>glyphosate (Rodeo®) in summer <strong>of</strong> 2002 and had follow-upspraying in summer <strong>of</strong> 2003. The spray-mowed site wasmowed to a level <strong>of</strong> approximately 14 cm during <strong>the</strong> spring<strong>of</strong> 2003 to remove dead stubble and encourage bird usage.The untreated Spartina area is a large meadow comprisingmore than 200 ha at <strong>the</strong> southwest end <strong>of</strong> <strong>the</strong> bay.Shorebird dataBeak probe density, footprint density, and fecaldropping density (#/0.25 m 2 ) data were collected on May 13,2003, using five replications per habitat per location, withfive subsample counts per replication. For each replication,comparative habitats (treatments) were located within 20feet <strong>of</strong> each o<strong>the</strong>r. Remote monitoring <strong>of</strong> sites was doneusing video cameras in winter/spring 2003. A Mitsubishi1000500Bare mudflat011/22/03 12/12/03 1/1/04 1/20/04 2/9/04Date <strong>of</strong> observationFig. 2. Visual counts <strong>of</strong> <strong>the</strong> major shorebird species in Willapa Bay inWinter 03/04 as a funciton <strong>of</strong> <strong>the</strong> Spartina control method.Time Lapse Security Recorder, Model #HS-1280U, wasused to record <strong>the</strong> black and white image from a SuperCircuits PC23C camera with a 12-mm 1/3” CS TV lens.Power was provided using three 12-volt deep cycle marinebatteries and a 16-watt solar pane with a DC to AC, 12-volt,150-watt inverter. Cameras were mounted in wea<strong>the</strong>rpro<strong>of</strong>camera housing on 7-m poles 133 m from <strong>the</strong> native marsh.The camera focal area for each site varied slightly, rangingfrom approximately 90 to 180 m 2 . Total bird usage(shorebird and waterfowl) from each tape was recordedevery 30 seconds and <strong>the</strong> data was converted to mean dailyflux densities (#/m 2 /hour). For shorebirds, daily fluxdensities were based only on time periods during <strong>the</strong> daywhen <strong>the</strong> tideflats were exposed. The total number <strong>of</strong> days <strong>of</strong>complete data collection, from February 18, 2003 to May 14,2003, ranged from 20 to 40 depending on <strong>the</strong> site. Visualobservation <strong>of</strong> bird usage in <strong>the</strong> winter <strong>of</strong> 2003/04 was doneusing a single observer. Three plots (one hectare each) persite were observed for 10 minute intervals using a spottingscope. Observations were timed to coincide with peak usageat each site, just prior to tidal submergence or after tidalwithdraw. Observation frequency was at least once a week.Bird species and behavior were noted.Soil and plant dataIntact cores to <strong>the</strong> bottom <strong>of</strong> <strong>the</strong> root system (80+ cm)were collected by digging a 1-m wide and deep trench.- 208 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementDistance from native marsh (m)10008006004002000Salicornia virginicaTriglochin maritimumCotula coronopifoliaSpergularia canadensisTilled 2001/2002Sprayed 2002/2003Fig. 3. The mean maximum distance from <strong>the</strong> shoreline that salt marshspecies were located after Spartina was controlled. Data were collected June2004.Standard soil science methodology was used to determineporosity, bulk density, and <strong>the</strong> core sample composition. Bywashing <strong>the</strong> trench wall it was possible to identify allgrowing point meristems and record <strong>the</strong>ir points <strong>of</strong> origin(depth). The change in depth from <strong>the</strong> first occurringmeristem to <strong>the</strong> current growing points over <strong>the</strong> 8-to-10-yearperiod this meadow had been growing was assumed to be achange in tidal elevation resulting from Spartina-inducedaccretion. Data on vascular plant density (#/m 2 ) by specieswere collected in June 2004 from multiple transects from <strong>the</strong>native marsh line out to 500 m through each treatment site.RESULTSShorebird foragingBased on visual and remote observation data during <strong>the</strong>time course <strong>of</strong> this study, none <strong>of</strong> <strong>the</strong> Spartina controlmethods resulted in shorebird usage comparable to <strong>the</strong> baremudflats (Figures 1 & 2). Flux density <strong>of</strong> shorebirds duringwinter and spring <strong>of</strong> 2003 was higher in <strong>the</strong> bare mudflatthan in <strong>the</strong> tilled areas based on remote sensing (Figure 1),but comparable between <strong>the</strong> sites based on visualobservations (Figure 2). Flux densities <strong>of</strong> shorebirds on <strong>the</strong>tilled site were higher than <strong>the</strong> sprayed or spray-mowed site(Figures 1 & 2). During 480 hours <strong>of</strong> video recordings, noshorebirds were ever observed at <strong>the</strong> Spartina meadow site.Shorebirds (Western sandpipers) were only noted once over14 visual observations in <strong>the</strong> Spartina meadow, and <strong>the</strong>yfloated in on a Spartina wrack. From a behavioralperspective, it appeared that <strong>the</strong> tilled site had <strong>the</strong> lowestpercentage <strong>of</strong> skittish feeding and <strong>the</strong> bare mudflat <strong>the</strong>lowest percentage <strong>of</strong> resting (data not shown).Based on short-term comparisons in shorebirdfootprints, fecal droppings and beak probe densities, <strong>the</strong>rewere major differences in shorebird microsite habitatpreferences (Table 1). All types <strong>of</strong> dead Spartina stubble orlive Spartina stems drastically interfered with shorebirdforaging. For three locations within Porter Point, <strong>the</strong>re wasalmost no evidence <strong>of</strong> any shorebird usage where <strong>the</strong>re wasTable 1. Shorebird usage <strong>of</strong> Spartina habitat at Porter Point based onfootprint, fecal dropping and beak print densities as a function <strong>of</strong>accessibility.*General sitelocationSprayedmeadow, 80 mfrom nativemarsh; stubble25-50cm highEdge <strong>of</strong> sprayedmeadowadjacent to baremudflat, 80 mfrom nativemarsh; stubble25 to 50 cm highTilled strips inbetween sprayedSpartina, 1000’from nativemarsh; stubblewas mowed at25 cm heightShorebirdaccessdescriptions Footprints*Live SpartinacanopyDensity (#/0.25m 2 )**FecaldroppingsBeakprints0 0 0.8±0.4Dead stubble 50±50 0.9±0.4 29±4ExposedmudflatLive Spartinacanopy134±11 1.8±0.3 68±100.2±0.1 0 3±2Dead stubble 8±2 6±0.3 13±4Exposedmudflat35±5 0.8±0.2 47±5Dead stubble 87±12 0.5±0.2 25±4Exposedmudflat130±19 1.3±0.3 76±9*Data collected 5-13-03; 5 replications per habitat per location, with 5subsample counts per replication. For each replication comparative habitats(treatments) were located with 20 feet <strong>of</strong> each o<strong>the</strong>r.** mean ± standard errorlive Spartina growing. The bare mud and dead stubblelocations usually displayed high counts <strong>of</strong> beak and footprints and fecal droppings. Bare mud usually had twice asmuch shorebird usage as dead stubble.Soil and plant dataSpartina meadows rapidly began a transition to nativemiddle to upper salt marsh as soon as <strong>the</strong> Spartina was killed(Figure 3). Within two years <strong>of</strong> treatment, four salt marshplant species extended 400 m out from <strong>the</strong>ir native marshhabitat. At this particular site, <strong>the</strong> transition from mudflat toSpartina meadow to salt marsh has all occurred with tenyears and represents a permanent loss <strong>of</strong> hundreds <strong>of</strong>hectares <strong>of</strong> prime shorebird habitat. In an analysis <strong>of</strong> soilparameters at this meadow (data not shown), we have found<strong>the</strong> dead Spartina root mat extends down to 35 cm, with <strong>the</strong>bulk <strong>of</strong> <strong>the</strong> soil volume being comprised <strong>of</strong> organic matterand pore space. Only 15% <strong>of</strong> <strong>the</strong> elevation rise could beaccounted for by sediment accretion.DISCUSSIONRestoring mudflats back to <strong>the</strong>ir original form andfunction will be extremely difficult. Even with tilling andseveral years <strong>of</strong> follow-up chemical control, as well as- 209 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinanatural restoration processes occurring over several years,Spartina-affected mudflats are far from having shorebirdusage comparable to what normally occurred on a bare tidalmudflat. This may be especially true for low tidal energysites in <strong>the</strong> sou<strong>the</strong>rn half <strong>of</strong> Willapa Bay, where <strong>the</strong>re istraditionally high shorebird usage. At <strong>the</strong>se sites, <strong>the</strong>landscape-scale changes in bathymetry via Spartina-inducedaccretion and root mass accumulation make it unlikely thatany restoration effort would be able to bring back <strong>the</strong>original bathymetry. This is especially true since nativemarsh is already succeeding in <strong>the</strong>se areas. Once <strong>the</strong>se siteshave transitioned to stable salt marshes, <strong>the</strong>re will be littlelikelihood that <strong>the</strong>y could ever become functional mudflatsagain. To prevent irreversible loss <strong>of</strong> prime shorebirdhabitat, it is <strong>the</strong>refore absolutely essential to eradicate allexisting Spartina in <strong>the</strong>se critical sites as quickly as possible.Can we realistically achieve restoration <strong>of</strong> functionalshorebird habitat at Spartina-affected tidelands post-control?If <strong>the</strong> site has undergone major elevation changes, it is likelythat it will become a stable salt marsh, and achievingshorebird habitat over <strong>the</strong> long term will be problematic. If<strong>the</strong> site has not undergone major Spartina-induced elevationchanges, <strong>the</strong>n habitat restoration is feasible. Restoration maybe expedited with a process that breaks up root masses andremoves stubble and traces <strong>of</strong> Spartina canopy, such astilling. This process is not inexpensive. Tilling <strong>of</strong> largeSpartina meadows is cost-prohibitive, requires veryspecialized equipment and is very slow (

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementWHERE DO WE GO FROM HERE?ALTERNATIVE CONTROL AND RESTORATIONTRAJECTORIES FOR A MARINE GRASS (SPARTINA ANGLICA)INVADER IN DIFFERENTHABITAT TYPESS.D. HACKER 1 AND M.N. DETHIER 21 Department <strong>of</strong> Zoology, Oregon State University, 3029 Cordley Hall, Corvallis, Oregon 97331;hackers@science.oregonstate.edu2 Department <strong>of</strong> Biology, University <strong>of</strong> Washington, Friday Harbor Laboratories, 620 University Road, Friday Harbor,Washington 98250; mdethier@u.washington.eduLittle is known about <strong>the</strong> effects <strong>of</strong> removing invasive species and subsequent consequences forcommunity restoration. Invasive species removal can have positive effects for some communitiesbut may cause unexpected changes that lead <strong>the</strong> system to an alternative state. The consequences <strong>of</strong>invasive species removal are likely to be context-dependent with restoration occurring readily undersome situations but not o<strong>the</strong>rs. English cordgrass, Spartina anglica, has invaded large areas <strong>of</strong>protected shoreline in <strong>the</strong> Puget Sound, Washington State, USA, and is <strong>the</strong> target <strong>of</strong> intensiveremoval efforts. It invades and modifies a variety <strong>of</strong> habitat types, from unvegetated mudflats andcobble beaches to established low and high salinity native marshes. It binds sediment around itsdense root system and changes biogeochemical processes, all <strong>of</strong> which can have significantconsequences for shorebirds, infauna, and commercial aquaculture. Cordgrass invasion,modification, and removal varies among <strong>the</strong> different habitat types but post-removal colonizationpredictably results in colonization <strong>of</strong> native vascular plants. Although <strong>the</strong>se plants are <strong>the</strong> dominantspecies in salt marsh communities, <strong>the</strong>y are uncommon in mudflat and cobble beach communities,and thus do not represent a restored post-removal state. Instead, <strong>the</strong> legacy effects <strong>of</strong> cordgrassproduce alternative outcomes. We hypo<strong>the</strong>size, based on <strong>the</strong> interaction between recruitment,physical disturbance (water movement) and sediment accretion, that cobble beach and high salinitymarshes will be restored but that mudflats and low salinity marshes will retain <strong>the</strong> legacy <strong>of</strong> <strong>the</strong>invasion for <strong>the</strong> long term.Keywords: English cordgrass, Spartina anglica, habitat modification, salt marsh, mudflat, cobblebeach, restoration, alternative states, Puget Sound, Washington.INTRODUCTIONIn recent years <strong>the</strong>re has been a strong focus on <strong>the</strong>impacts <strong>of</strong> invasive species both at community andecosystem levels (Parker et al. 1999; Ruiz et al. 1999;Grosholz 2002). Some <strong>of</strong> <strong>the</strong>se invasive species, known asfoundation, dominant, or ecosystem engineering species(Jones et al. 1994; Power et al. 1996; Bruno and Bertness2001; Crooks 2002), can transform communities, resulting inboth positive and negative effects for native as well asnonindigenous species. These species can have a largeinfluence on community structure relative to <strong>the</strong>ir biomass;<strong>the</strong>y alter ecological processes in multiple ways and <strong>of</strong>tencreate positive feedbacks that benefit <strong>the</strong>ir continuedexpansion and impact.Much less is known about <strong>the</strong> consequences <strong>of</strong>removing invasive species, especially dominant orfoundation species (Hobbs and Humphries 1995; Myers etal. 2000; Zavalata et al. 2001, Hacker and Dethier 2009).Invasive species removal can have positive effects for somecommunities, with restoration occurring soon after removal(Fig. 1A; Myers et al. 2000). However, in many cases, <strong>the</strong>results have been mixed, with unexpected and widespreadimpacts on natural communities (Zavalata et al. 2001;D’Antonio and Meyerson 2002). Communities may notsimply return to <strong>the</strong>ir former state in a straightforwardreversal <strong>of</strong> <strong>the</strong> invasion process but, instead, could be somodified by <strong>the</strong> invasion that <strong>the</strong>y are not easily restored(Fig. 1B; Hobbs and Humphries 1995; D’Antonio andMeyerson 2002). These modifications are likely to vary indegree, depending on characteristics <strong>of</strong> <strong>the</strong> invader, <strong>the</strong>invaded community, and <strong>the</strong> time since invasion, but couldprevent full recovery after <strong>the</strong> invader is removed. Just how<strong>the</strong> ‘legacy’ <strong>of</strong> <strong>the</strong> invasion influences post-removalcommunity structure is poorly understood; yet suchunderstanding is critical to be able to confidently predictwhe<strong>the</strong>r management goals, originally intended to restore<strong>the</strong> integrity <strong>of</strong> highly invaded communities, will be met byremoval alone or whe<strong>the</strong>r additional measures may berequired. Given <strong>the</strong> numerous removal programs underway,development <strong>of</strong> testable <strong>the</strong>ories that predict post-removalcommunity dynamics are needed to better understand <strong>the</strong>benefits and risks <strong>of</strong> removing invaders that have largemodifying effects (Hacker and Dethier 2009).- 211 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaA. Best casescenarioRestorationB. More likelyscenarioAlternativeFig. 1. Stages <strong>of</strong> invasive species removal (arrow) under A) <strong>the</strong> best casescenario for which <strong>the</strong> invader modification is lost and habitat is restored,and B) <strong>the</strong> more likely scenario for which <strong>the</strong> modification remains andcommunity follows an alternative trajectory. From Hacker and Dethier 2009In this paper, we explore <strong>the</strong> possible consequences <strong>of</strong>removing <strong>the</strong> invasive English cordgrass, Spartina anglica,which has invaded <strong>the</strong> shoreline <strong>of</strong> Puget Sound inWashington State, USA. Spartina anglica colonizescommunities with different species assemblages andphysical conditions, and thus produces variable degrees <strong>of</strong>invader modification. Our goal is to use <strong>the</strong>se patterns <strong>of</strong>differential invasion along with a simple conceptual modelto make predictions about post-removal communitystructure. Based on this analysis, we predict that somecommunities will be readily restored while o<strong>the</strong>rs willfollow alternative states. These states will depend on how<strong>the</strong> modifications and physical disturbance interact toinfluence both species recruitment and communitymaintenance through time.STUDY SYSTEMSpartina anglica was first introduced to Puget Soundfrom England in 1961. It did not become a managementpriority until <strong>the</strong> late 1990s when it had spread to a total <strong>of</strong>3,300 hectares (ha) <strong>of</strong> intertidal habitat at 77 sites (Hacker etal. 2001). Cordgrass is a strong ecosystem modifier thataccretes sediment around its dense root system and changessediment biogeochemistry, which can have importantcommunity-wide consequences (Thompson 1991; Daehlerand Strong 1996). Species affected by <strong>the</strong> invasion includenative and commercial invertebrates (infauna and epifaunasuch as clams and oysters; Zipperer 1996; O’Connell 2002),plants (Hacker and Dethier 2006), and birds (Goss-Custardand Moser 1988; Triplet et al. 2002).Cordgrass grows in a range <strong>of</strong> communities in PugetSound from mudflats and cobble beaches, which arenormally devoid <strong>of</strong> vascular plants, to low and high salinitymarshes, where native vascular plants are <strong>the</strong> mainbiological component (Hacker et al. 2001). Our researchshows that cordgrass invasion, removal, and post-removalpatterns vary dramatically between <strong>the</strong>se communities(Hacker et al. 2001; Reeder and Hacker 2004; Dethier andHacker 2005; Hacker and Dethier 2006), thus suggestingthat its removal will result in different post-removalcommunity structure depending on habitat.Habitat dependent invasion and modification by cordgrassThe abundance and distribution <strong>of</strong> English cordgrassvaries between four habitat types within Puget Sound(Hacker et al. 2001). Low salinity marsh and mudflat siteshave much larger infestations than cobble beach and highsalinity marsh sites. These differences are driven mostly byvariability in physical conditions across <strong>the</strong> four habitattypes although biological interactions play a small role(Dethier and Hacker 2005). In a seed addition experiment(Dethier and Hacker 2005), we found that mudflat and lowsalinity marshes had <strong>the</strong> greatest cordgrass germination,survival, and growth <strong>of</strong> all <strong>the</strong> habitats. In mudflatcommunities, which are naturally devoid <strong>of</strong> vascular plants,seed germination was high and surviving seedlings grewquickly. This growth pattern eventually results in <strong>the</strong>coalescence <strong>of</strong> individual plants and widespread growthacross all <strong>the</strong> intertidal elevations. Low salinity marshes, on<strong>the</strong> o<strong>the</strong>r hand, are dominated by a diverse native vascularplant assemblage. Cordgrass had high seedling germination,survival and growth; this pattern eventually results in largeswards that outcompete native plants and form densemonocultures. In high salinity marsh habitats, cordgrassgrows mainly in sparsely vegetated low intertidal streamchannel areas. Seed additions show high germination andsurvival in low elevation areas without plant neighbors butnot higher in <strong>the</strong> intertidal zone, pointing to <strong>the</strong> importance<strong>of</strong> high salinity and native plant interactions in restrictingcordgrass distribution. Finally, cordgrass is rare in cobblebeach habitats despite <strong>the</strong> lack <strong>of</strong> native vegetation. Hereshifting cobble sediments contribute to <strong>the</strong> low germination,survival, and growth <strong>of</strong> <strong>the</strong> plant.Cordgrass generally modifies its habitat by accretingsediment around its large root system, forming an elevatedroot mat and changing sediment biogeochemistry(Thompson 1991; Maricle and Lee 2002; Hacker and- 212 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementDethier 2006). In addition, it can provide substantialaboveground structure not normally present in somehabitats. These structural and biogeochemical modificationshave direct and indirect effects on o<strong>the</strong>r plant and animalspecies as mentioned earlier. In our study system, Englishcordgrass differentially modifies four habitat types, resultingin variable, community-wide consequences (Hacker andDethier 2006). For example, we know that <strong>the</strong> accumulation<strong>of</strong> sediment by cordgrass differs between <strong>the</strong> four habitats,with mudflats experiencing <strong>the</strong> greatest sediment accretion(~30 cm) and low and high salinity marshes, <strong>the</strong> least (~10-15 cm) (Hacker and Dethier 2006). Cordgrass growing incobble beaches accretes much less sediment (~20 cm) thanmudflats. A likely explanation for this difference is waveenergy; water movement is lower in mudflats but higher incobble beaches.In addition to accreting sediment, thus changing <strong>the</strong>structural characteristics <strong>of</strong> <strong>the</strong> substrate, cordgrass causes anumber <strong>of</strong> chemical transformations. We found changes insediment water content, redox potential (a proxy for oxygencontent), and salinity in invaded versus native areasdepending on community type (Hacker and Dethier 2006).For example, cordgrass caused a decline in sediment watercontent in mudflats and high salinity marshes suggesting that<strong>the</strong> elevated root mat had less tidal inundation, betterdrainage, and/or more water uptake by cordgrass. In cobblebeaches, <strong>the</strong> opposite was seen; root mat sediments hadhigher water content than <strong>the</strong> unmodified cobble sediments.In low salinity marshes, <strong>the</strong>re was no change in sedimentwater content with <strong>the</strong> presence <strong>of</strong> cordgrass compared tonative vascular plants. Cordgrass generally increases oxygencontent in <strong>the</strong> sediments <strong>of</strong> all <strong>the</strong> communities. This may berelated, in part, to better drainage seen in some <strong>of</strong> <strong>the</strong>communities but is more likely due to oxygen leakage fromroot aerenchyma (Maricle and Lee 2002). Finally, surfacesalinities were generally lower in sediments with cordgrassalthough <strong>the</strong>y did not change in low salinity sites with nativeplant cover. These results suggest that cordgrass shades <strong>the</strong>sediment surface thus decreasing water evaporation and saltaccumulation compared to unvegetated areas although thishypo<strong>the</strong>sis was untested. We have not investigated cordgrasseffects on nutrient cycling but it is clear that cordgrass is amajor carbon source unlike any o<strong>the</strong>r in <strong>the</strong>se communitiesand likely modifies microbial and macr<strong>of</strong>aunal resource use.English cordgrass modifications have significantconsequences for community structure. We compared nativevascular plant and algal abundance in areas with and withoutcordgrass in all communities (Hacker and Dethier 2006). Inthis study, we found that in mudflats, cobble beaches, andhigh salinity marshes, cordgrass invasion caused an increasein native vascular plant cover and decline in algal cover. Byelevating sediments, increasing oxygen content, anddecreasing salinity, cordgrass clearly provides a moresuitable habitat for native vascular salt marsh plants but aless suitable habitat for algae, which require greater tidalinundation to avoid desiccation. In low salinity marshes,native vascular plants declined precipitously, presumably aconsequence <strong>of</strong> cordgrass competitive dominance. We alsocompared marine invertebrates in mud and cobble sedimentsversus adjacent cordgrass patches. Cobble areas withoutcordgrass had oligochaetes, bivalves, and a variety <strong>of</strong>crustaceans; cobble with cordgrass had less infauna because<strong>of</strong> dense root mat formation. Uninvaded mudflats <strong>of</strong>ten hadabundant clams and a variety <strong>of</strong> polychaete worms, whilethose with cordgrass had fewer infauna but more epifaunalcrustaceans such as amphipods (presumably because <strong>of</strong> <strong>the</strong>three–dimensional structure provided by <strong>the</strong> vegetation).Invertebrate studies (Zipperer 1996; O’Connell 2002) inareas invaded by Spartina alterniflora in Willapa Bay, WA,similarly found that certain taxa are excluded by cordgrass(esp. polychaetes and bivalves) while o<strong>the</strong>rs are increased(esp. dipteran larvae and spiders). Although we have notquantitatively measured possible changes in epifaunalcommunities, we have observed a general increase in highermarsh epifauna such as grasshoppers, spiders, snails, mice,and marsh wrens within cordgrass meadows and a decline inshorebirds, some species <strong>of</strong> snails, mussels, and oysters.Cordgrass removal and consequences for post-removalrestorationRemoval <strong>of</strong> English cordgrass, involving mowing andherbicide applications, began in 1997 and has caused amodest 10-20% decline as <strong>of</strong> 2002 (Hacker et al. 2001;Dethier and Hacker 2004). Although local eradication hasoccurred at some sites with minimal treatment repetition,most sites have required repeated removal spanning multipleyears to achieve eradication. To investigate <strong>the</strong> factorshindering removal success, we conducted a multiple sitestudy that linked removal data with ecological factors andremoval methodologies. We found that removal successdepended on removal regime and community type (Dethierand Hacker 2004). The bulk <strong>of</strong> cordgrass decline was due toconsistent, multi-year removal (once per year) in certaincommunities. For example, cobble beaches, high salinitymarshes, and mudflats showed <strong>the</strong> most promising responseto consistent removal even when years <strong>of</strong> removal effortwere similar to low salinity marshes. When removal wasintermittent ( 2 years missed), low salinity marshes andmudflats were <strong>the</strong> least responsive to removal.The pattern <strong>of</strong> regrowth with removal regime points to<strong>the</strong> resiliency <strong>of</strong> cordgrass. If photosyn<strong>the</strong>sis is notcontinually interrupted to slowly kill <strong>the</strong> plant, removalsuccess is compromised due to its ability to regrow. In a set<strong>of</strong> manipulative experiments, we found that biomass gainsunder intermittent removal greatly outweigh losses accruedunder consistent removal both because <strong>of</strong> <strong>the</strong> highlyproductive nature <strong>of</strong> <strong>the</strong> species, <strong>the</strong> benefits <strong>of</strong> competitiverelease, and <strong>the</strong> habitat modifications that facilitate fur<strong>the</strong>rgrowth (Reeder and Hacker 2004).- 213 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaPreliminary study <strong>of</strong> post–removal community structuresuggests that <strong>the</strong> response largely depends on howmodifications created by <strong>the</strong> invader interact with habitattype (Reeder and Hacker 2004). In this study, all sitesexhibited similar community responses following cordgrassremoval. Native vascular plants increased with consistentremoval in all communities, but declined under intermittentremoval. In addition, <strong>the</strong> plant assemblages in <strong>the</strong> two saltmarsh communities were different after cordgrass removalthan <strong>the</strong>ir native counterparts. As a result, post-removalcommunity structure and habitat restoration is more complexthan simply removing cordgrass and anticipating a reversal<strong>of</strong> <strong>the</strong> invasion process.IMPLICATIONS AND FUTURE DIRECTIONSIn this system, we found that removing invasive S.anglica resulted in an increase in native vascular plantsirrespective <strong>of</strong> habitat type (Reeder and Hacker 2004).Although <strong>the</strong>se plants are <strong>the</strong> dominant species in salt marshcommunities, <strong>the</strong>y are uncommon in mudflat and cobblebeach communities, and thus do not represent a restoredpost-removal state. Instead, <strong>the</strong> legacy effects <strong>of</strong> <strong>the</strong>cordgrass infestation produced alternative short termoutcomes, which may or may not continue for some time.Legacy effects include elevated, stabilized sediments andaltered biogeochemical processes that make conditionsbetter for vascular plants and poorer for macroalgae andinfauna. We have observed a similar pattern <strong>of</strong> vascularplant colonization on relic cordgrass root mats both in <strong>the</strong>U.S. (Willapa Bay, Washington) and in o<strong>the</strong>r countries (Baiede Somme, France and Invercargill, New Zealand)suggesting that <strong>the</strong> effect is widespread (personalobservation). Interestingly, in <strong>the</strong> Baie de Somme, nativevegetation that colonizes cordgrass removal areas, inparticular <strong>the</strong> salt marsh plants Salicornia and Atriplex, areharvested and sold as a specialty food. However, in mostcases, long-term vascular plant colonization, especially inmudflat habitats, is likely to prolong <strong>the</strong> main managementissues surrounding invasive cordgrass (i.e. reducing habitatfor shorebirds, infauna, and commercially importantshellfish).Given <strong>the</strong> potential that removal <strong>of</strong> cordgrass can resultin alternative community structure trajectories, wedeveloped a conceptual model that helps predict <strong>the</strong> possibleoutcomes <strong>of</strong> cordgrass removal (Hacker and Dethier 2009).It is based on alternative stable state <strong>the</strong>ory which explains<strong>the</strong> observation that different species assemblages can occurin <strong>the</strong> same general locality at different times (or differentlocalities at <strong>the</strong> same time) because historical events orcontingencies play an important role in creating communitystructure (Lewinton 1969; Su<strong>the</strong>rland 1974; Petraitis andLatham 1999). We suggest that it can provide a usefulframework for identifying <strong>the</strong> processes important to postremovalcommunity structure by identifying <strong>the</strong> factors thatAlternative,post-removal stateMAINTENANCETRANSITIONSA. Cobble Beach and Mud flat CommunitiesMaintenance <strong>of</strong> cordgrasssediment accretion,biogeochemicalprocessesVascular PlantsDominateDecreasealgal, infaunalrecruitmentIncrease vascularplant, epifaunalrecruitmentAlgalRecruitmentVascular PlantRecruitmentWater MovementB. Low and High Salinity Marsh CommunitiesHigher IntertidalVascular PlantsDominateMaintenance <strong>of</strong> cordgrasssediment accretion,biogeochemicalprocessesDecreaseoriginal plantrecruitmentIncreasehigh intertidal plantrecruitmentOriginal PlantRecruitmentHigher IntertidalPlant RecruitmentWater MovementRestored,post-removal stateMAINTENANCEAlgae, InfaunaDominateLoss <strong>of</strong> cordgrasssediment accretion,biogeochemicalprocessesIncreasealgal, infaunalrecruitmentDecrease vascularplant, epifaunalrecruitmentOriginalVascular PlantsDominateLoss <strong>of</strong> cordgrasssediment accretion,biogeochemicalprocessesIncreaseoriginal plantrecruitmentDecreasehigh intertidal plantrecruitmentFig. 2. Depiction <strong>of</strong> <strong>the</strong> hypo<strong>the</strong>sized processes controlling alternative vs.restored community structure for ei<strong>the</strong>r A) cobble beach and mudflathabitats (original communities lack native vascular plants) or B) high andlow salinity marshes (original communities dominated by native vascularplants). Modified from Hacker and Dethier (2009).could lead a community toward or away from a restoredstate.In our model (Hacker and Dethier 2009 based onPetraitis and Latham 1999) <strong>the</strong>re are two community states:1) a restored state defined as <strong>the</strong> replacement <strong>of</strong> <strong>the</strong> lostcommunity assemblage, and its function, after <strong>the</strong> invader isremoved, and 2) an alternative state defined as one in whicha new species assemblage colonizes and persists; it couldalso include reinvasion by <strong>the</strong> original invading species.There are transitional processes that include disturbance andstress, recruitment <strong>of</strong> species, and biological interactions. Inaddition, <strong>the</strong>re are positive feedback processes in which <strong>the</strong>existing species assemblage acts to reinforce and maintainits current structure and function.Applying <strong>the</strong>se ideas to post-removal communitystructure, we can hypo<strong>the</strong>size what may happen to <strong>the</strong> fourhabitats when cordgrass is removed (Hacker and Dethier2009). We predict that cobble beaches will assume arestored state due to <strong>the</strong> interaction <strong>of</strong> both transition andmaintenance processes as outlined in Fig. 2. If we assumethat vascular plant recruitment occurs in cobble beaches, butplant density is low due to high water movement, <strong>the</strong>n analternative state can only be produced if vascular plants can- 214 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and Managementmaintain elevated sediments and altered biogeochemicalprocesses (Fig. 2A, left side). Active water movement and<strong>the</strong> scouring action <strong>of</strong> shifting cobbles and gravel shouldhamper this process. Ultimately, we predict that analternative state will not be maintained due to <strong>the</strong> transitionalprocess <strong>of</strong> water movement increasing sediment erosionaround plant roots and decreasing subsequent vascular plantand infaunal recruitment. As a result, negative feedbackprocesses in <strong>the</strong> maintenance component <strong>of</strong> <strong>the</strong> alternativestate will shift cobble beach community structure to arestored state (Fig. 2A, right side). Algal recruitment willincrease as sediment erosion occurs and cobbles re-emerge,pushing <strong>the</strong> community into a positive feedback loop thatincludes loss <strong>of</strong> sediment, decreased native vascular plantrecruitment, and continued increases in algal and infaunalrecruitment (Fig. 2A, right side).We hypo<strong>the</strong>size that mudflat habitats, because <strong>the</strong>yexperience lower wave action, will have increased vascularplant recruitment, and allow for <strong>the</strong> maintenance <strong>of</strong>cordgrass sediment accretion and biogeochemical processes(Fig. 2A, left side). We expect that a positive feedback loopgenerated by slower water movement and <strong>the</strong> presence <strong>of</strong>vascular plants will maintain sediment characteristicscreated by cordgrass, decrease algal and infaunalrecruitment, and continue to increase native vascular plantrecruitment (Fig. 2A, left side).Finally, we suggest that high and low salinity marshespreviously invaded by cordgrass will experience recruitment<strong>of</strong> higher intertidal vascular plant assemblages, ra<strong>the</strong>r than<strong>the</strong> original assemblage, due to <strong>the</strong> increased tidal elevationproduced by cordgrass-accreted sediments. We expect that<strong>the</strong>se plants will be good at maintaining sediment depth andbiogeochemical processes originally created by cordgrass.Their presence will be maintained via a positive feedbackloop that increases <strong>the</strong>ir own continued recruitment whiledecreasing that <strong>of</strong> <strong>the</strong> lower intertidal plant communitypresent before <strong>the</strong> invasion (Fig. 2B, left side). However, ifwater movement is sufficient to erode sediments or if plantrecruitment is low or delayed, marsh communities may shiftinto a restored state (Fig. 2B, right side).Testing <strong>the</strong>se hypo<strong>the</strong>ses will require both large andsmall-scale experiments in areas where cordgrass has beenremoved. Ultimately, this research will provide natural areamanagers with better ways <strong>of</strong> predicting and evaluating <strong>the</strong>consequences <strong>of</strong> post-removal impact across differenthabitat types. It may be that active removal <strong>of</strong> nativevascular plant assemblages will be necessary in habitats suchas mudflats where restoration could be continually hinderedby <strong>the</strong> recruitment and positive feedbacks produced by <strong>the</strong>seplants. Although invasive species removal and restoration isa critical component to <strong>the</strong> management <strong>of</strong> many naturalareas, most research is site or species-specific. The proposedresearch will help establish <strong>the</strong> importance <strong>of</strong> a general<strong>the</strong>ory on <strong>the</strong> contextual dependence <strong>of</strong> invasive speciesremoval that can be applied to communities with varyingspecies assemblages, physical conditions, and degrees <strong>of</strong>invasion.ACKNOWLEDGEMENTSWe thank <strong>the</strong> many hands that assisted with <strong>the</strong> researchincluding C. Catton, C. Gozart, E. Hellquist, A. Hysert, T.Reeder, T. Riordon, and T. Royce. The research wassupported by Washington and National Sea Grant Programs.REFERENCESBruno, J., and M.D. Bertness. 2001. Habitat modification andfacilitation in benthic marine communities. Pages 201-220 In:M.D. Bertness, S.D. Gaines, and M.E. Hay, eds., MarineCommunity Ecology. Sinauer Associates: Sunderland, MA, pp.201-210.Crooks, J. 2002. Characterizing ecosystem-level consequences <strong>of</strong>biological invasions: <strong>the</strong> role <strong>of</strong> ecosystem engineers. Oikos97:153-166.D’Antonio, C.D., and L.A. Meyerson. 2002. Exotic plant species asproblems and solutions in ecological restoration: a syn<strong>the</strong>sis.Restoration Ecology 10:703–713.Daehler, C.C., and D.R. Strong. 1996. Status, prediction andprevention <strong>of</strong> introduced cordgrass, Spartina spp. invasions inPacific estuaries, USA. Biological Conservation 78:51-58.Dethier, M.N., and S.D. Hacker. 2005. Physical factors vs. bioticresistance in controlling <strong>the</strong> invasion <strong>of</strong> an estuarine marsh grass.Ecological Applications 15:1273-1283.Dethier, M.N., and S.D. Hacker. 2004. Improving managementpractices for invasive cordgrass in <strong>the</strong> Pacific Northwest: a casestudy <strong>of</strong> Spartina anglica. Washington Sea Grant Publications,Seattle, WA, USA.Goss-Custard, J.D., and M.E. Moser. 1988. Rates <strong>of</strong> change in <strong>the</strong>numbers <strong>of</strong> Dunlin, Calidris alpina, wintering in British estuariesin relation to <strong>the</strong> spread <strong>of</strong> Spartina anglica. Journal <strong>of</strong> AppliedEcology 25:95-109.Grosholz, E. 2002. Ecological and evolutionary consequences <strong>of</strong>coastal invasions. Trends in Ecology and Evolution 17:22–27.Hacker, S.D., and M.N. Dethier. 2006. Community modification bya grass invader has differing impacts for marine habitats. Oikos113:279-286.Hacker, S.D., and M.N. Dethier. 2009. Differing consequences <strong>of</strong>removing ecosystem-modifying invaders: Significance <strong>of</strong> impactand community context to restoration potential. In: J.A. Rilovand J.A. Crooks, eds., Biological Invasions in MarineEcosystems. Springer-Verlag: Berlin Heidelberg, pp. 375-385.Hacker, S.D., D. Heimer, C.E. Hellquist, T.G. Reeder, B. Reeves,T. Riordan, and M.N. Dethier. 2001. A marine plant (Spartinaanglica) invades widely varying habitats: potential mechanisms<strong>of</strong> invasion and control. Biological Invasions 3:211-217.Hobbs, R.J., and S.E. Humphries. 1995. An integrated approach to<strong>the</strong> ecology and management <strong>of</strong> plant invasions. ConservationBiology 9:761–770.Jones, C.G., J.H. Lawton, and M. Shachak. 1994. Organisms asecosystem engineers. Oikos 689:373-386.Lewinton, R.C. 1969. The meaning <strong>of</strong> stability. In: Diversity andstability in ecological systems. Brookhaven Symposia in BiologyNo. 22. Brookhaven National Laboratories, Brookhaven, NY,USA, pp. 13-24.Maricle, B.R., and R.W. Lee. 2002. Aerenchyma development andoxygen transport in <strong>the</strong> estuarine cordgrasses Spartinaalterniflora and S. anglica. Aquatic Botany 74:109-120.Myers, J.H., D. Simberl<strong>of</strong>f, A.M. Kuris, and J.R. Carey. 2000.Eradication revisited: dealing with exotic species. Trends inEcology and Evolution 15:316-320.- 215 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaO’Connell, K.A. 2002. Effects <strong>of</strong> invasive Atlantic smoothcordgrass(Spartina alterniflora) on infaunal macroinvertebratecommunities in sou<strong>the</strong>rn Willapa Bay, WA. MS Thesis, WesternWashington University, Bellingham, WA.Parker, I.M., D. Simberl<strong>of</strong>f, W.M. Lonsdale, K. Goodell, M.Wonham, P.M. Kareiva, M.H. Williamson, B. Von Holle, P.B.Moyle, J.E. Byers, and L. Goldwasser. 1999. Impact: toward aframework for understanding <strong>the</strong> ecological effects <strong>of</strong> invaders.Biological Invasions 1:3-19.Petratis, P.S., and R.E. Latham. 1999. The importance <strong>of</strong> scale intesting <strong>the</strong> origins <strong>of</strong> alternative community states. Ecology80:429-442.Power, M.E., D. Tilman, J.A. Estes, B.A. Menge, W.J. Bond, L.S.Mills, G. Daily, J.C. Castilla, J. Lubchenco, and R.T. Paine.1996. Challenges in <strong>the</strong> quest for keystones. Bioscience 46:609–620.Reeder, T.G., and S.D. Hacker. 2004. Factors contributing to <strong>the</strong>removal <strong>of</strong> a marine grass invader (Spartina anglica) andsubsequent potential for habitat restoration. Estuaries 27:244-252.Ruiz, G.M., P. F<strong>of</strong>on<strong>of</strong>f, A.H. Hines, and E.D. Grosholz. 1999.Nonindigenous species as stressors in estuarine and marinecommunities: assessing impacts and interactions. Limnology andOceanography 44:950–972.Su<strong>the</strong>rland, J.P. 1974. Multiple stable points in naturalcommunities. American Naturalist 108: 859-873.Triplet, P., C. Fagot, S. Van Imbeck, A. Sournia, and F. Sueur.2002. Rôle de la végétation dans l'utilisation de l'estran par leslimicoles. Alauda 70:445-449.Thompson, J.D. 1991. The biology <strong>of</strong> an invasive plant: Whatmakes Spartina anglica so successful? BioScience 41:393-401.Zavaleta, E.S., R.J. Hobbs, and H.A. Mooney. 2001. Viewinginvasive species removal in a whole-ecosystem context. Trendsin Ecology and Evolution 16:454-459.Zipperer, V.T. 1996. Ecological effects <strong>of</strong> <strong>the</strong> introducedcordgrass, Spartina alterniflora, on <strong>the</strong> benthic communitystructure <strong>of</strong> Willapa Bay, WA. MS Thesis, University <strong>of</strong>Washington, Seattle, WA.- 216 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementSPARTINA CONTROL STRATEGY AND EXPERIENCE IN THE SAN FRANCISCO ESTUARYE.K. GRIJALVASan Francisco Estuary Invasive Spartina Project, 2612-A 8 th Street, Berkeley, CA 94710; ekgrijalva@spartina.orgIn 2004, <strong>the</strong> San Francisco Estuary Invasive Spartina Project (ISP) initiated its first year <strong>of</strong> a regionwide,coordinated Spartina control program. The strategies for selection <strong>of</strong> <strong>the</strong> treatment locationsand <strong>the</strong> coordination <strong>of</strong> <strong>the</strong> projects were developed jointly by <strong>the</strong> ISP and its partners, which includea number <strong>of</strong> local, regional, state, and federal agencies and non-pr<strong>of</strong>it organizations. Fundingfor <strong>the</strong> work was contributed by many <strong>of</strong> <strong>the</strong> partners, by <strong>the</strong> ISP’s host agency (California StateCoastal Conservancy), by grants from a state-federal consortium, and o<strong>the</strong>r sources. The controlplans were developed by <strong>the</strong> ISP, and implemented by <strong>the</strong> partners. Approximately 176 hectares (ha)(435 acres [ac]) <strong>of</strong> non-native Spartina (Spartina alterniflora and hybrids, S. densiflora, and S. patens)were treated in 2004 using a variety <strong>of</strong> methods including covering, digging, and treatment withaquatic herbicide. The ISP facilitates acquisition <strong>of</strong> permits, grants, and contracts to implement <strong>the</strong>control work. Successful control <strong>of</strong> Spartina in <strong>the</strong> San Francisco Estuary is complicated by severalfactors, including an extremely short treatment season (September 1 to mid-October) to avoid disturbingendangered California clapper rails during <strong>the</strong>ir breeding season, and <strong>the</strong> greater-thanexponentialspread <strong>of</strong> <strong>the</strong> S. alterniflora hybrid swarm. At <strong>the</strong> end <strong>of</strong> <strong>the</strong> 2004 control season, approximately627 ha (1,550 ac) <strong>of</strong> non-native Spartina remained untreated in <strong>the</strong> San Francisco Estuary.In order to address this rapidly-expanding infestation, <strong>the</strong> ISP seeks to have control plans inplace for all non-native Spartina in <strong>the</strong> estuary by <strong>the</strong> end <strong>of</strong> 2005, even though treatment may notbe implemented on all sites until <strong>the</strong> following years to minimize impact on endangered species.One <strong>of</strong> <strong>the</strong> tools that <strong>the</strong> ISP hopes will help assure a successful control program is <strong>the</strong> aquatic herbicideimazapyr, which should be registered for use in California by <strong>the</strong> summer <strong>of</strong> 2005. The highefficacy and suitability <strong>of</strong> this herbicide for estuarine use bodes well for control efforts around SanFrancisco Bay. Building upon <strong>the</strong> structure and partnerships developed during <strong>the</strong> 2004 Spartinacontrol season, <strong>the</strong> ISP believes it is possible to set in place coordinated, sustainable, estuary-widemanagement and control <strong>of</strong> non-native Spartina.Keywords: Spartina, cordgrass, hybrids, hybrid swarm, imazapyr, glyphosate, estuary, Californiaclapper railThe spread <strong>of</strong> non-native Spartina in <strong>the</strong> San FranciscoEstuary has been a topic <strong>of</strong> concern in <strong>the</strong> region since <strong>the</strong>early 1970s, following <strong>the</strong> initial introduction <strong>of</strong> several species:S. densiflora (Chilean cordgrass), S. anglica (Englishcordgrass), and S. patens (salt meadow cordgrass), but especiallyS. alterniflora (smooth cordgrass). By <strong>the</strong> early1990s, ecologists and land managers in <strong>the</strong> estuary recognized<strong>the</strong> increasing impacts <strong>of</strong> <strong>the</strong> invasion as marshlandwithin <strong>the</strong>ir respective jurisdictions began to support everexpandingpopulations <strong>of</strong> non-native Spartina species.Around this time, a hybridization event between <strong>the</strong> nonnativeS. alterniflora and <strong>the</strong> native S. foliosa (Pacificcordgrass) produced <strong>the</strong> first individual clones <strong>of</strong> a vigoroushybrid swarm that swiftly outpaced its alien parent species inits ability to colonize a wide range <strong>of</strong> habitats in <strong>the</strong> SanFrancisco Estuary.In 2000, <strong>the</strong> California State Coastal Conservancy(Conservancy) initiated <strong>the</strong> Invasive Spartina Project (ISP)to stave <strong>of</strong>f <strong>the</strong> invasion <strong>of</strong> non-native cordgrass and its potentialimpacts. The ISP was intended to be a regionallycoordinatedeffort <strong>of</strong> federal, state and local agencies, privatelandowners, and o<strong>the</strong>r interested parties, with <strong>the</strong> ultimategoal <strong>of</strong> arresting and reversing <strong>the</strong> spread <strong>of</strong> non-nativeSpartina in <strong>the</strong> estuary. It was not until <strong>the</strong> late summer <strong>of</strong>2004 that all <strong>of</strong> <strong>the</strong> required permitting and planning hadbeen completed sufficient to initiate <strong>the</strong> first season <strong>of</strong>Spartina control work estuary-wide.The process to produce an Environmental ImpactStatement and Environmental Impact Report (FEIS/R), satisfyingboth Federal and State requirements, began at <strong>the</strong> inception<strong>of</strong> <strong>the</strong> ISP in 2000. In October <strong>of</strong> 2003, following <strong>the</strong>conclusion <strong>of</strong> <strong>the</strong> public comment period on <strong>the</strong> draft documentand preparation <strong>of</strong> comment responses, <strong>the</strong> Final EIS/Rwas completed by <strong>the</strong> Conservancy and <strong>the</strong> U.S. Fish andWildlife Service (California Coastal Conservancy andUSFWS 2003). The FEIS/R was <strong>the</strong>n used by <strong>the</strong> FWS toproduce a Programmatic Biological Opinion (PBO)(USFWS 2004(a)) for ISP’s Control Program, which outlined<strong>the</strong> requirements for creating detailed individual BiologicalOpinions (BOs) for each proposed Spartina treatment- 217 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinasite. Issuance <strong>of</strong> site-specific BOs depended on <strong>the</strong> development<strong>of</strong> site-specific Spartina control plans by <strong>the</strong> ISP andits partners.SITE-SPECIFIC SPARTINA CONTROL PLANSConcurrent with, and subsequent to <strong>the</strong> completion <strong>of</strong><strong>the</strong> FEIS/R and PBO, individual site-specific Spartina controlplans (SSPs) (California Coastal Conservancy 2004(b))for 2004 were drafted by <strong>the</strong> ISP and its partners for 16 sitesscattered throughout <strong>the</strong> estuary. These sites were chosenbased on a variety <strong>of</strong> characteristics, including existing partnerinvolvement, infestation age and composition, endangeredspecies issues, access issues, adjacent land uses ando<strong>the</strong>r criteria. The 16 sites chosen for Spartina control in2004 included 45 sub-areas (areas fur<strong>the</strong>r delineated for ease<strong>of</strong> treatment) encompassing an estimated 181 net ha (447 netac) <strong>of</strong> Spartina, within just over 6,070 ha (15,000 ac) <strong>of</strong> estuarymarshland. Each <strong>of</strong> <strong>the</strong> SSPs included: Scope <strong>of</strong> Work Work Program defining <strong>the</strong> schedule <strong>of</strong> work Budget Impact Identification Matrix Impact Mitigation Matrix Spill Prevention Protocols Drift Reduction Protocols Marsh Safety Recommendations Site Maps Site PhotographsAs part <strong>of</strong> <strong>the</strong> requirements <strong>of</strong> <strong>the</strong> California EnvironmentalQuality Act (CEQA), each SSP contained an ImpactIdentification Matrix and an Impact Mitigation Matrix. TheImpact Identification Matrix evaluated <strong>the</strong> suite <strong>of</strong> Spartinacontrol methods proposed for each site for potential impactsto environmental resources. Included in <strong>the</strong> matrix were impactsrelated to geomorphology and hydrology, water quality,biological resources, air quality, noise, human health andsafety, visual resources, and cumulative impacts.Once <strong>the</strong> impacts associated with Spartina control workwere identified for a given site, an Impact Mitigation Matrixwas prepared. Each site’s matrix explicitly referenced thosemitigation measures defined in <strong>the</strong> FEIS/R that were to beimplemented in <strong>the</strong> field. Impact Mitigation Matrices weredesigned to require verification signatures next to each mitigationmeasure. Both <strong>the</strong> implementing agency representativeand a representative <strong>of</strong> <strong>the</strong> ISP were required to verifythat all mitigation measures were implemented.PERMITTING AND GRANTSThe completion <strong>of</strong> <strong>the</strong> SSPs enabled FWS review <strong>of</strong> <strong>the</strong>documents for issuance <strong>of</strong> site-specific BOs which analyzed<strong>the</strong> effect <strong>of</strong> <strong>the</strong> proposed work on <strong>the</strong> suite <strong>of</strong> endangeredspecies within <strong>the</strong> scope <strong>of</strong> <strong>the</strong> project (USFWS 2004(b)).Individual site-specific Environmental Assessments (EAs)were also prepared at this time to determine <strong>the</strong> effect <strong>of</strong> <strong>the</strong>proposed work on <strong>the</strong> greater environment beyond endangeredspecies issues (USFWS 2004(c)). The results <strong>of</strong> <strong>the</strong>seanalyses provided <strong>the</strong> basis for a Finding <strong>of</strong> No SignificantImpact (FONSI), which declared that <strong>the</strong> proposed Spartinacontrol work would not produce “significant impact” on <strong>the</strong>estuary’s natural resources (USFWS 2004(d)). Additionally,a Record <strong>of</strong> Decision (ROD) was published in <strong>the</strong> FederalRegister announcing <strong>the</strong> completion <strong>of</strong> <strong>the</strong> FEIS/R (USFWS2004(e)).Soon after its inception, <strong>the</strong> ISP worked to identify partneragencies and groups that would be willing to join in <strong>the</strong>effort to control non-native Spartina in <strong>the</strong> estuary. Many <strong>of</strong><strong>the</strong> partners involved in developing SSPs for <strong>the</strong>ir jurisdictionslacked <strong>the</strong> funds to initiate control work on <strong>the</strong>ir own,so outside funding for control operations were necessary forwork to proceed. In some cases, funding was not <strong>the</strong> constrainingfactor for <strong>the</strong> partner, but instead, clearing permittinghurdles, logistics or knowledge <strong>of</strong> control methodologiespresented <strong>the</strong> greatest challenge. In each case, <strong>the</strong> ISPprovided <strong>the</strong> necessary assistance to implement <strong>the</strong> Spartinacontrol plan. Where funding was <strong>the</strong> limiting factor, grantsfrom <strong>the</strong> Conservancy were provided to enable work to proceed.The terms <strong>of</strong> <strong>the</strong>se grant agreements were negotiatedduring late 2003 and <strong>the</strong> first half <strong>of</strong> 2004. Ultimately <strong>the</strong>Conservancy and <strong>the</strong> CalFed Bay Delta Authority issued atotal <strong>of</strong> $350,000 in grants for Spartina control work in <strong>the</strong>estuary for <strong>the</strong> 2004 control season. Official ISP partners forthis season included U.S. Fish and Wildlife Service, EastBay Regional Parks District, Alameda County Flood ControlDistrict, California Department <strong>of</strong> Transportation, SantaClara Valley Water District, City <strong>of</strong> Palo Alto, Friends <strong>of</strong>Corte Madera Creek Watershed, Tiburon Audubon Society,Marin Conservation Corps, and California Department <strong>of</strong>Parks and Recreation. Inclusion <strong>of</strong> any organization oragency as an <strong>of</strong>ficial ISP partner required that partner’s governingbody <strong>of</strong>ficially adopt <strong>the</strong> ISP’s FEIS/R, develop anSSP in coordination with <strong>the</strong> ISP, and adhere to any mitigationsdefined in <strong>the</strong> Impact Mitigation Matrix developed for<strong>the</strong> site within <strong>the</strong>ir jurisdiction.For each type <strong>of</strong> control method proposed for each site,certain permits needed to be acquired before work couldbegin. In <strong>the</strong> case <strong>of</strong> herbicide applications, all treatmentswere required to comply with <strong>the</strong> Environmental ProtectionAgency’s National Pollutant Discharge Elimination System(NPDES), administered by <strong>the</strong> California Water QualityControl Board’s (CWQCB) San Francisco Bay regional <strong>of</strong>fice.However, at <strong>the</strong> outset <strong>of</strong> 2004, <strong>the</strong> CWQCB had yet todevelop an NPDES permit to cover <strong>the</strong> discharge <strong>of</strong> aquaticpesticides, a novel application <strong>of</strong> this portion <strong>of</strong> <strong>the</strong> CleanWater Act that had, until recently, been reserved for <strong>the</strong>regulation <strong>of</strong> effluent discharge. As a consequence,herbicide-based control operations throughout <strong>the</strong> state wereon hold until <strong>the</strong> CWQCB revised its permitting procedures- 218 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and Managementto address aquatic herbicides. The revisions were adopted inJune <strong>of</strong> 2004, enabling ISP partners to issue Notices <strong>of</strong> Intent(NOIs) to be covered under <strong>the</strong> Statewide GeneralNPDES Permit for Aquatic Weed Control.The NPDES permit also requires developing an AquaticPesticide Application Plan (APAP) identifying <strong>the</strong> proceduresfor applying herbicides, <strong>the</strong> chemicals to be used, <strong>the</strong>species targeted for treatment, a description <strong>of</strong> <strong>the</strong> waterbodysystem, and o<strong>the</strong>r related information. Within <strong>the</strong> APAP is aWater Quality Monitoring Plan (WQMP), which defines <strong>the</strong>protocols for water quality sampling <strong>of</strong> treatment areaswhere herbicide is used. While <strong>the</strong> ISP developed <strong>the</strong> APAPon behalf <strong>of</strong> <strong>the</strong> ISP partners, <strong>the</strong>y contracted <strong>the</strong> services <strong>of</strong><strong>the</strong> San Francisco Estuary Institute (SFEI) to fur<strong>the</strong>r developand implement <strong>the</strong> WQMP for <strong>the</strong> work proposed by <strong>the</strong> ISPand its partners, and to report on its findings following <strong>the</strong>2004 treatment season (California Coastal Conservancy2004(c)).In addition, access permits from regional and localagencies were necessary for treatment on many <strong>of</strong> <strong>the</strong> 16sites slated for Spartina control and <strong>the</strong> ISP coordinated <strong>the</strong>irprocurement.TREATMENT METHODSA variety <strong>of</strong> methods can be used to treat non-nativeSpartina populations in <strong>the</strong> estuary. However, very rarely isany particular treatment method appropriate for all situations.In <strong>the</strong> FEIS/R, <strong>the</strong> ISP proposed a suite <strong>of</strong> controlmethods and approaches based on <strong>the</strong> cumulative expertise<strong>of</strong> Spartina control operations worldwide, especially <strong>the</strong>work being done to control S. alterniflora in Willapa Bay,Washington State, USA. In 2004, <strong>the</strong> ISP tested several possiblecontrol techniques to determine <strong>the</strong> relative benefits <strong>of</strong>each.Chemical ControlGlyphosate herbicide (Aquamaster® or Rodeo®) wasused in most <strong>of</strong> <strong>the</strong> area <strong>of</strong> Spartina infestation that waschemically treated in 2004. The application <strong>of</strong> herbicidescomprised <strong>the</strong> largest part <strong>of</strong> <strong>the</strong> Control Program for a host<strong>of</strong> reasons, not least <strong>of</strong> which was <strong>the</strong> need to control largeacreage at low cost to stay ahead <strong>of</strong> <strong>the</strong> rapid expansion rate<strong>of</strong> <strong>the</strong> invader. Moreover, this method has a relatively lowimpact on <strong>the</strong> treated marshlands in contrast to o<strong>the</strong>r treatmentoptions. Drawbacks to herbicide use include <strong>the</strong> potentialfor herbicide or fuel spills, negative public perception,timing restrictions, and potential water quality impacts. Glyphosateherbicide was applied via tracked amphibious vehicles(Hydrotraxx or Argo) fitted with spray apparatus, truckmountedspray equipment, backpack sprayers, shallowbottomedboats with outboard motors and broadcast aerialapplications via helicopter with a 30-foot boom fitted with anozzle array.Manual ControlFor those sites not yet fully established—where <strong>the</strong>rewere very small, pioneering infestations to be treated—manual control techniques were used. These methods were<strong>of</strong>ten incorporated into public outreach efforts, involvingvolunteer groups that worked on marshland restorationprojects. For smaller sites, digging <strong>the</strong> plants resulted ingood control, but swiftly exhausted worker enthusiasm andresulted in heavy impacts to <strong>the</strong> marsh. Digging plants ininfestations much larger than 25 square meters is prohibitivelyexpensive and has <strong>the</strong> potential to leave lasting damageto <strong>the</strong> treated marsh in <strong>the</strong> form <strong>of</strong> deep holes ortrenches. In such cases, tarping or covering <strong>of</strong> <strong>the</strong> plants wasused to good effect. Heavy syn<strong>the</strong>tic tarps were staked inplace over trampled stands <strong>of</strong> non-native Spartina in severallocations in <strong>the</strong> estuary, as well as at Point Reyes NationalSeashore on <strong>the</strong> outer coast north <strong>of</strong> <strong>the</strong> Golden Gate. Tarpswere left in place for at least eight months and required regularmonitoring to assure that <strong>the</strong> covering remained in placeduring tidal fluctuation, especially storms. Removal <strong>of</strong> <strong>the</strong>covering following successful control <strong>of</strong> <strong>the</strong> Spartina was<strong>of</strong>ten difficult because <strong>of</strong> <strong>the</strong> heavy accumulation <strong>of</strong> mud ontop <strong>of</strong> <strong>the</strong> tarps, requiring laborious excavations to unearth<strong>the</strong> covering. Drawbacks to covering include <strong>the</strong> high cost <strong>of</strong>installation, maintenance and removal and <strong>the</strong> attendantdamage to <strong>the</strong> marsh, as well as <strong>the</strong> negative environmentalimpacts <strong>of</strong> petrochemical production associated with <strong>the</strong>manufacture <strong>of</strong> <strong>the</strong> heavy syn<strong>the</strong>tic fabric used in this treatmentmethod. Two <strong>of</strong> <strong>the</strong> sub-areas identified for Spartinacontrol in 2004 used small-scale mowing <strong>of</strong> <strong>the</strong> infestation.Gas-powered hand mowers fitted with a tri-blade were ableto make quick work <strong>of</strong> <strong>the</strong> standing material which was ei<strong>the</strong>rleft in place in <strong>the</strong> marsh (early season, pre-seed set) orremoved from <strong>the</strong> marsh and disposed <strong>of</strong> at a landfill (lateseason, post seed-set). Mowing in this way serves to controlseed and pollen spread and may slow <strong>the</strong> vegetative expansion<strong>of</strong> <strong>the</strong> Spartina clone by usurping root reserves. Mowingalone does not kill <strong>the</strong> plants, however. Unless coupledwith o<strong>the</strong>r control methods, mowing would need to be implementedseveral times each year in perpetuity and yetwould still allow some vegetative expansion.O<strong>the</strong>r methods <strong>of</strong> Spartina control tried elsewhere werenot attempted in <strong>the</strong> 2004 control season for various reasons.Macerating or tilling was not used because <strong>of</strong> <strong>the</strong> concerns<strong>of</strong> land managers that <strong>the</strong> high amount <strong>of</strong> urban refuse presentin estuary marshes (tires, shopping carts, blocks <strong>of</strong> concrete,pieces <strong>of</strong> metal, wood debris embedded with bolts ornails) and natural wrack (driftwood logs), would result inhigh maintenance costs or operator injury. The extremelys<strong>of</strong>t nature <strong>of</strong> most <strong>of</strong> <strong>the</strong> estuary mud where control wouldtake place was also taken into consideration, as <strong>the</strong> machineryinvolved in <strong>the</strong>se methods is heavy and very likely to getstuck in place. Crushing <strong>the</strong> plants, where machinery is used- 219 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaFig. 1. 2004 Spartina treatment sites in <strong>the</strong> San Francisco Estuary, CA.to press <strong>the</strong> plants into <strong>the</strong> mud, was tried on a very smallscale, but <strong>the</strong> visual impacts <strong>of</strong> such a treatment method aswell as potential impacts to endangered species made thisoption unsuitable for large-scale use.TIMINGSpartina control work in <strong>the</strong> estuary is limited by morethan just <strong>the</strong> daunting amount <strong>of</strong> paperwork necessary toinitiate treatment. The estuary is a highly dynamic system,and <strong>the</strong> areas typically infested with non-native Spartinabear <strong>the</strong> brunt <strong>of</strong> much <strong>of</strong> this dynamism. Occupying manyhabitats, including <strong>the</strong> bayfront edge as well as more shelteredmarshlands, non-native Spartina infestations are subjectto daily tidal inundation, storms, wrack deposits, urbanrun<strong>of</strong>f and pollution, and sediment deposition. Treatmentmethods selected for <strong>the</strong>se areas must account for all <strong>of</strong><strong>the</strong>se factors. Since large-scale treatment in 2004 reliedheavily on herbicide use, <strong>the</strong> treatment parameters <strong>of</strong> glyphosatefur<strong>the</strong>r dictated <strong>the</strong> available treatment window in<strong>the</strong> estuary.Glyphosate herbicide has specific requirements thatmust be satisfied in order to achieve high treatment efficacies.For Spartina control <strong>the</strong>se are: long dry times (12-24- 220 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and Managementhours) where <strong>the</strong> plant is nei<strong>the</strong>r inundated at high tide norrained on, low amounts <strong>of</strong> silt and salt on <strong>the</strong> treated plants,and complete coverage <strong>of</strong> <strong>the</strong> treated plant (“spray-to-wet”).Where Spartina grows in <strong>the</strong> San Francisco Estuary, <strong>the</strong>serequirements can be difficult to meet. Tidal fluctuations precludemost low and mid-elevation sites from adequate drytime for glyphosate most <strong>of</strong> <strong>the</strong> year. Siltation is a constantproblem in <strong>the</strong> highly turbid waters <strong>of</strong> <strong>the</strong> estuary, <strong>of</strong>ten resultingin much <strong>of</strong> <strong>the</strong> herbicide binding to <strong>the</strong> accumulatedsilt on <strong>the</strong> plant’s leaves ra<strong>the</strong>r than entering <strong>the</strong> plant’s tissuesand translocating properly. Finally, complete coverage<strong>of</strong> <strong>the</strong> targeted Spartina can be difficult to achieve over largestands or in areas difficult to access. Given all <strong>of</strong> <strong>the</strong>se constraints,very few days are available in a given growing season(late May through mid-October) with <strong>the</strong> convergence <strong>of</strong>conditions necessary for efficacious treatment using thisherbicide.The San Francisco Estuary is also <strong>the</strong> habitat <strong>of</strong> <strong>the</strong>Federal and State endangered California clapper rail (Ralluslongirostris obsoletus), whose breeding season extends fromFebruary 1 through August 31 <strong>of</strong> each year. Many <strong>of</strong> <strong>the</strong>marshes where larger populations <strong>of</strong> clapper rail breed havebeen invaded by non-native Spartina, and represent <strong>the</strong> majority<strong>of</strong> <strong>the</strong> infestation in <strong>the</strong> estuary. No Spartina controlwork can occur during <strong>the</strong> clapper rail breeding season in<strong>the</strong>se marshes. Therefore, <strong>the</strong> larger infestations <strong>of</strong> nonnativeSpartina may only be treated between September 1and roughly <strong>the</strong> end <strong>of</strong> October every year. During this time,tidal windows <strong>of</strong> opportunity that allow for <strong>the</strong> required drytime for glyphosate are limited. Finally, in most sites in <strong>the</strong>estuary, late-morning or early-afternoon winds develop thatexceed spray drift reduction criteria for herbicide application(10 mph sustained winds). As a result, in <strong>the</strong> 2004 treatmentseason, Spartina control work within clapper rail occupiedmarsh was restricted to roughly 6-10 days where all necessaryconditions were met.2004 TREATMENTAt <strong>the</strong> outset <strong>of</strong> <strong>the</strong> 2004 treatment season, <strong>the</strong> SSPscalled for treatment <strong>of</strong> 45 individual sub-areas within <strong>the</strong>selected 16 sites. The total area targeted was 181 ha (447ac), which roughly coincided with <strong>the</strong> initial estimates <strong>of</strong>population size determined in 2001, just after <strong>the</strong> ISP’s inceptionand initial inventory mapping <strong>of</strong> <strong>the</strong> non-nativeSpartina in <strong>the</strong> estuary. Figure 1 shows <strong>the</strong> locations <strong>of</strong> <strong>the</strong>various control sites within <strong>the</strong> estuary.As <strong>of</strong> December 2004, 176 ha (435 ac) <strong>of</strong> targeted nonnativeSpartina had been treated, with roughly 0.8 ha (2 ac)slated for manual control work in January 2005. Herbicidetreatment <strong>of</strong> several sub-areas with large infestations wascancelled as a result <strong>of</strong> heavy October rains, which preventedvehicle travel on bay-mud levees. Planned use <strong>of</strong> anamphibious excavator (Aquamog) was also postponed because<strong>of</strong> rain.RATE OF SPREADIn <strong>the</strong> fall <strong>of</strong> 2004 <strong>the</strong> ISP Inventory Monitoring Programproduced a Monitoring Report based on selected sitessurveyed during 2003 (California Coastal Conservancy2004(a)). This report analyzed <strong>the</strong> change in area <strong>of</strong> nonnativeSpartina at 28 monitoring sites stratified across <strong>the</strong>estuary by subregion (latitude), site type and marsh type.Based on sample surveys, <strong>the</strong> report extrapolated an average244% increase in area between 2001 and 2003 from all nonnativeSpartina species in <strong>the</strong> estuary. Spartina alterniflorahybrids were found to be spreading at <strong>the</strong> fastest rate, by asmuch as 317% over that time period. In 2001, an estuarywideinventory mapped approximately 195 ha (470 acres) <strong>of</strong>non-native Spartina, with hybrids comprising all but five <strong>of</strong><strong>the</strong>se hectares. Applying <strong>the</strong> 317% rate <strong>of</strong> increase results inan estimate <strong>of</strong> as much as 793 ha (1960 acres) <strong>of</strong> S. alterniflorahybrids in <strong>the</strong> estuary in 2003.Based on <strong>the</strong>se numbers, 2004 treatment efforts resultedin some 20% <strong>of</strong> <strong>the</strong> estuary’s Spartina population receivingtreatment during <strong>the</strong> year, leaving 80% untreated. If <strong>the</strong> untreatedarea doubles by <strong>the</strong> 2005 treatment season, <strong>the</strong>recould be as much as 1,294 ha (3,200 ac) <strong>of</strong> non-nativeSpartina requiring treatment during that year. This numberbecomes even larger if efficacy rates (40-80%) <strong>of</strong> <strong>the</strong> varioustreatment methods are factored into <strong>the</strong> calculation.OUTLOOK FOR 2005 AND BEYONDGiven <strong>the</strong> “supra-exponential” expansion <strong>of</strong> <strong>the</strong>Spartina hybrid swarm in <strong>the</strong> estuary, <strong>the</strong> ISP has determinedthat only an aggressive, comprehensive strategyaimed at treating all <strong>of</strong> <strong>the</strong> Spartina during <strong>the</strong> 2005 controlseason has a realistic chance <strong>of</strong> eradicating <strong>the</strong> invader from<strong>the</strong> estuary. Building upon <strong>the</strong> partnerships and experiencedeveloped during <strong>the</strong> 2004 treatment season, <strong>the</strong> ISP willaim to put in place each <strong>of</strong> <strong>the</strong> four main components <strong>of</strong>Spartina treatment for each site or sub-area in <strong>the</strong> estuary.The components, broadly defined are:• Partner identification and buy-in. Identification <strong>of</strong><strong>the</strong> specific agency, landowner or land manager for<strong>the</strong> estimated 130 sub-areas requiring treatmentduring 2005 can be difficult given overlappingjurisdictional boundaries, property lines that bisectSpartina infestations, and o<strong>the</strong>r issues. Much <strong>of</strong> thiswork for public property has been accomplishedduring <strong>the</strong> 2004 treatment season; however,extensive public outreach will be necessary for <strong>the</strong>large number <strong>of</strong> individual private property ownerswhose properties are infested with non-nativeSpartina.• Development <strong>of</strong> Site-Specific Treatment Plans. Foreach sub-area or site a site-specific treatment planwill be produced following <strong>the</strong> model <strong>of</strong> <strong>the</strong> 2004SSPs, and incorporating elements <strong>of</strong> <strong>the</strong> FWS BOand EA. These plans will be developed in close- 221 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinacoordination with <strong>the</strong> ISP partners responsible for<strong>the</strong> infested area.• Procurement <strong>of</strong> Funding. Identifying a fundingsource for <strong>the</strong> proposed work on <strong>the</strong> individual sitecan take many forms. In addition to grants awardedby <strong>the</strong> Conservancy, <strong>the</strong> ISP will seek grants fromo<strong>the</strong>r sources and coordinate volunteer activitiesappropriate for each site based on <strong>the</strong> Site-SpecificPlans.• Obtain Necessary Permits for Work. In closecoordination with <strong>the</strong> regulatory agencies whosepurview encompasses <strong>the</strong> areas slated for Spartinatreatment, <strong>the</strong> ISP will aid partners in obtaining <strong>the</strong>necessary permits for work on <strong>the</strong>ir site.In 2005 treatment work will continue, and expandwhere appropriate, on <strong>the</strong> sites treated in 2004. As time is <strong>of</strong><strong>the</strong> essence in treating Spartina, several enhanced controltechniques will be employed to <strong>the</strong> fullest extent possible tomaximize <strong>the</strong> treatment area in 2005.Chief among <strong>the</strong>se is <strong>the</strong> use <strong>of</strong> <strong>the</strong> herbicide imazapyr(Habitat®), which is well suited to <strong>the</strong> challenges <strong>of</strong>Spartina control in an estuarine environment. Applicationrequirements for imazapyr herbicide are not as challengingas those for glyphosate; required dry times are shorter andplants need not be completely covered with <strong>the</strong> chemical forhigh efficacy (less chemical used). These qualities allowtreatment <strong>of</strong> larger areas <strong>of</strong> Spartina in a given time periodwhile also providing wider tidal windows for treatment.The incorporation <strong>of</strong> imazapyr herbicide into <strong>the</strong> ControlProgram also enables greater use <strong>of</strong> aerial (helicopter)treatments. An estimated 42% <strong>of</strong> <strong>the</strong> total area slated fortreatment may be suitable for aerial treatment. Some <strong>of</strong> thisarea is difficult to access on <strong>the</strong> ground; for <strong>the</strong>se sites, replacingground-based treatment with aerial applications willsave time and money, and enable personnel to target lessdemanding sites.Beyond <strong>the</strong> 2005 Spartina control ceason, <strong>the</strong> ISPanticipates establishing <strong>the</strong> initial stages <strong>of</strong> a land-managerbasedSpartina monitoring and control program, as a pilotproject for eventual dissolution <strong>of</strong> <strong>the</strong> ISP. The 2006 and2007 Spartina control seasons will require <strong>the</strong> ISP to be infull effect coordinating estuary-wide eradication efforts, butlater, and once <strong>the</strong> infestation has been reduced to moremanageable levels, individual land managers will graduallyneed to assume responsibility for keeping low-level remnantinfestations <strong>of</strong> non-native, invasive Spartina under control.In this way, <strong>the</strong> ISP seeks to render itself unnecessaryfollowing <strong>the</strong> 2010 Spartina treatment season.CONCLUSIONSThe infestation <strong>of</strong> non-native Spartina in <strong>the</strong> San FranciscoEstuary is spreading at a supra-exponential rate in <strong>the</strong>absence <strong>of</strong> widespread treatment. However, control effortsundertaken during <strong>the</strong> 2004 treatment season indicate thatcontrol and eventual eradication <strong>of</strong> non-native Spartina in<strong>the</strong> estuary is possible given continued funding, expandedpartner involvement, and political will. New tools, including<strong>the</strong> use <strong>of</strong> imazapyr herbicide, are expected to greatly assistin achieving this goal.ACKNOWLEDGEMENTSThe author would like to thank Peggy Ol<strong>of</strong>son and KatyZaremba <strong>of</strong> <strong>the</strong> Invasive Spartina Project, as well as all ISPpartners involved in Spartina control during <strong>the</strong> 2004 controlseason, and acknowledges <strong>the</strong> financial support from <strong>the</strong>California Coastal Conservancy (Contract #02-153).REFERENCESCalifornia Coastal Conservancy and U.S. Fish and Wildlife Service.2003. San Francisco Estuary Invasive Spartina Project:Spartina Control Program: Final Environmental Impact Statement/Environmental Impact Report (Vol. 1), and Appendices(Vol. 2). Berkeley, CA.California Coastal Conservancy. 2004(a). San Francisco EstuaryInvasive Spartina Project Monitoring Report for 2003. Producedby <strong>the</strong> Invasive Spartina Project for <strong>the</strong> California Coastal Conservancy.Berkeley, CA.California Coastal Conservancy. 2004(b). San Francisco EstuaryInvasive Spartina Project: Spartina Control Program, 2004 Site-Specific Spartina Control Plans. Prepared by SFEISP for <strong>the</strong>California Coastal Conservancy and <strong>the</strong> U.S. Fish & WildlifeService. Berkeley, CA.California Coastal Conservancy. 2004(c). Aquatic Pesticide ApplicationPlan for <strong>the</strong> San Francisco Estuary Invasive SpartinaProject. Produced by <strong>the</strong> Invasive Spartina Project for <strong>the</strong> CaliforniaCoastal Conservancy. Berkeley, CA.U.S. Fish and Wildlife Service. 2004(a). Programmatic FormalIntra-Service Endangered Species Consultation on <strong>the</strong> SanFrancisco Estuary Invasive Spartina Project. Sacramento, CA.U.S. Fish and Wildlife Service. 2004(b). Formal Intra-ServiceEndangered Species Consultation on implementation <strong>of</strong> <strong>the</strong> SanFrancisco Estuary Invasive Spartina Project: Spartina ControlProgram. Sacramento, CA.U.S. Fish and Wildlife Service. 2004(c). Environmental Assessment,San Francisco Estuary Invasive Spartina Project 2004Spartina Control Program, San Francisco Bay, California. Sacramento,CA.U.S. Fish and Wildlife Service. 2004(d). Finding <strong>of</strong> No SignificantImpact Statement for a Proposed Action to Control Non-nativeInvasive Spartina at 15 Locations In <strong>the</strong> San Francisco Bay Estuary,California. Sacramento, CAU.S. Fish and Wildlife Service. 2004(e). Record <strong>of</strong> Decision, SanFrancisco Estuary Invasive Spartina Project, Spartina ControlProgram, Final Environmental Impact Statement. Sacramento,CA.- 222 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementACOMPREHENSIVE LOOK AT THE MANAGEMENT OF SPARTINA IN WASHINGTON STATEK.C. MURPHY 1 ,W.BROWN 2 AND D. HEIMER 31 Aquatic Reserves Program Manager, Washington State Department <strong>of</strong> Natural Resources, P.O. Box 47027, Olympia, WA98504-7027; kyle.murphy@dnr.wa.gov2 Executive Coordinator, Washington Invasive Species Council, P.O. Box 40917, Olympia, WA 98504-09173 Noxious Weed Program Coordinator, Washington State Department <strong>of</strong> Fish and Wildlife, 4516 N 28 th , Tacoma, WA 98407Washington State has been fighting <strong>the</strong> spread <strong>of</strong> invasive Spartina since <strong>the</strong> early 1990’s. Untilrecently <strong>the</strong> progress in Puget Sound was slow or, in <strong>the</strong> case <strong>of</strong> Willapa Bay, non-existent.However, with appropriate funding, increased community, agency and legislative support, improvedtools, and better cooperation between <strong>the</strong> entities involved, <strong>the</strong> eradication in Puget Sound isprogressing at a rapid pace, and <strong>the</strong> tide is finally being turned in Willapa Bay.This paper discusses <strong>the</strong> challenges that have led to <strong>the</strong> current success in Washington State from anon-<strong>the</strong>-ground management prospective. Elements contributing to that success range from choosing<strong>the</strong> correct control tools to <strong>the</strong> importance <strong>of</strong> cultivating community support and cooperationnecessary for a successful program, regardless <strong>of</strong> <strong>the</strong> size <strong>of</strong> <strong>the</strong> infestation that is being treated.The lessons conveyed in this paper are intended to help o<strong>the</strong>rs confronting infestations <strong>of</strong> non-nativeSpartina and o<strong>the</strong>r invasive species to make <strong>the</strong> most <strong>of</strong> <strong>the</strong>ir available resources. Not repeating <strong>the</strong>same learning processes and avoiding <strong>the</strong> mistakes that have already been made can save resourcemanagers and field coordinators precious time and money and help to build a successful program toaddress invasive species problems.Keywords: Invasive Spartina, Spartina alterniflora, Spartina anglica, Spartina patens, Spartinadensiflora, Willapa Bay, Puget Sound, Hood Canal, Integrated Pest ManagementINTRODUCTIONSpartina species introduced to Washington State haveproven to be aggressive noxious weeds that severely disrupt<strong>the</strong> ecosystems <strong>of</strong> native saltwater estuaries. Theyoutcompete native vegetation and convert mudflats intomonotypic Spartina meadows. This is <strong>of</strong> great concernbecause estuaries serve as an important rearing area fornumerous fish species, important breeding, migration andwintering grounds for many migratory birds, waterfowl ando<strong>the</strong>r wildlife and provide a critical economic resource formany communities dependent on commercial fishing,mariculture, shipping and tourism (Patten and Stenvall2002). By 2002, in large areas <strong>of</strong> Willapa Bay, <strong>the</strong> area with<strong>the</strong> largest cordgrass infestation in <strong>the</strong> state, Spartinaalterniflora had reduced <strong>the</strong> available foraging time forshorebirds by 50% <strong>of</strong> winter daylight hours because <strong>of</strong> <strong>the</strong>reduction in mudflat acreage (Jaques 2002).In Washington, federal, state and local goverments,tribal entities, non-pr<strong>of</strong>it organizations, business interests,local universities and private citizens are working toge<strong>the</strong>r tocombat <strong>the</strong> invasion. This effort is becoming moresuccessful each year with reduction <strong>of</strong> invasive Spartinastatewide over <strong>the</strong> past two years. This report outlines <strong>the</strong>different components that have been instrumental in leading<strong>the</strong> Spartina eradication effort towards success, includingprogram support, planning and coordination, adaptivemanagement, and <strong>the</strong> evolution <strong>of</strong> <strong>the</strong> various tools used foreradication efforts.EXTENT OF INFESTATIONWashington State has one <strong>of</strong> <strong>the</strong> largest invasions <strong>of</strong>non-native Spartina species on <strong>the</strong> entire U.S. west coast(Patten and Stenvall 2002). The invasion is comprised <strong>of</strong>four species, Spartina alterniflora, Spartina anglica,Spartina patens and Spartina densiflora, spread throughoutfour main waterbodies, Willapa Bay, Grays Harbor, PugetSound and Hood Canal. Willapa Bay is known to containonly S. alterniflora; Grays Harbor contains both S.alterniflora and S. densiflora. Hood Canal is known tocontain S. alterniflora, S. anglica and S. patens, while PugetSound has infestations <strong>of</strong> S. alterniflora, S. anglica and S.densiflora.The current size <strong>of</strong> <strong>the</strong> infestation in Washington isapproximately 3,035 solid hectares (ha) (7,500 acres [ac]),affecting more than 7,300 ha (18,000 ac) <strong>of</strong> <strong>the</strong> intertidalmarine environment (Murphy 2004). In Puget Sound andHood Canal <strong>the</strong>re are 95 sites with current or historicalinfestations (Fig. 1). The infestations cover approximately260 solid ha (645 ac). The infestation in Willapa Bay coversmore than 2,800 solid ha (7,000 ac) and encompases almost<strong>the</strong> entire shoreline <strong>of</strong> <strong>the</strong> bay (Fig. 2).- 223 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaFig. 1. Map showing Spartina invasion sites found within Puget Sound inWashington State, USA.PROGRAM SUPPORTSupport for <strong>the</strong> Spartina eradication program inWashington is vital for success and can be categorized intothree types: community, agency, and legislative.Community support for <strong>the</strong> program has come fromprivate landowners, most <strong>of</strong> whom own property directlyaffected by Spartina or situated adjacent to infestations, butalso includes local industries, sportsmen, tribal entities withland impacted by Spartina invasions, environmental groups,and local universities. At <strong>the</strong> outset <strong>of</strong> <strong>the</strong> eradicationprogram in <strong>the</strong> early 1990s <strong>the</strong> biggest advocates for <strong>the</strong>program were those who depended on, managed, or ownedproperty in <strong>the</strong> intertidal environment impacted by Spartina.During <strong>the</strong> early stages <strong>of</strong> <strong>the</strong> infestation, communitysupport played a critical role in bringing <strong>the</strong> problem to <strong>the</strong>attention <strong>of</strong> both <strong>the</strong> goverment agencies who were in aposition to acknowledge and act on <strong>the</strong> problem and locallawmakers who were in a position to mandate and fundappropriate actions.Support for <strong>the</strong> program also has come from <strong>the</strong> variousfederal, state and local government agencies that must dealwith <strong>the</strong> Spartina invasion ei<strong>the</strong>r as part <strong>of</strong> <strong>the</strong>ir landFig. 2. Map showing Spartina invasion sites found within Willapa Bay inWashington State, USA. Infestation data provided by Washington StateDepartment <strong>of</strong> Natural Resources.management responsibilities or through <strong>the</strong>ir various legalmandates. These agencies include <strong>the</strong> U.S. Fish and WildlifeService, which manages Willapa National Wildlife Refugein Willapa Bay, as well as <strong>the</strong> Washington StateDepartments <strong>of</strong> Fish and Wildlife, Natural Resources,Ecology and State Parks. County noxious weed controlboards are also involved with and support <strong>the</strong> Spartinaeradication program as a part <strong>of</strong> <strong>the</strong>ir legislative mandates.Finally, <strong>the</strong> Washington State Department <strong>of</strong> Agriculture ismandated by law to lead <strong>the</strong> statewide eradication effort.In <strong>the</strong> beginning stages <strong>of</strong> <strong>the</strong> eradication program, <strong>the</strong>early 1990s, some higher-level individuals within keyagencies were reluctant to recognize Spartina as a threat.However, after field personnel familiar with <strong>the</strong> problemeducated <strong>the</strong>m on <strong>the</strong> pertinent issues, and with <strong>the</strong> passage<strong>of</strong> state legislation, <strong>the</strong>se agencies became fully supportive<strong>of</strong> <strong>the</strong> effort. The most important step in this educationprocess was <strong>the</strong> continued efforts by <strong>the</strong> field personnel andbiologists working on <strong>the</strong> Spartina problem in educating<strong>the</strong>ir pr<strong>of</strong>essional peers, making <strong>the</strong>m aware <strong>of</strong> Spartina’sthreat to <strong>the</strong> larger ecosystem and viability <strong>of</strong> <strong>the</strong> estuaries.- 224 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementAno<strong>the</strong>r aspect <strong>of</strong> agency support is interagencycooperation. The ability <strong>of</strong> each agency to work closely wi<strong>the</strong>ach o<strong>the</strong>r, share equipment, resources and personnel hasalso been a very important part <strong>of</strong> <strong>the</strong> eradication program’ssuccess. The various partners in Washington have signed <strong>of</strong>fon a Memorandum <strong>of</strong> Understanding (MOU) that helps t<strong>of</strong>oster this cooperative process.Legislative support has also been extremely importantin helping to motivate <strong>the</strong> governmental agencies to makeSpartina eradication a priority. Most importantly, strongpolitical support from both local and federal politicians hasensured <strong>the</strong> program <strong>of</strong> adequate funding. Such support mustbe ongoing to ensure that adequate funding remainsavailable.PLANNING AND COORDINATIONThe Washington State Department <strong>of</strong> Agriculture(WSDA) has been identified by law, RCW 17.26 (RevisedCode <strong>of</strong> Washington), as <strong>the</strong> lead agency <strong>of</strong> <strong>the</strong> Spartinaeradication program. Lead agency responsibilities includeensuring that <strong>the</strong> proper permits are in place, producingyearly progress reports for <strong>the</strong> Washington State Legislature,maintaining eradication program records, monitoring waterquality in relation to herbicide applications, providing publiceducational information and developing management plansin cooperation with <strong>the</strong> o<strong>the</strong>r entities. Having one agency in<strong>the</strong> state manage <strong>the</strong>se responsibilities has allowed <strong>the</strong> o<strong>the</strong>rentities to prioritize on-<strong>the</strong>-ground eradication efforts.However, no legal authority was given to WSDA requiringpartners to follow specific WSDA plans or directives. Thisled to beneficial innovation in control methods as variouspartners tried control tools which <strong>the</strong>y believed aided ineradication and built friendly competition between <strong>the</strong>agencies. However, <strong>the</strong> absence <strong>of</strong> lead-agency authorityalso sometimes resulted in unilateral decisions and a lack <strong>of</strong>budgetary transparency.Spartina eradication efforts in Washington are plannedcooperatively. In Puget Sound, <strong>the</strong> North Puget SoundSpartina Eradication Task Force, made up <strong>of</strong> representativesfrom local and state agencies, tribal entities, and non-pr<strong>of</strong>itorganizations meets throughout <strong>the</strong> year to developmanagement plans and discuss <strong>the</strong> successes and failures <strong>of</strong>different approaches. In Willapa Bay, a technical groupmade up <strong>of</strong> field coordinators and biologists develops yearlymanagement plans and brings <strong>the</strong>m before an advisorycommittee for review and critique. While <strong>the</strong>re has notalways been agreement on all <strong>the</strong> specifics contained in <strong>the</strong>management plans, <strong>the</strong>re has always been consensus on <strong>the</strong>overall goal <strong>of</strong> eradicating Spartina. With this commonground, <strong>the</strong> partners continue to return to <strong>the</strong> planning table.ADAPTIVE MANAGEMENTRCW 17.15.020 directs state agencies in Washington toimplement integrated pest management (IPM) practiceswhen carrying out <strong>the</strong> agency's or institution's duties relatedto pest control. IPM effectively extends <strong>the</strong> operating seasonand allows techniques to be applied when and where <strong>the</strong>ywork best (DOI 1996).Adaptive management is a key element <strong>of</strong> a successfulIPM program. It allows management policies and practicesto continually improve by learning from <strong>the</strong> interimoutcomes <strong>of</strong> <strong>the</strong> eradication program.http://www.for.gov.bc.ca/hfp/amhome/Admin/index.htmWashington’s Spartina eradication program contains fiveadaptive management components: pre-treatmentmonitoring, implementing eradication practices on <strong>the</strong>proper scale, post-treatment monitoring and evaluation,adjusting future treatment based on evaluation, andcontrolling in a consistent manner.Pre-treatment monitoring sets a baseline for comparison<strong>of</strong> post-treatment monitoring results. During <strong>the</strong> spring <strong>of</strong>2002, <strong>the</strong> partners involved in control work developed acomprehensive monitoring plan that included protocols forpre-treatment monitoring. This ensured an accurate baseline<strong>of</strong> Spartina stem counts.The next component <strong>of</strong> adaptive management isimplementing eradication practices on an appropriate scale.It is important to choose <strong>the</strong> proper tools to use and set agoal <strong>of</strong> treating <strong>the</strong> entire site. For example, using backpacksprayers to completely treat a 200-hectare meadow wouldnot be appropriate, whereas using a helicopter applicationwould be time- and cost-efficient. Post-treatment monitoringhelped determine that using <strong>the</strong> proper tool(s) forretreatment <strong>of</strong> a large Spartina infestation takes a minimum<strong>of</strong> three years. Ensuring that <strong>the</strong> entire site receivestreatment is <strong>of</strong>ten dependent on funding, so matchingappropriate funding to <strong>the</strong> infestation size is essential.Through field experience and research, it has become clearthat consistency <strong>of</strong> treatment is as important as <strong>the</strong> methodchosen for control. Early failures in Washington, in part,resulted from not treating <strong>the</strong> entire site, leaving viableplants to re-seed and spread back into <strong>the</strong> treated portion <strong>of</strong><strong>the</strong> site.Once eradication practices are implemented, posttreatmentmonitoring is conducted. The monitoring plandeveloped in 2002 ensured that enough samples werecollected to allow for accurate comparison with <strong>the</strong> pretreatmentsampling. Through this comparison, managers andfield coordinators were able to evaluate <strong>the</strong> effectiveness <strong>of</strong><strong>the</strong> eradication practices employed.Adjusting <strong>the</strong> treatments based on <strong>the</strong> completedevaluation becomes <strong>the</strong> next step in <strong>the</strong> process. Forinstance, crushing Spartina was found to be ineffective when<strong>the</strong> substrate was too firm and, as a result, <strong>the</strong> site inquestion was sprayed <strong>the</strong> next season. With adequate preandpost-treatment monitoring and evaluation, managers andfield coordinators get a good picture <strong>of</strong> <strong>the</strong> success or failure- 225 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaFig. 3. Field crew member mowing Spartina anglica with hand-held brushcutters.Fig. 4. Field crew member treating Spartina patens with backpack herbicidesprayer.<strong>of</strong> a treatment method. This allows for <strong>the</strong> adjustment <strong>of</strong>treatment methods for better control.The most imporant component <strong>of</strong> adaptive managementin Spartina eradication is consistent control from year toyear. Research has shown that invasive Spartina is highlyresilient and requires multiple years <strong>of</strong> consistent treatmentto cause substantial declines. In some instances inconsistentcontrol, specifically, when two treatment seasons werefollowed by a year without treatment, resulted in a 100% to500% increase in cover and number <strong>of</strong> tillers (Reeder andHacker 2004). This underscores <strong>the</strong> importance <strong>of</strong>consistently treating a chosen site year after year to achieveeventual reduction and eradication.EVOLUTION OF TOOLSOver <strong>the</strong> more than ten years <strong>of</strong> invasive Spartinaeradication efforts in Washington State, <strong>the</strong> treatment toolshave evolved and become more effective. In <strong>the</strong> beginning,crews treated infestations <strong>of</strong> all sizes with small hand heldbrushcutters (Fig. 3) and backpack herbicide sprayers (Fig.4). While <strong>the</strong>se tools are effective on small infestations andare still used to this day, <strong>the</strong>y are simply too slow foreffectively treating anything greater than a few hectares. Theeradication program in Washington now employs helicoptersand large, tracked amphibious machines to treat largemeadows, and airboats with high pressure herbicide spraysytems to treat scattered clones and re-growth. Thehelicopters, applying herbicides at low carrier volumes, 97lliters per hectare (l/ha) (10 gallons per acre [10 gpa]), cantreat more than 160 ha (400 ac) per day. While slower than<strong>the</strong> helicopters, <strong>the</strong> tracked machines with broadcastapplication systems are still able to treat upwards <strong>of</strong> 16 ha(40 ac) per day. Airboats may only be able to treat 3.2 ha (8ac) per day at most, but <strong>the</strong>y are able to manuever in <strong>the</strong>estuaries and on mudflats, allowing for rapid transportbetween scattered infestations and produce virtually noenvironmental footprint.Having <strong>the</strong> ability to treat large infestations quickly andeffectively has increased <strong>the</strong> success <strong>of</strong> <strong>the</strong> eradicationprogram in Washington .CONCLUSIONOver <strong>the</strong> past two years, <strong>the</strong> eradication program inWashington State has become more efficient and effective.In 2003, a total <strong>of</strong> 2,700 solid ha (6,695 ac) out <strong>of</strong>approximately 3,400 solid ha (8,500 ac) were treated(Murphy 2003). These treatments resulted in <strong>the</strong> first overallreduction in <strong>the</strong> statewide infestation. In 2004, a total <strong>of</strong>2,500 solid ha (6,200 ac) out <strong>of</strong> an estimated 3,000 solid ha(7,500 ac) were treated statewide (Murphy 2004). Managersare confident that <strong>the</strong> 2004 treatments will also result in aFig. 5. Helicopter herbicide application, Willapa Bay, Washington, USA- 226 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementFig. 6. Tracked amphibious machine with ground broadcast herbicide applicationequipment, Willapa Bay, Washington, USA.Fig. 7. Airboat with high-pressure spray equipment, Willapa Bay, Washington,USA.continued overall decline <strong>of</strong> Washington’s Spartinainfestation.Many things have contributed to <strong>the</strong> current success <strong>of</strong><strong>the</strong> program. Consistent control made possible with adequatefunding led to effective treatment. Washington’smanagement plans called for a multi-year commitment totreatment sites, controlling a site until eradication has beenachieved. Permitting consolidation also contributed to <strong>the</strong>program’s success. WSDA is granted an National PollutantDischarge Elimination System (NPDES) permit that allowsfor <strong>the</strong> chemical treatment <strong>of</strong> Spartina. As <strong>the</strong> permit holder,WSDA extends coverage to o<strong>the</strong>r entities to treat Spartinainfestations, while WSDA retains <strong>the</strong> responsibily formonitoring and reporting. This allows <strong>the</strong> o<strong>the</strong>r entities t<strong>of</strong>ocus <strong>the</strong>ir resources towards on-<strong>the</strong>-ground control, ra<strong>the</strong>rthan on manuevering through <strong>the</strong> bureaucratic permittingprocess.Finally, through a long and painful learning curve,managers and field coordinators have discovered what worksbest to control Spartina infestations within a complicated,and sometimes controversial, environmental, social andpolitical context. Garnering public support, havinglegislation in place to provide incentive and funding,working collaboratively with o<strong>the</strong>r agencies, having accessto a wide breadth <strong>of</strong> tools, and knowing where and how tobest use those tools have all contributed to <strong>the</strong> currentsuccess <strong>of</strong> <strong>the</strong> program. While <strong>the</strong>re is still much progress tobe made, <strong>the</strong> partners involved are confident that <strong>the</strong>ireradication goal will be achieved. The effort in Washingtonwill continue searching for better tools and equipment toincrease efficacy and efficiency, while minimizing costs.2010 UPDATEAt <strong>the</strong> time <strong>of</strong> this article’s publication, <strong>the</strong> cooperativeSpartina eradication effort in Washington State hascontinued to yield great success in effectively reducing <strong>the</strong>infestation statewide. During 2009 <strong>the</strong> partners involved in<strong>the</strong> eradication effort treated approximately 110 solid acres<strong>of</strong> Spartina. WSDA now estimates that, as a result <strong>of</strong> <strong>the</strong>2009 effort, <strong>the</strong> infestation statewide has been reduced toonly 40 solid acres (Phillips pers. com. 2010). Thisrepresents a 99.5% overall reduction from <strong>the</strong> historic high<strong>of</strong> 9,260 solid acres estimated in 2003.This continued success provides evidence that <strong>the</strong>management effort, originally detailed in this article in 2004,was and continues to be an effective approach. Since <strong>the</strong>2004 <strong>International</strong> Invasive Spartina <strong>Conference</strong>, <strong>the</strong> effortin Washington has continued to seek improvements in <strong>the</strong>tools and approaches used to eradicate Spartina. Imazapyr,first aerially applied to 2000 solid acres during <strong>the</strong> 2004season (Murphy 2005), proved to be a huge success and abig reason for <strong>the</strong> success <strong>of</strong> <strong>the</strong> overall Washington Stateeffort. Ano<strong>the</strong>r obvious reason for this success was <strong>the</strong>continued program support and funding.As major reductions were made in <strong>the</strong> overall size <strong>of</strong> <strong>the</strong>infestation, <strong>the</strong> remaining Spartina was scattered throughouta large area. This required <strong>the</strong> partners to modify <strong>the</strong> on-<strong>the</strong>groundapproach. The effort went from focusing on largescaleaerial and ground broadcast applications, to smallscale,targeted treatments <strong>of</strong> scattered infestations. Thishighlights <strong>the</strong> need to constantly re-evaluate <strong>the</strong> program toensure that <strong>the</strong> most effective and efficient tools andapproaches are used to continue <strong>the</strong> success <strong>of</strong> <strong>the</strong>cooperative Spartina eradication effort.REFERENCESBritish Columbia Ministry <strong>of</strong> Forests & Range. 2008. Definitions<strong>of</strong> Adaptive Management. Posted athttp://www.for.gov.bc.ca/hfp/amhome/Admin/index.htm.- 227 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaJaques, D. 2002. Shorebird status and effects <strong>of</strong> Spartina alternifloraat Willapa National Wildlife Refuge. Progress Reportto <strong>the</strong> WNWR.Murphy, K. 2003. Report to <strong>the</strong> Legislature: Progress <strong>of</strong> <strong>the</strong> 2003Spartina Eradication Program. WSDA Pub # AGR PUB 850-110 (N/1/04).Murphy, K. 2004. Report to <strong>the</strong> Legislature, Progress <strong>of</strong> <strong>the</strong> 2004Spartina Eradication Program. WSDA Pub # AGR PUB 850-132 (N/1/05).Murphy, K. 2005. Report to <strong>the</strong> Legislature, Progress <strong>of</strong> <strong>the</strong> 2005Spartina Eradication Program. WSDA Pub # AGR PUB 850-151 (N/1/06).Patten, K. and C. Stenvall. 2002. Control <strong>of</strong> Smooth Cordgrass(Spartina alterniflora): A Comparison between mechanical andchemical control methods for efficacy, cost and aquatic toxicity.In: <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> 11th Int. Conf. on Aquatic Invasive Species,2001,Pembroke, Ontario, Canada. 25-28 February 2002,Alexandria, VA. The Pr<strong>of</strong>essional Edge, Pembroke, ON. Canada:340-350.Reeder, T.G., S.D. Hacker. 2004. Factors Contributing to <strong>the</strong> Removal<strong>of</strong> a Marine Grass Invader (Spartina anglica) and SubsequentPotential for Habitat Restoration. Estuaries 27: 244-253.U.S. Dept <strong>of</strong> Interior, Fish and Wildlife Service, Willapa NationalWildlife Refuge. 1996. Iterim Environmental Assessment: Control<strong>of</strong> Smooth Cordgrass (Spartina alterniflora) on Willapa NationalWildlife Refuge in 1996: 20-21.- 228 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementWHY DID IT TAKE SO LONG TO TURN THINGS AROUND IN WILLAPA BAY?THE HUMAN SIDE OF THE SPARTINA INVASIONM. WECKERP.O. Box 160, Naselle, WA 98638; mwecker@wwest.netOyster growers first noticed clones <strong>of</strong> exotic cordgrass growing in Willapa Bay in <strong>the</strong> 1950s andbrought <strong>the</strong>ir concerns about <strong>the</strong> alien grass to <strong>the</strong> attention <strong>of</strong> <strong>the</strong> staff <strong>of</strong> <strong>the</strong> Willapa Refuge. Atthat time, <strong>the</strong> infestation involved no more than a handful <strong>of</strong> clones. Spartina now infests over15,000 acres <strong>of</strong> mudflats in Willapa Bay, seriously degrading one <strong>of</strong> <strong>the</strong> most ecologicallyproductive and o<strong>the</strong>rwise healthy estuarine areas along <strong>the</strong> entire Pacific Coast. The human storythat parallels <strong>the</strong> expansion <strong>of</strong> Spartina reveals more than just <strong>the</strong> inevitable physical, logistical, andbiological challenges <strong>of</strong> responding to an aggressive invasive species. Many <strong>of</strong> our environmentalorganizations were unprepared for this problem and reacted in ways that made <strong>the</strong> problem farworse. Our government structures were also not well designed to meet <strong>the</strong> challenges <strong>of</strong> a highlydamaging but very natural process. This paper will present <strong>the</strong> views <strong>of</strong> one person who has beenboth an observer from <strong>the</strong> sidelines and an involved participant in <strong>the</strong> Willapa Bay Spartina programfor <strong>the</strong> past 13 years. Since 1995, Miranda Wecker has served as leader <strong>of</strong> <strong>the</strong> marine researchprogram <strong>of</strong> <strong>the</strong> University <strong>of</strong> Washington’s Olympic Natural Resources Center. In that capacity, shesupervised research on <strong>the</strong> feasibility <strong>of</strong> biological control <strong>of</strong> Spartina as well as spatial analysis <strong>of</strong><strong>the</strong> weed’s spread. She also collaborated with local leaders in <strong>the</strong> development and implementation<strong>of</strong> a long-term strategy for eradication based on <strong>the</strong> comparison <strong>of</strong> alternative control options usinggeographic information system (GIS) tools. It is important to acknowledge errors that have beenmade and draw lessons from <strong>the</strong>m in order to assure that <strong>the</strong> same mistakes are not repeated whenwe face <strong>the</strong> next invasion.Keywords: management lessons, long-term strategyINTRODUCTIONMy task is to summarize <strong>the</strong> key lessons that I havelearned during <strong>the</strong> more than 15 years that we have beenfighting <strong>the</strong> spread <strong>of</strong> Spartina in Willapa Bay. My aim isnot to insult or embarrass anyone. My hope is that an honestappraisal <strong>of</strong> our problems will allow us to makeimprovements. I also hope our counterparts in San FranciscoBay will find it helpful to hear my opinions about <strong>the</strong> lessonswe have learned.What is at stake, it seems to me, is something muchmore than winning <strong>the</strong> battle against Spartina. I believe that<strong>the</strong> mismanagement <strong>of</strong> <strong>the</strong> Spartina control effort in WillapaBay has damaged local attitudes towards government and itsabilities to solve societal problems. During <strong>the</strong> past tenyears, Willapa residents have watched Spartina continue tospread at an alarming rate despite <strong>the</strong> expenditure <strong>of</strong> millions<strong>of</strong> dollars <strong>of</strong> public funds. The ongoing failure to beat <strong>the</strong>Spartina invasion has added ano<strong>the</strong>r chapter to a longhistory <strong>of</strong> rural skepticism towards government programs.Poor performance creates a double bind. If governmentagencies perform poorly, vocal criticism puts funding at risk.Without such criticism, poor performance is allowed tocontinue unquestioned. Without reform, <strong>the</strong> program wouldexhaust <strong>the</strong> public will and funds available to deal with <strong>the</strong>problem and fail anyway.In <strong>the</strong> following sections, I will present my thoughtsregarding what we have learned in fighting Spartina inWillapa Bay. These comments are intended to be subjectivepersonal conclusions based my experiences in <strong>the</strong> humandimensions <strong>of</strong> <strong>the</strong> Spartina battle. For <strong>the</strong> battle was not justabout deploying <strong>the</strong> physical resources to kill 20,000 acres<strong>of</strong> an invasive weed. Success required assembly <strong>of</strong> logistical,legal, financial, and political resources. Success depended onorganizing an array <strong>of</strong> people and institutions to act in aconcerted and effective manner.LESSON 1: NEVER TOO MANY MAPS, BUT THEY CAN BETOO EXPENSIVEA biological invasion is an inherently spatial problem.Before you can formulate a sensible management plan, youneed a good idea <strong>of</strong> <strong>the</strong> scale <strong>of</strong> <strong>the</strong> problem. You need toknow where <strong>the</strong> enemy plants are. If your aim is to assist andfacilitate management, you have to be concerned about <strong>the</strong>costs and <strong>the</strong> time delays associated with <strong>the</strong> variousmapping options. Rarely does a weed control program havesurplus funding. To see that most <strong>of</strong> <strong>the</strong> money goes towarderadication activities in <strong>the</strong> field, you must avoid overlyexpensive and labor intensive methods <strong>of</strong> mapping. GIS- 229 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaanalytic services are <strong>of</strong>ten very expensive. Most agencieshave geographic information system (GIS) expertise, but<strong>the</strong>y are assigned to <strong>the</strong> highest priority projects for <strong>the</strong>agency. The Spartina infestation was not considered to besufficiently important to justify dedication <strong>of</strong> internal GISservices by any <strong>of</strong> Washington’s state agencies. Fortunately,precision is not critical. You really don’t need maps andanalysis worthy <strong>of</strong> scientific publications. For that reason, Ibelieve that <strong>the</strong> use <strong>of</strong> standard-accuracy Global PositioningSystem (GPS) technology is an acceptable way to map formanagement purposes. We found that a basic survey <strong>of</strong>Willapa Bay could be done in seven-day period bydelineating <strong>the</strong> outer edge <strong>of</strong> <strong>the</strong> infestation using a GPSunit. GPS surveys are inherently ground-tru<strong>the</strong>d and georeferenced.The data from a GPS survey can be easilyexported and displayed as a GIS layer. Our total cost wasless than $2,000. Infrared aerial photography and itsinterpretation, by comparison, was so expensive thatagencies could only afford to have it done every third year.Image processing and interpretation required years <strong>of</strong> work.Katy Zaremba presented information earlier in this Spartinaconference indicating that aerial photography misses a greatmany plants. While aerial photography may have academicvalue, I believe that we have learned that it is not <strong>the</strong> optimalway to serve day-to-day management needs.In terms <strong>of</strong> quantity, however, I would say that it wouldbe difficult to have too many maps. During <strong>the</strong> first fiveyears <strong>of</strong> <strong>the</strong> Willapa Spartina program, we had too few. It isbest to have surveys done at <strong>the</strong> beginning as well as at <strong>the</strong>end <strong>of</strong> each season. I would recommend that you have GISand GPS services at your beck and call. You’ll need maps torespond quickly to new challenges as <strong>the</strong>y arise.For example, in 2002, Dr. Kim Patten <strong>of</strong> WashingtonState University concluded that herbicide applications wouldonly deliver optimal kill rates if a minimum <strong>of</strong> 12 hours <strong>of</strong>dry time followed treatment. Our challenge was to determinewhe<strong>the</strong>r Spartina in <strong>the</strong> lowest elevations <strong>of</strong> Willapa Baywere exposed for at least 12 hours during any <strong>of</strong> <strong>the</strong> tidesexpected in <strong>the</strong> upcoming season. Using <strong>the</strong> new LiDARbathymetry layer generated through <strong>the</strong> University <strong>of</strong>California at Davis and <strong>the</strong> NOAA Coastal Services Centerproject, Teresa Alcock and Keven Bennett <strong>of</strong> <strong>the</strong> University<strong>of</strong> Washington’s Olympic Natural Resources Center(ONRC) formulated a method for presenting spatiallyexplicittidal predictions. Ano<strong>the</strong>r example <strong>of</strong> a spatialinformation challenge was <strong>the</strong> identification <strong>of</strong> owners <strong>of</strong>infested parcels. In many areas <strong>of</strong> Willapa Bay, <strong>the</strong> pattern<strong>of</strong> ownership is extremely complex with hundreds <strong>of</strong> plotsshaped like jigsaw pieces. Before treatment could take place,landowners had to be contacted and permission forms had tobe completed. ONRC staff generated maps with ownershipsand Spartina coverage displayed.In addition to its use as a monitoring and planning tool,GIS also played an important role in communications andpublic relations. An animated movie clip we generatedgraphically displayed how <strong>the</strong> Spartina infestation wouldexpand throughout Willapa Bay. This short movie proved tobe one <strong>of</strong> <strong>the</strong> most effective educational tools we created tocommunicate <strong>the</strong> risks. In addition, GIS played aninvaluable role in generating concrete images to helpfacilitate discussion <strong>of</strong> alternative strategies. Because we hada GIS staff at <strong>the</strong> ready, we were able to rapidly react to <strong>the</strong>emerging information needs as <strong>the</strong>y arose.LESSON 2: TEACH THE PUBLIC WELLIn Willapa Bay, it was obvious from <strong>the</strong> beginning thatbroad-based public support for <strong>the</strong> control program wasnecessary. Roughly half <strong>the</strong> infestation occurred on privatelands. The full costs <strong>of</strong> control far exceeded <strong>the</strong> local abilityto pay, so help from <strong>the</strong> state and federal government wasneeded. The only practical option was a concerted andhighly visible effort by public agencies. So it was obviousthat <strong>the</strong> Spartina control program could not succeed without<strong>the</strong> consent, understanding and au<strong>the</strong>ntic support <strong>of</strong> <strong>the</strong>citizenry. Controversy would also have complicated <strong>the</strong>ability <strong>of</strong> lawmakers to secure <strong>the</strong> millions <strong>of</strong> dollars <strong>of</strong>funding needed to sustain <strong>the</strong> eradication effort to itsconclusion. The public relations challenge was successfullymet by focussing time and energy on educating andinvolving local residents; this kept <strong>the</strong> Spartina issue at <strong>the</strong>top <strong>of</strong> <strong>the</strong> priorities list. We also provided numerousopportunities for citizens to track control program progressand comment on its evolution. Because for years <strong>the</strong>Spartina control effort was not making progress, many <strong>of</strong>our public meetings provided <strong>the</strong> opportunity for expressions<strong>of</strong> harsh criticism. In response, agency personnel becameexcessively defensive.To overcome <strong>the</strong> climate <strong>of</strong> pessimism that surrounded<strong>the</strong> Spartina issue, we resorted to some ra<strong>the</strong>runconventional approaches to public education. Weunderstood that <strong>the</strong> standard environmental jargon andtactics that work well in receptive urban populations wouldonly trigger sarcasm and hostile reactions in our area. Onenew way in which we delivered detailed information toaudiences was through educational placemats. We designedplacemats presenting key information along with beautifulimages and distributed <strong>the</strong>m free <strong>of</strong> charge to local c<strong>of</strong>feeshops around Pacific County. Customers in restaurantsproved to be very willing to read whatever was available as<strong>the</strong>y waited for <strong>the</strong>ir orders. Copies <strong>of</strong> <strong>the</strong> placemats werealso carried away and shared with o<strong>the</strong>rs. Ano<strong>the</strong>r successfuldevice was our series <strong>of</strong> “science for <strong>the</strong> people” meetings inwhich research scientists presented progress updates on <strong>the</strong>irwork to <strong>the</strong> public. One <strong>of</strong> <strong>the</strong> more entertaining <strong>of</strong> thoseevenings occurred when Dr. Donald Strong <strong>of</strong> <strong>the</strong> University- 230 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and Management<strong>of</strong> California at Davis engaged in a lively debate over <strong>the</strong>feasibility <strong>of</strong> biological control with a senior <strong>of</strong>ficial <strong>of</strong> <strong>the</strong>Washington State Department <strong>of</strong> Agriculture. Citizenslearned about <strong>the</strong> uncertainties common to research and <strong>the</strong>ways in which enterprising scientists respond to <strong>the</strong>m. Theywere able to draw <strong>the</strong>ir own conclusions after hearingfirsthand <strong>the</strong> various sides <strong>of</strong> a debate.LESSON 3: TIME IS OF THE ESSENCEWhen <strong>the</strong> Willapa Bay Spartina invasion becameevident as a serious threat, <strong>the</strong> outside world was not readyto see it as such. In <strong>the</strong> late 1980s, local civic leaders becameconvinced that <strong>the</strong> Spartina invasion was fast becoming anecological disaster. Unfortunately, <strong>the</strong> majority <strong>of</strong>environmental pr<strong>of</strong>essionals across <strong>the</strong> country had not yetrecognized <strong>the</strong> extent <strong>of</strong> <strong>the</strong> threat posed by invasive species.Within <strong>the</strong> ranks <strong>of</strong> <strong>the</strong> many public agencies withjurisdiction, pr<strong>of</strong>essional staff held varying views. Conflictsover <strong>the</strong> acceptability <strong>of</strong> <strong>the</strong> impacts tied to controltreatments made a prompt response to <strong>the</strong> spread <strong>of</strong> Spartinaimpossible.Responses were delayed because <strong>of</strong> <strong>the</strong> availability <strong>of</strong>legal mechanisms for individuals who held contrarian viewsto block agency action despite support for <strong>the</strong> program by<strong>the</strong> vast majority <strong>of</strong> <strong>the</strong> public. In Willapa Bay, suchchallenges deflected energies and added to <strong>the</strong> costs <strong>of</strong> <strong>the</strong>management effort. Contrarian activists made repeatedefforts to derail <strong>the</strong> Spartina control program in <strong>the</strong> name <strong>of</strong>environmental concern. Their “monkeywrenching” tacticswere ultimately defeated by <strong>the</strong> active participation <strong>of</strong> <strong>the</strong>Washington Chapter <strong>of</strong> <strong>the</strong> Nature Conservancy andAudubon Society Washington. It is vital to enlist <strong>the</strong> supportand involvement <strong>of</strong> credible pr<strong>of</strong>essional environmentalorganizations in order to prevent o<strong>the</strong>rs from framing <strong>the</strong>issue in <strong>the</strong>ir own terms. During that early period <strong>of</strong> delay,an aggressive control program would have solved <strong>the</strong>problem with <strong>the</strong> least costs and <strong>the</strong> least ecologicalimpacts. Instead, <strong>the</strong> invasion was allowed to accelerate.LESSON 4: SOMEBODY HAS TO BE IN CHARGEIn Willapa Bay, three state agencies and one federalagency are directly engaged in field operations. They areindependent <strong>of</strong> one ano<strong>the</strong>r, have separate lands, and pursuedifferent missions. In 1995, <strong>the</strong> Washington Legislatureenacted RCW 17.26—a statute calling for coordination <strong>of</strong>Spartina control efforts and strong leadership under a singlelead agency. The Legislature designated <strong>the</strong> WashingtonState Department <strong>of</strong> Agriculture (WSDA) as <strong>the</strong> leadagency, but also directed that each agency take responsibilityfor Spartina control on <strong>the</strong>ir own lands. WSDA hasinterpreted its lead role as comprised <strong>of</strong> two major functions:it convenes interagency meetings and drafts <strong>the</strong> annual reportto <strong>the</strong> Legislature. WSDA declined to interpret its assignmentexpansively, in o<strong>the</strong>r words to include all mission criticalleadership functions.Without a single agency in charge, it was impossible for<strong>the</strong> agencies to develop and implement a comprehensive andcoherent long-term strategy with an appropriate allocation <strong>of</strong>assignments and responsibilities. Without a single agency incharge, individual agencies were free to pursue <strong>the</strong>irindividual near-term interests ra<strong>the</strong>r than subordinate <strong>the</strong>m to<strong>the</strong> interests <strong>of</strong> <strong>the</strong> overall program. No one agency was in aposition to demand <strong>of</strong> o<strong>the</strong>rs a timetable for success or toassess <strong>the</strong> level <strong>of</strong> funding needed for overall success. No oneagency was held accountable for <strong>the</strong> success or failure <strong>of</strong> <strong>the</strong>program. No one agency was motivated to demand adherenceto an overall vision and coordinating scheme. As a result, <strong>the</strong>control effort in Willapa Bay drifted down a path that wasunnecessarily expensive and slow. Each agency was free topoint to <strong>the</strong> inadequacies <strong>of</strong> <strong>the</strong> program withoutresponsibility for <strong>the</strong> whole. From presentations during thisconference, it is obvious that <strong>the</strong> California CoastalConservancy’s Invasive Spartina Project has very competent,hard-working staff. They have generated a number <strong>of</strong>impressive documents that will go a long way toward gainingcooperation through persuasion. Our experience tells us thatcoordination will require more than persuasion. Legal andcoercive authority will probably be needed.LESSON 5: SELL THE NEED FOR SCIENCELooking back, I now think that one <strong>of</strong> our biggest errorswas that we did not argue strenuously enough for a seriousmonitoring and research effort in <strong>the</strong> early years. We neverpersuaded our political leadership <strong>of</strong> <strong>the</strong> importance <strong>of</strong>monitoring and research designed to inform <strong>the</strong> managementeffort. From <strong>the</strong> outset, we should have more adequatelycountered <strong>the</strong> strong antipathy towards spending dedicatedfunds on research. Public hostility towards research wasunderstandable: substantial sums had been spent exploring<strong>the</strong> non-target impacts <strong>of</strong> control tools that had beenapproved for years and used without problems in a widevariety <strong>of</strong> settings. Some <strong>of</strong> <strong>the</strong> scientists involved in earlyresearch were seen as biased. They were more concernedwith <strong>the</strong> short-term and minor effects <strong>of</strong> control tools thanwith <strong>the</strong> long-term ecological devastation wrought by <strong>the</strong>invasion itself. In at least one case, a scientist was politicallyaffiliated with groups opposed to <strong>the</strong> use <strong>of</strong> herbicides. Thisbiased him to propose a research plan designed to ignore <strong>the</strong>longterm implications <strong>of</strong> <strong>the</strong> Spartina invasion whilehighlighting <strong>the</strong> very transient benefits <strong>of</strong> Spartina increating edge habitat. O<strong>the</strong>r scientists participating in <strong>the</strong>project objected to <strong>the</strong> inherent bias and forced a redesign <strong>of</strong><strong>the</strong> workplan. In <strong>the</strong> end, because <strong>the</strong> results were not to <strong>the</strong>liking <strong>of</strong> <strong>the</strong> politically motivated scientist, he refused topublish <strong>the</strong> results. This intentional blurring <strong>of</strong> science andpolitics undermined trust in good science. It also led <strong>the</strong>- 231 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinapolitical leadership to restrict <strong>the</strong> use <strong>of</strong> funds. Agencieswere forbidden from using funds for research, even studiesthat would directly improve <strong>the</strong> efficiency and effectiveness<strong>of</strong> <strong>the</strong> control program. Ra<strong>the</strong>r than <strong>of</strong>ficially-sanctionedstudies, evaluations <strong>of</strong> <strong>the</strong> costs and benefits <strong>of</strong> controltechniques were done using grant funds cobbled toge<strong>the</strong>r.Control technique trials were not conducted to providereliable information on <strong>the</strong> efficacy <strong>of</strong> <strong>the</strong> tools. Alternativestrategies could not be compared. Good consistent recordswere not kept. No agency could compare <strong>the</strong>ir work witho<strong>the</strong>rs. The control program was conducted year after yearwithout a good understanding <strong>of</strong> how well <strong>the</strong> tools and <strong>the</strong>application strategies actually worked.The lack <strong>of</strong> objective quantitative information harmed<strong>the</strong> personal relations among both <strong>the</strong> allied agencies andlocal stakeholders as well. More and more frequently,interagency meetings provided <strong>the</strong> occasion for majorsquabbles over whe<strong>the</strong>r particular treatment techniques weresufficiently effective to continue <strong>the</strong>ir use. Withoutdefinitive studies, <strong>the</strong> subjectivity inherent in our argumentscould not be overcome. The tension and distrust generatedby <strong>the</strong>se irresolvable debates eventually led factions to stoptalking.If I had it to do over, I would have spent much moretime selling our politicians and <strong>the</strong> public on <strong>the</strong> need forgood rigorous science and monitoring. We should have triedto learn how to kill Spartina effectively and affordably at <strong>the</strong>very start. A rigorous comparison <strong>of</strong> <strong>the</strong> tools should havebeen conducted to determine <strong>the</strong> conditions for maximumefficiency <strong>of</strong> each. We should also have gotten an earlierstart on <strong>the</strong> feasibility <strong>of</strong> biological control, and startedlooking for alternative chemicals soon after hearing from ourfriends in Australia and New Zealand that glyphosate doesnot work well on Spartina.LESSON 6: THINK AHEAD AND SHARE YOUR VISIONMy experience has convinced me that long-termplanning is very necessary despite <strong>the</strong> discomfort andhumility you may feel in <strong>the</strong> face <strong>of</strong> many seemingly criticalunknowns. By taking a decent stab at planning, youcommunicate that you are serious about solving <strong>the</strong> problemin a logical and efficient manner. If all <strong>the</strong> facts are notavailable, assumptions will have to do for <strong>the</strong> time being.The concept <strong>of</strong> adaptive management is nothing more than<strong>the</strong> explicit promise to update your plan as you learn fromexperience. You are likely to attract <strong>the</strong> money you needonly if you have a compelling and well-defined plan <strong>of</strong>action. You are also likely to draw important allies if yourvision is clear and convincing. If you are asking citizens toaccept <strong>the</strong> use <strong>of</strong> chemical applications as a major part <strong>of</strong> <strong>the</strong>program, you had better have an endpoint in mind and beprepared to explain why it will take you that long to get<strong>the</strong>re. You need a flexible timetable. You need graphicimages showing how you can succeed with <strong>the</strong> resourcesyou have at hand. Then you need to take your vision forsuccess on <strong>the</strong> road and sell it to <strong>the</strong> key stakeholders.LESSON 7: CONSOLIDATED SYSTEMATIC PROGRESSOver <strong>the</strong> years, it became more and more apparent thatwe would make no progress unless we could eliminate largecontiguous blocks <strong>of</strong> Spartina in a systematic fashion.Ra<strong>the</strong>r than dispersing <strong>the</strong> control effort here and <strong>the</strong>rearound <strong>the</strong> Bay, we needed to consolidate our efforts. Tolimit re-infestation from nearby untreated areas, we adopteda strategy <strong>of</strong> working in consolidated blocks. Theoreticalweed management models advise that all outliers should betreated first in order to minimize <strong>the</strong> overall long-term costs.Logistical realities in Willapa Bay make <strong>the</strong> “outliers first”strategy untenable. Few tides are sufficiently low to allowtime for effective treatment <strong>of</strong> all outliers before o<strong>the</strong>r parts<strong>of</strong> <strong>the</strong> infestation are tackled. Most <strong>of</strong> <strong>the</strong> outliers were alsoon private lands, while an agency’s first priorities are <strong>the</strong>irown lands. Chemical applications <strong>of</strong> glyphosate required drytimes rarely available where <strong>the</strong> outliers occur. The timerequired and costs <strong>of</strong> moving heavy equipment dictate thatcrews limit hopping, skipping, and backtracking as much aspossible. It is also far easier to keep track <strong>of</strong> where you haveleft <strong>of</strong>f if <strong>the</strong> progression <strong>of</strong> treatments is simple andmethodical. Based on <strong>the</strong>se practical considerations, ourcrews have adopted a strategy <strong>of</strong> sweeping Spartina from <strong>the</strong>eastern side <strong>of</strong> <strong>the</strong> Bay first and <strong>the</strong>n finishing up on <strong>the</strong>west side. The federal crews are moving from south to north,while <strong>the</strong> state crews in general are operating from north tosouth.For many years, most parties were driven by narrowself-interest. Under those conditions, it was impossible tocreate a coherent and effective strategy. The lack <strong>of</strong> any—even a flawed—long-term vision allowed agencies to pursue<strong>the</strong>ir own near-term interests without political consequences.No one had <strong>the</strong> authority or permission to impose <strong>the</strong>discipline <strong>of</strong> working toge<strong>the</strong>r to treat some lands first,o<strong>the</strong>rs later, and some at <strong>the</strong> very end. Without such a multiyearcommitment, no one had confidence <strong>the</strong> program wouldlast long enough to serve <strong>the</strong>ir own interests. The brevity <strong>of</strong>budget periods constrained long-term thinking. With federalbudgets set on an annual basis and state budgets set everytwo years, <strong>the</strong>re were no absolute guarantees that <strong>the</strong>funding would be maintained to <strong>the</strong> very end. Still, with awell thought-out plan and <strong>the</strong> support <strong>of</strong> all parties, you candraw toge<strong>the</strong>r enough political will to have a good shot atsustaining <strong>the</strong> program over <strong>the</strong> required number <strong>of</strong> years.LESSON 8: LOSE THE THIN SKINI have to conclude with some cautionary words about<strong>the</strong> emotional toll that comes with difficult challenges likecontrolling a widespread invasive weed. Weed eradicationwork in <strong>the</strong> field is tough and dirty business. People get hurt,- 232 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and Managementequipment breaks down, and tempers fray. In Willapa Bay,<strong>the</strong> social and political dynamics among agency staff andstakeholders were as challenging as <strong>the</strong> physical tests.Natural resource pr<strong>of</strong>essionals should remember to keepperspective and understand why <strong>the</strong> affected stakeholdersget very angry at times. In <strong>the</strong> Willapa Bay area, manypeople rely on <strong>the</strong> resources <strong>of</strong> <strong>the</strong> Bay for <strong>the</strong>ir livelihood.Their jobs and lifestyle are in jeopardy. Even if noteconomically tied to <strong>the</strong> Bay, many o<strong>the</strong>rs feel very deepaffection for <strong>the</strong>ir surroundings. Large amounts <strong>of</strong> publicmoney were being expended, yet, year after year, Spartinaspread and no progress was evident. It is no wonder thatmeetings <strong>of</strong>ten became contentious. In my view, agencystaff <strong>of</strong>ten reacted as if <strong>the</strong> criticisms <strong>of</strong> <strong>the</strong> program werepersonal affronts. Ra<strong>the</strong>r than acknowledging <strong>the</strong> reasons forfrustration and <strong>the</strong> importance <strong>of</strong> re-evaluation in light <strong>of</strong> <strong>the</strong>slow progress, <strong>the</strong>y decided to reduce <strong>the</strong>ir exposure to localstakeholders. The heightened sensitivity to criticismdisplayed by state agency staff involved in <strong>the</strong> WillapaSpartina program became an obstacle that prevented seriousevaluation and program improvement. If <strong>the</strong> agencies hadtaken an attitude that welcomed criticism as a means to seekprogram improvements, <strong>the</strong> agencies would have gone along way to enlist public support and confidence. Becausemany <strong>of</strong> <strong>the</strong> agency staff asked to manage this complicatedprogram were relatively young and inexperienced, it isunderstandable that <strong>the</strong>ir public relations skills wereinadequate. Less understandable was <strong>the</strong> disinterestdisplayed by <strong>the</strong>ir supervisors in requiring conduct thatwould improve relations with local constituents. It appearedthat <strong>the</strong> distrust <strong>of</strong> government decision-making commonamong citizens was met by its mirror image within agencies:a distaste felt by agency employees for public involvementin decision-making.For better or worse, natural resource issues do belong toeveryone. Pr<strong>of</strong>essionals who choose this line <strong>of</strong> work shouldaccept that <strong>the</strong> people most affected by <strong>the</strong> success or failure<strong>of</strong> <strong>the</strong>ir work will want to have a voice in decision-making.An attitude <strong>of</strong> intolerance towards criticism will only invitemore criticism. When <strong>the</strong> results visible on <strong>the</strong> ground areextremely unsatisfactory, little else counts for much. Soulsearching and healthy criticism are helpful after all. Themission to succeed should outweigh all petty resentments. In<strong>the</strong> end, <strong>the</strong>re is no choice but to work toge<strong>the</strong>r.- 233 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementDISCOVERY AND MANAGEMENT OF SPARTINA ANGLICA IN THE FRASER RIVER ESTUARY,BRITISH COLUMBIA,CANADAG. WILLIAMS 1 ,J.BAUMANN 2 ,D.BUFFETT 3 ,R.GOLDSTONE 4 ,V.KUCY 5 ,P.LIM 6 ,W.MATHER 7 ,K.MOORE 89, 10AND T. MURRAY1 GL Williams & Associates Ltd., 2907 Silver Lake Place Coquitlam, BC, V3C 6A2; glwill@telus.net2 Port Metro Vancouver, 100 The Pointe, 999 Canada Place, Vancouver, BC, V6C 3T43 Ducks Unlimited Canada, Unit 511, 13370 - 78 Avenue, Surrey, BC, V3W 0H64 Ministry <strong>of</strong> Environment, Parks & Protected Areas, PO Box 220, Brackendale, BC, V0N 1H05 The Corporation <strong>of</strong> Delta, 4500 Clarence Taylor Crescent, Delta, BC, V4K 3E26 Fisheries & Oceans Canada, Major Projects Review Unit, Habitat Enhancement Branch, Suite 200 – 401 Burrard Street,Vancouver, BC, V6C 3S47 Metro Vancouver Regional District, Regional Parks, 4330 Kingsway, Burnaby, BC V5H 4G88 Environment Canada, Canadian Wildlife Service, R.R.1, 5421 Robertson Road, Delta, BC V4K 3N29 Vancouver Aquarium Marine Science Centre, PO Box 3232, Vancouver, BC V6B 3X810 Current address: Greater Vancouver Invasive Plant Council, 6435 Marine Drive, Burnaby, BC V3N 2Y5Spartina anglica (English cordgrass) was discovered in August 2003, during intercauseway marshsurveys <strong>of</strong> Roberts Bank for a proposed Vancouver Port Authority (VPA) container terminalexpansion. This was a new species discovery for <strong>the</strong> estuary and <strong>the</strong> province <strong>of</strong> British Columbia,and a control program was immediately initiated by VPA. The control program involved collectingpreliminary data on <strong>the</strong> S. anglica infestation in September (measuring stalk density and height,conducting GPS surveys, GIS mapping) followed by manual removal in October. Spartina clones,measuring 3–4.5 meter (m) diameter, had an average stalk density <strong>of</strong> approximately 800 stalks persquare meter (m 2 ) and average height <strong>of</strong> 1 m, but most <strong>of</strong> <strong>the</strong> growth was in <strong>the</strong> form <strong>of</strong> seedlingsand small tussocks

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaon <strong>the</strong> North Arm. The concentrated port activity includes aconstant movement <strong>of</strong> barges and ships through <strong>the</strong> estuary.S. anglica is <strong>of</strong> particular concern in <strong>the</strong> Fraser estuarybecause <strong>of</strong> <strong>the</strong> known ecological impacts, includingconversion <strong>of</strong> mudflats to monoculture Spartina marshes(Gray et al. 1991; Huckle et al. 2000), displacement <strong>of</strong>waterbirds (Goss-Cutard and Moser 1988; Nairn 1986), fish(Sullivan 2001 cited in Anon. 2002; Dethier and Hacker2004), and invertebrates (Hedge and Kriwoken 2000;Jackson et al. 1985), accretion <strong>of</strong> sediment (Ranwell 1964;Gray 1991; Dethier and Hacker 2004), disruption <strong>of</strong> tidaldrainage patterns (Gray et al. 1991), and impacts on nativeintertidal marsh, eelgrass beds, and unvegetated habitats(Hedge and Kriwoken 2000; Hacker et al. 2001). Experiencein Washington State and elsewhere has shown that a quickresponse is required to avoid costly Spartina eradicationprograms. For example, $1.5 million was spent in 2003 totreat 6,000 ac <strong>of</strong> solid Spartina (70% <strong>of</strong> <strong>the</strong> infestation) inWillapa Bay and 694 solid ac (90% <strong>of</strong> <strong>the</strong> infestation) inPuget Sound (Murphy 2003). This paper summarizes <strong>the</strong>work undertaken in 2003 and 2004 in Roberts Bank andBoundary Bay and provides suggestions for fur<strong>the</strong>r work toeffectively control S. anglica in <strong>the</strong> Fraser River estuary.METHODS AND PROGRAM PLANNINGFig.1. Location <strong>of</strong> Fraser River estuary.The large tidal flats provide vast mudflat, eelgrass, andbrackish and salt marsh habitats for birds, fish and wildlife(Adams and Williams 2004; Harrison and Dunn 2004;Schaefer 2004). Over one million migratory birds use <strong>the</strong>outer delta for critical feeding and staging areas annually(Butler and Campbell 1987; Important Bird Area (IBA) SiteSummary, BC017, www.ibscanada.com; Pacific EstuaryConservation Program, www.ramsar.org). Numerous fishand invertebrate species utilize <strong>the</strong> flats, including Pacificsalmon and Dungeness crab that support importantcommercial, recreational, and First Nations fisheries (Hoosand Packmann 1974; Levings 2004). Much <strong>of</strong> <strong>the</strong> area isprotected within provincial Wildlife Management Areas, isrecognized as a nationally Important Bird Area, and exceeds<strong>the</strong> criteria for Ramsar and Western Hemisphere ShorebirdNetwork sites (Important Bird Area, IBA, Site Summary,BC017, www.ibacanada.com). The 586-hectare AlaskanNational Wildlife Area on Westham Island, adjacent toRoberts Bank, is a designated Ramsar site.Fraser River estuary is also a hub for domestic andinternational shipping with three port authorities: VPA’scontainer and bulk terminals at Roberts Bank, Fraser Port’scoastal and deep sea terminals on <strong>the</strong> Main Arm, and PortNorth Fraser’s coastal shipping and log booming facilities2003 VPA Spartina Control Program for Roberts BankFollowing discovery <strong>of</strong> S. anglica in Roberts Bank inAugust, GL Williams & Associates Ltd. was retained byVPA to establish a control program. Since <strong>the</strong> infestationwas limited in coverage and total area, <strong>the</strong> objective was tomanually remove <strong>the</strong> plants as quickly as possible without<strong>the</strong> use <strong>of</strong> herbicide. The VPA control program consisted <strong>of</strong>confirming identification <strong>of</strong> <strong>the</strong> Roberts Bank Spartina,researching scientific literature to guide managementoptions, conducting field GPS/GIS mapping and datacollection, and undertaking plant removal.Plant IdentificationPlant specimens were collected, digital images obtained,and a web search conducted for information about S.anglica. The field identification manual prepared by <strong>the</strong>California Coastal Conservancy (www.spartina.org) wasespecially useful for preliminary identification and providingdescriptions <strong>of</strong> <strong>the</strong> five Spartina species found along <strong>the</strong>Pacific Coast. Several wetland specialists in BritishColumbia and Washington State were contacted forappropriate keys and information. The indentification <strong>of</strong> <strong>the</strong>Roberts Bank specimens were confirmed using <strong>the</strong> Spartinakey developed by Barkworth (2003). Specimens were alsosubmitted to <strong>the</strong> Univeristy <strong>of</strong> British Columbia herbarium.Fresh S. anglica specimens from Roberts Bank andBoundary Bay were shipped to Dr. Debra Ayres, University<strong>of</strong> California at Davis, for DNA analysis (Ayres and Strong- 236 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and Management2001). DNA analysis was undertaken to confirm <strong>the</strong>identification and determine if <strong>the</strong> origin <strong>of</strong> <strong>the</strong> S. anglicacould be established. Finally, specimens from B.C. werecompared with samples <strong>of</strong> S. anglica and S. alterniflorafrom Padilla Bay during a meeting with S. Riggs at <strong>the</strong>Padilla Bay Estuarine Research Reserve in Mount Vernon,Washington.Field Surveys and MappingIn late September, S. anglica plants were located using ahandheld Garmin XL GPS, and a GIS map was producedusing Fraser River estuary orthophotos. To prevent <strong>the</strong>release <strong>of</strong> seeds, above-ground leaves and all inflorescenceswere cut and collected in plastic bags for disposal. Densityand height were measured for a small subsample (n=4)within <strong>the</strong> largest clone.Spartina RemovalOn October 24, 2003, <strong>the</strong> first Fraser Spartina Busterswere assembled on Roberts Bank The group was comprised<strong>of</strong> 21 volunteer representatives from VPA, federal, andprovincial agencies, Ducks Unlimited Canada (DUC),Langely Environmental Partners Society, and TsawwassenFirst Nations (TFN). The removal was conducted over onelow-tidal cycle and involved digging up plants with shovels,washing <strong>the</strong> roots and rhizomes in pools to removesediment, placing <strong>the</strong> washed plants in industrial-strengthplastic bags, and collecting <strong>the</strong>m in a shore-based disposalbin for incineration at <strong>the</strong> GVRD incinerator. All abovegroundmaterial was removed except for three clonesmeasuring 3 m or greater in diameter.Hovercraft Survey <strong>of</strong> <strong>the</strong> Outer Fraser EstuaryThe Canadian Coast Guard hovercraft, operated byFisheries and Oceans Canada (FOC), was used to survey <strong>the</strong>mudflat-marsh interface on Sturgeon Bank, Roberts Bankand Boundary Bay during low tide on November 19.Observers included <strong>the</strong> author (GW) and biologists fromFOC and VPA. Environment Canada, GVRD, and FOCprovided funding for <strong>the</strong> survey. GPS locations <strong>of</strong> S. anglicawere recorded and <strong>the</strong> distribution mapped.Outreach and EducationTo increase <strong>the</strong> awareness <strong>of</strong> <strong>the</strong> ecological impacts <strong>of</strong>S. anglica to <strong>the</strong> Fraser River estuary, a PowerPointpresentation was shown to several local naturalist groups.Articles were published in <strong>the</strong> Vancouver Natural HistorySociety newsletter (Williams 2004) and Botanical ElectronicNews (Williams 2004a). Interviews were also conducted andarticles published in local newspapers. CanadianBroadcasting Corporation (CBC) radio ran a short feature onS. anglica during a newscast. VPA is also consideringestablishing an environmental stewardship program withTFN that would include <strong>the</strong> monitoring and removal <strong>of</strong>Spartina in Roberts Bank.2004 Spartina anglica Control ProgramCoordination <strong>of</strong> ManagementFollowing <strong>the</strong> confirmation that S. anglica was notrestricted to Roberts Bank, <strong>the</strong> invasive species coordinatorfor FOC, Pat Lim, was contacted and a committeeestablished to deal with <strong>the</strong> S. anglica threat. OnDecember 4, 2003, representatives <strong>of</strong> <strong>the</strong> agenices involvedin <strong>the</strong> Roberts Bank control program met with invitedSpartina eradication specialists from Washington State todevelop an action plan for <strong>the</strong> control and removal <strong>of</strong> <strong>the</strong> S.anglica from Boundary Bay and Roberts Bank. WashingtonState representatives <strong>of</strong>fered to assist in <strong>the</strong> control andremoval program.The action plan, developed over several meetings,consisted <strong>of</strong> using volunteers to manually dig out plantsduring low tides in June, with a follow-up removal in <strong>the</strong>fall. A preliminary budget was determined to cover hiring avolunteer coordinator, producing educational materials, andsupporting a field operation. Matching funds werenegotiated with partner agencies and organizations, andconsiderable in-kind support was obtained.Mapping S. anglica DistributionOn May 31 and June 4, 2004, FOC mapped <strong>the</strong> easternarea <strong>of</strong> Boundary Bay between 104 th and 112 th Streets inDelta using handheld GPS units (Garmin eTrex Vista andGPS 76 Marine Navigator) with an accuracy <strong>of</strong> ± 5 m..Spartina was classified according to one <strong>of</strong> four classes:seedling, clone 0.3 m to 1 m diameter. A GIS map was produced using a recentorthophoto base map. The map was useful in indicating <strong>the</strong>amount <strong>of</strong> Spartina present and <strong>the</strong> allotment <strong>of</strong> resourcesfor removal.New growth <strong>of</strong> S. anglica was mapped in Roberts Bankin June and October using a Garmin 60 C handheld unit anda GIS map produced by DUC. On November 24, ahovercraft survey was conducted in Boundary Bay to mapnew and existing S. anglica. The GPS locations were addedto <strong>the</strong> GIS map produced by DUC.A survey <strong>of</strong> selected sites in Burrard Inlet (e.g.Maplewood, Port Moody mudflats, and Spanish Banks) wasconducted at low tide on November 21, 2004. No S. anglicawas observed at <strong>the</strong> Burrard Inlet sites.Outreach and EducationOn June 5, World Oceans Day was organized at BlackieSpit Park and a S. anglica poster and materials were put ondisplay. The event also provided an opportunity to recruitvolunteers for <strong>the</strong> S. anglica removal in Boundary Bayscheduled for June 17–19. FOC produced a fact sheet fordistribution at <strong>the</strong> event and o<strong>the</strong>r forums.PowerPoint presentations were made to several NGO’s,including naturalist groups and environmental training- 237 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaGel image <strong>of</strong> S. anglica (first 2 lanes) and S. alterniflora x foli hybridsBand is absent in S. alt (and S. foli)S. maritima band in S. anglicaFig. 2. Spartina anglica on Roberts Bank in August 2003.programs. Ducks Unlimited Canada and <strong>the</strong> Corporation <strong>of</strong>Delta issued several press releases. Delta established a website on Spartina anglica as part <strong>of</strong> <strong>the</strong>ir outreach program.An oral presentation describing <strong>the</strong> Fraser Spartina Bustersprogram was also made at <strong>the</strong> 3 rd <strong>International</strong> InvasiveSpartina <strong>Conference</strong> in San Francisco, November 8-10,2004.Volunteer Recruitment and ManagementVolunteer organizations and individuals were contactedby email providing information on <strong>the</strong> meeting time andplace, as well as mapping and field needs for <strong>the</strong> BoundaryBay S. anglica removal. A meeting room at <strong>the</strong> Delta AirPark was made available to <strong>the</strong> organizers for volunteer signup, supply and equipment staging, and brief orientation.Laminated S. anglica identification cards were given to allvolunteers to assist in accurate identification duringremovals.Several volunteer removal efforts were held from Juneand October in Boundary Bay and Roberts Bank. Eightyeightvolunteers participated in <strong>the</strong> June 17–19 removal. Afall follow-up was held October 13–15 supported by 19volunteers.Volunteers received a brief instruction on manualremoval methods in <strong>the</strong> meeting room during <strong>the</strong> Juneremoval, with field demonstrations conducted in subsequentremovals.Manual and Mechanical Removal MethodsThe manual removal methods were similar to <strong>the</strong>Roberts Bank program. However, an all-terrain vehicle,provided by Delta, was used to collect and transport bags to<strong>the</strong> dike. Filled bags were piled at collection points on top <strong>of</strong><strong>the</strong> dike and subsequently loaded by backhoe into a truckFig. 3. DNA analysis gel image <strong>of</strong> Roberts Bank S. anglica (left samples)showing S. maritima bands and absence <strong>of</strong> S. alterniflora (courtesy <strong>of</strong> D.Ayres, University <strong>of</strong> California, Davis).and taken to <strong>the</strong> GVRD incinerator, where <strong>the</strong>y wereweighed prior to incineration.For larger clones (i.e., >1 m diameter), an amphibiousswamp excavator operated by Concord Excavating andContracting Ltd. was used to bury <strong>the</strong> plants in deep holesexcavated in <strong>the</strong> mudflat. The machine ran on large trackedpontoons and left a light footprint on <strong>the</strong> ground (< 1 psi).Excavated S. anglica was buried in holes about 3–5 m deepand backfilled with excavated material, which was graded to<strong>the</strong> surrounding substrate (referred to earlier as in situburial). At some locations, GPS was used to record <strong>the</strong>location for future monitoring. The excavator was also usedto remove several large clones in <strong>the</strong> western section <strong>of</strong>Boundary Bay in June.Two removal efforts were also conducted at <strong>the</strong> RobertsBank site. On August 5, <strong>the</strong> swamp excavator was used tocomplete in situ burial <strong>of</strong> <strong>the</strong> three large clones. A volunteerremoval <strong>of</strong> seedings occurred on September 28, involvingeight person-days <strong>of</strong> effort.RESULTSSpartina IdentificationS. anglica was quite easily identified based on habitatand plant characteristics. Most plants colonized mudflatelevations well below native salt marsh. The oblique bladeand vertical spikelet arrrangements were distinguishingcharacteristics in <strong>the</strong> field (Fig. 2).DNA analysis showed <strong>the</strong> distinctive S. maritima bands,lacking in S. alterniflora (Fig. 3). However, due to <strong>the</strong>limited genetic variation in S. anglica, it was not possible todetermine <strong>the</strong> source <strong>of</strong> <strong>the</strong> Fraser River estuary specimens(D. Ayres, pers. comm.).- 238 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementFig 5. November 2003 GPS locations <strong>of</strong> main S. anglica infestations recordedduring hovercraft survey.Fig. 4. Comparison <strong>of</strong> S. anglica distribution on Roberts Bank in 2003 and2004.S. anglica Distribution on Roberts Bank and BoundaryBayIn 2003, S. anglica was present on Roberts Bank inthree 3-m diameter clones and approximately 50 scatteredseedlings and tussocks up to 1 m in diameter. The totalmudflat area affected was 60 ha (148 ac), but <strong>the</strong> actual totalarea covered by S. anglica was only about 0.1 ha (0.2 ac).The average plant density and height in <strong>the</strong> main clone(n=4) was 217 ± 42 shoots/0.25 m 2 and 108 ± 6 cm,respectively). Individual seedlings were much smaller,ranging from 0.1–0.3 m (4–12 in.) in height.In 2004, new seedling growth was found on RobertsBank, despite previous clearing in 2003 (Fig. 4). Seedlingdistribution, mapped in June and October, showed that <strong>the</strong>distribution <strong>of</strong> <strong>the</strong> plants continued to expand through <strong>the</strong>summer. Above-average temperatures experienced in <strong>the</strong>summer <strong>of</strong> 2004 appeared to increase <strong>the</strong> spread <strong>of</strong> S.anglica.In 2003, S. anglica in Boundary Bay was concentratedin two nodes near Beach Grove, in <strong>the</strong> west and between 96 thand 112 th Streets in <strong>the</strong> central portion (Fig.5). Most <strong>of</strong> <strong>the</strong>observed S. anglica was present as seedlings or smalltussocks, with occasional clones ranging from 1–3mdiameter. Total area impacted was about 135 ha (330 ac),but total coverage was estimated at 1% or 1.35 ha (3 ac).Similar to Roberts Bank, Boundary Bay seedlingabundance increased in 2004, and new areas weredocumented in Surrey and Mud Bay in <strong>the</strong> eastern section <strong>of</strong>Boundary Bay. The most densely infested area was locatedwest <strong>of</strong> 112 th Street, outside <strong>the</strong> FOC mapped area. Thecomposite map <strong>of</strong> Spartina anglica locations is shown inFig. 6.Spartina RemovalIn 2003, <strong>the</strong> one-day removal effort (total <strong>of</strong> 21 persondays)was successful in removing all S. anglica plants fromRoberts Bank, except for those in <strong>the</strong> three 3-meter diameterclones. Roberts Bank sediments were quite s<strong>of</strong>t which madewalking and digging more difficult. Manual removal <strong>of</strong> <strong>the</strong>dense growth in <strong>the</strong> clones was slow, and only half <strong>of</strong> one <strong>of</strong><strong>the</strong> 3-m diameter clone was removed by six volunteersworking a full day (Fig. 7). The sediments were washedfrom <strong>the</strong> plants to reduce <strong>the</strong> weight <strong>of</strong> collection bags. InBoundary Bay, <strong>the</strong> sediments were firmer (i.e. sand) whichfacilitated removal. The use <strong>of</strong> an all-terrain vehicle toremove <strong>the</strong> bags greatly reduced <strong>the</strong> labour. The 140 persondays<strong>of</strong> removal effort in 2004 was not sufficient to removeall <strong>the</strong> S. anglica.The excavator and in situ burial used in 2004 proved toFig. 6. Distribution <strong>of</strong> S. anglica on Roberts Bank and Boundary Bay.- 239 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinabe a more efficient removal method for larger clones. It isunlikely that S. anglica will grow from a depth <strong>of</strong> 2–3 m.Total plant material disposed <strong>of</strong> in 2004 was about8,000 kilograms (kg) and it is estimated that approximatelyfive times that amount was removed using in situ burial. Thetotal weight included plants and some sediment.Hovercraft Survey <strong>of</strong> <strong>the</strong> Outer Fraser EstuaryA hovercraft provides an efficient survey method forcovering large areas <strong>of</strong> tidal flats, especially where mudsubstrates make walking difficult. Seedlings are easilyobserved from ei<strong>the</strong>r inside or from <strong>the</strong> foredeck. The onboarddigital charts and GPS provide accurate locating. Thecraft can travel ei<strong>the</strong>r rapidly or at a very low speed, and iscapable <strong>of</strong> stopping if <strong>the</strong>re is a need to obtain plantspecimens.Outreach and EducationThe outreach component <strong>of</strong> <strong>the</strong> VPA Roberts Bankprogram included PowerPoint presentations, localnewspaper and radio, and printed/electronic publications.The PowerPoint presentations were well received and <strong>of</strong>tenled to invitations from o<strong>the</strong>r groups. The newspaper articlesprovided broad local coverage, which appeared to reach alarger audience than radio. The Vancouver Natural HistorySociety (VNHS) newsletter article led to a request to preparea more detailed article for <strong>the</strong> VNHS journal, Discovery.The electronic publication in Botantical Electronic News(BEN), being available on <strong>the</strong> web, led to communicationsand exchange <strong>of</strong> information with scientists in France andOregon.As well as increasing awareness <strong>of</strong> <strong>the</strong> S. anglica threat,<strong>the</strong> outreach was effective in recruiting volunteers. Duringconversations with removal volunteers, many said <strong>the</strong>y weremotivated to provide assistance after seeing <strong>the</strong> PowerPointpresentation or attending local World Oceans Day eventswhere information about <strong>the</strong> problem was presented.Program CostsThe 2003 VPA S. anglica management program wasdeveloped and coordinated by <strong>the</strong> author (GW), andvolunteers conducted <strong>the</strong> actual removal. The total budgetwas approximately $15,000, covering consultant fees,disposal trucking, incineration, and supplies (e.g. plasticbags, gloves, etc.). The hovercraft survey cost about $1500,including reporting and mapping. There were also in-kindcontributions <strong>of</strong> hovercraft fuel (provided by Fisheries andOceans Canada) and staff salaries.The 2004 program costs were over $50,000 plus amatching amount in-kind from many agencies involved in<strong>the</strong> Spartina removal program.DISCUSSION AND RECOMMENDATIONSThe management efforts in 2003 and 2004 were usefulin developing strategeies to effectively control S. anglica inFig. 7. Fraser Spartina Busters on Roberts Bank in October 2003.<strong>the</strong> Fraser estuary. Spartina. anglica was fully controlled atRoberts Bank by <strong>the</strong> end <strong>of</strong> October 2004, althoughcontinued management will be required to address newrecuitment from existing seedbank and new invasions. Thedistribution <strong>of</strong> S. anglica in Boundary Bay has been greatlyreduced, but <strong>the</strong>re is still a substantial amount remaining east<strong>of</strong> 112 th St., and scattered new seedlings were observedthroughout Boundary Bay in November 2004. The Fraserestuary program shows that control is possible without <strong>the</strong>use <strong>of</strong> herbicides, provided that <strong>the</strong> infestation is caughtearly and aggressive removal programs implemented(Daehler and Strong 1996; Anon 2002).Manual effort was effective in removing S. anglica in<strong>the</strong> Fraser estuary, but in situ burial is more effective forlarge clones exceeding 1 m. Digging and bagging plants isefficient for seedlings and tussocks, but <strong>the</strong> use <strong>of</strong> all-terrainvehicles for collection and transport to shore for disposal isrequired for large scale programs. Firmer substrates permit<strong>the</strong> use <strong>of</strong> mechanized equipment and are generally easier towork on than muddy sediments.The Fraser estuary Spartina programs depended onvolunteer manpower, but dedicated work crews arerecommended for Boundary Bay. Boundary Bay will requirea greater effort than 107 person-days <strong>of</strong> removal to reach <strong>the</strong>status <strong>of</strong> Roberts Bank. However, much <strong>of</strong> <strong>the</strong> S. anglica inBoundary Bay consists <strong>of</strong> seedlings, which can be easilyremoved. Denser coverage occurs in localized areas, such as<strong>the</strong> six, 3 m-diameter clones west <strong>of</strong> 104 th Street or mixedwithin native marsh west <strong>of</strong> 112 th Street.Mapping was a critical component <strong>of</strong> <strong>the</strong> 2003 and 2004programs. Handheld GPS units provide sufficient accuracyfor mapping S. anglica, and classification <strong>of</strong> <strong>the</strong> growth asseedling, tussock (0.1–0.3 m), small clone (0.3–1m) or largeclone (>1 m) provides sufficient detail for planning manual- 240 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and Managementor mechanical removal. Seedlings and tussocks can beremoved manually, while large clones will requiremechanical, in situ burial. Mapping should be conducted inJune when <strong>the</strong> plants have matured and are easily located,and removal efforts should be conducted during July.Flowering occurs from June to September, but may be aslate as November, and most seed is set in September(Dethier and Hacker 2004).Following <strong>the</strong> summer removals, FOC participation in<strong>the</strong> program was discontinued and Environment Canadaassumed <strong>the</strong> chair <strong>of</strong> <strong>the</strong> steering committee. Senior agenciesshould make a higher commitment to S. anglicamanagement and provide sufficient manpower and financialsupport to bring <strong>the</strong> infestation under control. Withoutsufficient follow-up S. anglica will spread and make controland removal much more difficult and expensive in <strong>the</strong>future.Training is an important component <strong>of</strong> S. anglicaremoval. Some volunteers had difficulty distinguishing itfrom native salt marsh species. Although complete removal<strong>of</strong> above-ground plants and below-ground roots andrhizomes was stressed during site demonstrations, someremovals were incomplete. These concerns can beminimized by providing better training and by implementinginspections by qualified supervisors during removal efforts.During both 2003 and 2004 <strong>the</strong>re were substantial inkindcontributions from many particpating agencies. Forexample, FOC covered <strong>the</strong> hovercraft fuel costs; Deltaprovided <strong>the</strong> all-terrain vehicle, bag collection, and dikeaccess; FOC provided GPS and GIS support, a Spartina factsheet, and meeting rooms; MWLAP arranged for <strong>the</strong>excavator; and DUC arranged for portable toilets in <strong>the</strong> field,GIS mapping, and press releases.Increased surveillance is required, which benefits frombetter outreach. Providing presentations to interested partiesis useful for increasing awareness and building support for<strong>the</strong> program. Attempts were made to enlist <strong>the</strong> help <strong>of</strong>birders from <strong>the</strong> VNHS, but <strong>the</strong>re was no active follow-up tomake it more effective. A Spartina contact telephonenumber should established. Material should be prepared toassist in <strong>the</strong> identification <strong>of</strong> S. anglica and o<strong>the</strong>r species(e.g. S. alterniflora and S. densiflora). Outreach should alsobe extended to Vancouver Island and <strong>the</strong> Gulf Islands, wheresuitable conditions exist for establishment.POST PRESENTATION UPDATE:To manage Spartina infestations in British Columbia,<strong>the</strong> BC Spartina Working Group (BCSWG) was formed, ledby Ducks Unlimited Canada. The BCSWG has preparedannual reports and commissioned <strong>the</strong> BC Spartina ResponsePlan (Dresen et al. 2010). Spartina mapping and relatedinformation pertaining to <strong>the</strong> BCSWG efforts are providedon <strong>the</strong> Community Mapping Network websitewww.Spartina.ca.ACKNOWLEDGEMENTSThe authors would like to thank all <strong>the</strong> volunteers forparticipating in <strong>the</strong> 2003 and 2004 Fraser Spartina Bustersremoval efforts, including <strong>the</strong> Langley EnvironmentalPartners Society, Vancouver Aquarium RiverWorks, TheNature Trust, Tsawwassen First Nations, NorthwestWildlife, American partipants from Whatcom County,Whatcom County Noxious Weed Board, Puget SoundAction Team, and many agency staff and individuals.Cheryl Lynch and Kay Kennes, FOC, conducted GPSsurveys and GIS mapping <strong>of</strong> <strong>the</strong> Spartina infestations and<strong>the</strong>ir contribution is appreciated. FOC librarians MarciaCroy VanWely, Louise Archibald and John Alexander,researched <strong>the</strong> eology <strong>of</strong> S. anglica for <strong>the</strong> fact sheet.We gratefully acknowledge Dr. Adolf Ceska, Dr. DebraAyres, Dr. Sally Hacker, Sharon Riggs, Dave Heimer, KyleMurphy, and Cindy Sayre for providing information andassisting with plant identification.The authors appreciate <strong>the</strong> financial contributions <strong>of</strong> <strong>the</strong>Western Regional Panel on Aquatic Nuisance Species;Darrell Desjardin, VPA; Ted Lea; MWLAP, for funding S.anglica removal and surveillance; GVRD and EnvironmentCanada for funding 2003 surveillance efforts; and B. Naitoand B. Mason, FOC, for arranging <strong>the</strong> hovercraft surveys.The authors thank <strong>the</strong> Galiano Conservancy forpermission to use <strong>the</strong> estuary satellite image in Fig. 1. Theeditorial comments from Kristy Williams were appreciated.REFERENCESAdams, M.A., and G.L. Williams. 2004. Tidal marshes <strong>of</strong> <strong>the</strong> FraserRiver estuary: composition, structure, and a history <strong>of</strong> marshcreation efforts to 1997: In: B.J. Groulx, D.C. Mosher, J.L.Luternauer, and D.E. Bilderback , eds. Fraser River Delta, BritishColumbia: Issues <strong>of</strong> an Urban Estuary, Geol. Surv. Can. Bull.567:147-172.Anon. 2002. Strategy for <strong>the</strong> management <strong>of</strong> rice grass (Spartinaanglica) in Tasmania, Australia. Natural Heritage Trust andTasmania Department <strong>of</strong> Primary Industries, Water and Environment,Hobart, Tasmania.Ayres, D.R., University <strong>of</strong> California, Evolution and Ecology,Davis, personal communication.Ayres, D. R., and D. R. Strong. 2001. Origin and genetic diversity<strong>of</strong> Spartina anglica (Poaceae) using nuclear DNA markers.American Journal <strong>of</strong> Botany. 88(10): 1863-1867.Barkworth, M.E. 2003. Spartina Schreb., pp. 240-251. In: Barkworth,M.E., K.M. Carpels, S. Long, and M. B. Piep, eds., Flora<strong>of</strong> North America North <strong>of</strong> Mexico, Vol. 25: Magnoliophyta:Commelinidae (in part): Poaceae, part 2, Oxford UniversityPress.Butler, R.W., and W. Campbell. 1987. The birds <strong>of</strong> <strong>the</strong> FraserRiver delta: populations, ecology and international significance.Canadian Wildlife Service Occasional Paper 65. EnvironmentCanada.- 241 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaDaehler, C.C., and D. R. Strong. 1996. Status, prediction and prevention<strong>of</strong> introduced cordgrass Spartina spp. Invasions in Pacificestuaries, USA. Biological Conservation 78:51-58.Dethier, M.N., and S.D. Hacker. 2004. Improving managementpractices for invasive cordgrass in <strong>the</strong> Pacific Northwest: a casestudy <strong>of</strong> Spartina anglica. Washington Sea Grant Program, College<strong>of</strong> Ocean and Fishery Sciences, University <strong>of</strong> Washington,WSG-AS-04-05, Seattle, WA.Dresen, K., L. (EDI Environmental Dynamics Ltd.), L. Scott (Eco-Matters Consulting), and G. L. Williams (GL Williams & AssociatesLtd.) 2010. BC Spartina response plan 2010. Prepared forDucks Unlimited Canada, Surrey, BC: 73 p.Goss-Custrad, J.D., and M.E. Moser. 1988. Rates <strong>of</strong> change in <strong>the</strong>numbers <strong>of</strong> dunlin, Calidris alpine, wintering in British estuariesin relation to <strong>the</strong> spread <strong>of</strong> Spartina anglica. Journal <strong>of</strong> AppliedEcology 25:95-109.Gray, A.J., D.F. Marshall, and A.F. Raybould. 1991. A century <strong>of</strong>evolution in Spartina anglica. Advances in Ecological Research21:1-62.Hacker S.D., D. Heimer, C.E. Hellquist, T.G. Reeder, B. Reeves,T.J. Riordan, and M.N. Dethier. 2001. A marine plant (Spartinaanglica) invades widely varying habitats: potential mechanisms<strong>of</strong> invasion and control. Biological Invasions 3: 211-217.Harrison, P.G., and M. Dunn. 2004. Fraser River delta seagrassecosystems, <strong>the</strong>ir distributions and importance to migratorybirds. In: Groulx, B.J., D.C. Mosher, J.L. Luternauer, and D.E.Bilderback, eds. Fraser River Delta, British Columbia: Issues <strong>of</strong>an Urban Estuary, Geol. Surv. Can. Bull. 567:173-188.Hedge, P., and L. Kriwoken. 2000. Evidence for effects <strong>of</strong> Spartinaanglica invasion on benthic macr<strong>of</strong>auna in Little Swanport estuary,Tasmania. Applied Ecology 25:150-159.Hoos, L.M., and G.A. Packman. 1974. The Fraser River estuarystatus <strong>of</strong> environmental knowledge to 1974. Environment CanadaSpecial Estuary Series No. 1.Huckle, J.M., J.A. Potter, and R.H Marrs. 2000. Influence <strong>of</strong> environmentalfactors on <strong>the</strong> growth and interactions between saltmarsh plants: effects <strong>of</strong> salinity and waterlogging. Journal <strong>of</strong>Ecology 88: 492-505.Important Bird Area IBS Site Summary, BC017,www.ibacanada.com.Jackson, D., C.F. Mason, and S.P. Long. 1985. Macro-invertebratepopulations and production on salt-marsh in east England dominatedby Spartina anglica. Oecologia 65:406-411.Levings, C.D. 2004. Knowledge <strong>of</strong> fish ecology and its applicationto habitat management. In: Groulx, B.J., D.C. Mosher, J.L.Luternauer, and D.E. Bilderback, eds. Fraser River Delta, BritishColumbia: Issues <strong>of</strong> an Urban Estuary, Geol. Surv. Can. Bull.567:213-236.Murphy, K.C. 2003. Report to <strong>the</strong> Legislature: progress <strong>of</strong> <strong>the</strong>2003 Spartina eradication program. Washington State Department<strong>of</strong> Agriculture, Pub. 805-110 (N/1/04), Olympia, WA: 58 p.Nairn, R.G.W. 1986. Spartina anglica in Ireland and its potentialon wildfowl and waders – a review. Irish Birds 3:215-228.Pacific Estuary Conservation Program, www.ramsar.com.Ranwell, D.S., 1964. Spartina saltmarshes in Sou<strong>the</strong>rn England, II.Rate and seasonal pattern <strong>of</strong> sediment accretion. Journal <strong>of</strong>Ecology 52:79-94.Schaefer, V. 2004. Ecological settings <strong>of</strong> <strong>the</strong> Fraser River delta andits urban estuary. In: Groulx, B.J., D.C. Mosher, J.L. Luternauer,and D.E. Bilderback, eds. Fraser River Delta, British Columbia:Issues <strong>of</strong> an Urban Estuary, Geol. Surv. Can. Bull. 567:35-47.Sullivan, A.J. 2001. A comparison between fish assemblages in ricegrass (Spartina anglica) and adjacent mudflat habitats in <strong>the</strong>Tamar estuary, Tasmania. B.App.Sc. Thesis, Australian MaritimeCollege.Thompson, J.D. 1991. The biology <strong>of</strong> an invasive plant: whatmakes Spartina anglica so successful? BioScience 41(6):393-401.Williams, G.L. 2004. English cordgrass (Spartina anglica). VancouverNaturalist 6(1):15.Williams, G.L. 2004a. Discovery <strong>of</strong> a new salt marsh invasive toBritish Columbia, English cordgrass (Spartina anglica C.E.Hubb.) and management initiatives in 2003. Botanical ElectronicNews No.324.- 242 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementIMPLEMENTING A STRATEGY FOR MANAGEMENT OF RICE GRASS, SPARTINA ANGLICA, INTASMANIA,AUSTRALIAP. HEDGE 1 ,C.SHEPHERD 2 AND C. DYKE 31 National Oceans Office, GPO Box 2139 Hobart Tasmania Australia 7001; Paul.Hedge@oceans.gov.au2 Department <strong>of</strong> Primary Industries, Water and Environment, GPO Box 44 Hobart Tasmania Australia 70013 Rice Grass Advisory Group, PO Box 83 Triabunna Tasmania Australia, 7190INTRODUCTIONSpartina anglica, commonly referred to in Australia asrice grass, was intentionally introduced to Tasmania in <strong>the</strong>1930s because <strong>of</strong> its potential value for coastal engineeringand agriculture. Many decades later, however, <strong>the</strong> vastmajority <strong>of</strong> <strong>the</strong> benefits provided by rice grass have beenovershadowed by <strong>the</strong> ecological, social and economic costs<strong>of</strong> its continued spread (Hedge and Kriwoken 2000;Kriwoken and Hedge 2000). In 1996 <strong>the</strong> Rice GrassAdvisory Group (RGAG) was established to provide adviceto <strong>the</strong> Tasmanian Government on future managementoptions for rice grass infestations. The RGAG facilitated <strong>the</strong>formation <strong>of</strong> a partnership between <strong>the</strong> AustralianGovernment, Tasmanian Government, Tasmanian FishingIndustry Council and <strong>the</strong> broader community to fund <strong>the</strong>development <strong>of</strong> <strong>the</strong> Strategy for <strong>the</strong> Management <strong>of</strong> RiceGrass (Spartina anglica) in Tasmania, Australia (DPIWE2002).The Management Strategy set out a range <strong>of</strong> broadobjectives and associated tasks to reduce <strong>the</strong> area <strong>of</strong> ricegrass infestations in Tasmania and identified <strong>the</strong> Department<strong>of</strong> Primary Industries, Water and Environment (DPIWE) aslead management agency. It also identified area-basedobjectives for each <strong>of</strong> Tasmania’s infested regions andpointed to Fusilade® (active constituent: 212 grams per liter(g/L) fluazifop-P present as butyl ester) as <strong>the</strong> only costeffectiveand practicable technique for controlling anderadicating infestations. It recommended <strong>the</strong> establishment<strong>of</strong> a small, multi-disciplinary team to strategically reducerice grass infestations in collaboration with affectedindustries and communities.In 1997 DPIWE sought funds to implement <strong>the</strong>Management Strategy. After considerable lobbying by <strong>the</strong>RGAG, <strong>the</strong> Natural Heritage Trust Fisheries Action Programprovided US $690,000 in funding, including US $487,000 <strong>of</strong>in-kind support. The DPIWE was charged with <strong>the</strong>responsibility to administer <strong>the</strong> project and a steeringcommittee was established to guide its development.This paper focuses on <strong>the</strong> implementation <strong>of</strong> <strong>the</strong>Management Strategy during <strong>the</strong> period from 1998 to 2002.It initially identifies <strong>the</strong> priority risks to program success andlinks <strong>the</strong>se to key management tasks, including <strong>the</strong>development <strong>of</strong> area-based management plans,environmental monitoring programs and targeted research.The tools and approaches used to develop and maintainstakeholder support, environmental monitoring programsand reduce rice grass infestations are also identified anddiscussed. Lessons learned and current challenges for ricegrass management in Tasmania are summarized.IDENTIFICATION OF KEY RISKS TO THE MANAGEMENTPROGRAMReducing rice grass infestations presents a vast range <strong>of</strong>challenges to managers, including complexities associatedwith identification <strong>of</strong> suitable control techniques, logistics <strong>of</strong>field-based operations, and stakeholder support for controlprograms (Kriwoken and Hedge 2000; Hedge et al. 2003).Understanding and effectively addressing <strong>the</strong>se challengeswas recognized by <strong>the</strong> steering committee as being critical toprogram success. Table 1 lists <strong>the</strong> priority risks to programsuccess and links <strong>the</strong>se to key management responses.An essential approach to understanding and effectivelyaddressing priority program risks involved learning frompast and current experiences <strong>of</strong> rice grass management ino<strong>the</strong>r locations. Ga<strong>the</strong>ring published information on <strong>the</strong>ecology and management <strong>of</strong> rice grass provided considerableinformation about rice grass ecology, biology, ecologicalimpacts and overviews <strong>of</strong> various approaches to controlprograms in o<strong>the</strong>r regions, including Australia, New Zealandand <strong>the</strong> United States. Such information was useful forunderstanding <strong>the</strong> range <strong>of</strong> control techniques and strategiesused to manage rice grass. Extensive collaboration andcommunications with experienced rice grass managementteams in Victoria, Australia and Washington State, USAprovided particularly useful insights into <strong>the</strong>se risks andchallenges in <strong>the</strong> field.KEY ELEMENTS OF THE MANAGEMENT PROGRAMEngaging and working with stakeholdersA high level <strong>of</strong> stakeholder engagement was critical to<strong>the</strong> success <strong>of</strong> <strong>the</strong> management program. The Rice GrassAdvisory Group, established in 1996, provided an importantmechanism to bring toge<strong>the</strong>r relevant stakeholder groups sothat <strong>the</strong>ir concerns, perspectives and ideas could be sharedand understood. The Management Strategy recognized thisgroup as <strong>the</strong> central integrating and advisory body for ricegrass management in Tasmania.The Management Strategy required area-basedManagement Plans to be developed, in close collaborationwith local stakeholders, to meet <strong>the</strong> recommendedmanagement objectives for each infested area. Three main- 243 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaTable 1: Priority risks to <strong>the</strong> Tasmanian Rice Grass Management Program and summary <strong>of</strong> key management responses.Priority Risks to Program SuccessLack <strong>of</strong> key stakeholder support formanagement programEfficacy <strong>of</strong> control techniques is lessthan 90%Actual and perceived toxicologicalimpact <strong>of</strong> herbicide on <strong>the</strong>environmentActual and perceived impact <strong>of</strong> ricegrass removal to environmentActual and perceived impact <strong>of</strong>herbicide on survival growth andmarketing <strong>of</strong> commercial Pacificoysters, Crassostrea gigasLack <strong>of</strong> funds to continueimplementation <strong>of</strong> managementprogramExponential increases in area <strong>of</strong> ricegrass infestationsManagement Responses• Develop and review area-based Management Plans with stakeholders• Establish regional ‘champions’ and establish appropriate mechanisms forongoing communication• Include experienced weed management persons in management team• Identify key variables and establish system to monitor performance or teammembers and conduct annual review <strong>of</strong> control technique efficacy• Commission risk assessment on toxicological risks <strong>of</strong> herbicide to <strong>the</strong>environment• Develop targeted monitoring program• Provide scholarships/funding to encourage student research on <strong>the</strong> impact <strong>of</strong>herbicide to non-target organisms• Commission risk assessment that considers toxicological effects <strong>of</strong> herbicideon juvenile and adult C. gigas• Conduct field and laboratory-based research on effects <strong>of</strong> herbicide on C.gigas• Demonstrate commitment to program objectives, wise use <strong>of</strong> funds andaccountability• Seek key opportunities for media releases• Identify priority areas for control• Develop realistic seasonal goals for priority control areas and developsystematic approach to reduction <strong>of</strong> infestationsmechanisms were used to engage stakeholders in developing<strong>the</strong>se plans: 1) identifying and involving champions, 2)convening regional meetings, and 3) holding targetedstakeholder meetings. Trusted community members whowere well-informed on coastal natural resource managementissues assisted with outreach to <strong>the</strong> local communities wheregrass infestations occurred, and identified <strong>the</strong> range <strong>of</strong> localstakeholder interests to be considered in developing <strong>the</strong>seplans. Champions for <strong>the</strong> intertidal oyster industry were alsoidentified and involved to ensure that <strong>the</strong>ir interests,concerns and potential contributions were clearlyrecognized/understood.In collaboration with <strong>the</strong>se champions, regionalmeetings were organized to seek community and industry, inparticular <strong>the</strong> oyster industry, support for recommendedmanagement objectives, proposed control techniques andfield-based operations. Stakeholder contributions during<strong>the</strong>se meetings were very useful for identifying coastalaccess points, preferred periods for control activities andpriority areas for control. Regional meetings providedvaluable input to <strong>the</strong> development <strong>of</strong> area-basedmanagement plans, including agreement on <strong>the</strong> roles andresponsibilities <strong>of</strong> relevant stakeholders.In some cases regional meetings were followed up withmore targeted stakeholder meetings to better understandspecific stakeholder concerns about <strong>the</strong> potential ecologicaleffects <strong>of</strong> various control techniques. These meetings <strong>of</strong>tenrequired substantial preparation and time but were importantfor generating broad stakeholder support for <strong>the</strong> program.Media releases and a quarterly newsletter keptstakeholders informed <strong>of</strong> rice grass management progress.Media opportunities were sought to highlight <strong>the</strong>achievement <strong>of</strong> key management program milestones.Environmental monitoring and targeted research.During <strong>the</strong> mid 1990s <strong>the</strong> Department <strong>of</strong> NaturalResources, Victoria, Australia, conducted a range <strong>of</strong> studiesthat investigated <strong>the</strong> efficacy <strong>of</strong> control techniques and <strong>the</strong>irecological impacts. The studies collectively pointed to <strong>the</strong>highly selective post-emergent herbicide Fusilade as <strong>the</strong>preferred means <strong>of</strong> controlling rice grass infestations.Similar studies by <strong>the</strong> DPIWE in Tasmania also pointed toFusilade as an environmentally responsible, safe, and costeffectivetechnique. However, Fusilade is not licensed foruse in <strong>the</strong> coastal environment. In 1998 <strong>the</strong> DPIWE used<strong>the</strong>se research findings to support an application for an <strong>of</strong>flabelpermit to use Fusilade to control rice grass inTasmania, which was later approved that year by <strong>the</strong>National Registration Authority for Agricultural andVeterinary Chemicals.Initially, DPIWE voluntarily restricted its use <strong>of</strong>Fusilade to <strong>the</strong> very small infestations (e.g., Derwent Riverand St Helens sites) pending <strong>the</strong> completion <strong>of</strong> anindependent environmental risk assessment on <strong>the</strong> use <strong>of</strong>- 244 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementFusilade to control rice grass (Davies 1999). This assessmentfound that although <strong>the</strong> available information was limited,<strong>the</strong> use <strong>of</strong> Fusilade to control rice grass represents a lowenvironmental risk. The report also contained a range <strong>of</strong>recommendations that DPIWE <strong>the</strong>n used to refine itsmanagement program (e.g. initially limiting spray events to0.5 hectares per day as a precautionary approach in view <strong>of</strong>limited available information), design an environmentalmonitoring program, and target future research.The risk assessment recommended that fur<strong>the</strong>r studiesbe conducted to better understand <strong>the</strong> potential effects <strong>of</strong>Fusilade on Crassostrea gigas. Laboratory-based researchwas conducted, in collaboration with <strong>the</strong> Pacific oysteraquaculture industry, to investigate <strong>the</strong> potential effects <strong>of</strong>Fusilade on survival and development <strong>of</strong> C. gigas (Hedge etal. 1999). The investigation also included a laboratory/fieldcomponent that examined ingestion and depuration <strong>of</strong> <strong>the</strong>Fusilade active ingredient (fluazifop-P-butyl) and its primarydegradation product (flauzifop-P) in C. gigas when exposedto expected field concentrations. The findings <strong>of</strong> <strong>the</strong>investigation provided fur<strong>the</strong>r evidence that Fusiladepresented a low environmental risk and was also animportant factor in securing <strong>the</strong> support <strong>of</strong> <strong>the</strong> Pacific oysteraquaculture industry for <strong>the</strong> management program.In 1999, an environmental monitoring program wasdesigned to include <strong>the</strong> following components:• Water quality monitoring to assess fluazifop-P butyl andfluazifop-P concentrations at a variety <strong>of</strong> locations ineach estuary during and following initial controloperations;• Shellfish sampling to monitor ingestions <strong>of</strong> fluazifop-Pbutyl in C. gigas;• Small-scale, comprehensively designed biomonitoringprogram <strong>of</strong> benthic macroinvertebrates in areas sprayed;and• Mapping <strong>the</strong> location and size <strong>of</strong> rice grass infestationsfollowing treatments with Fusilade.Collectively, results from research on C. gigas and <strong>the</strong>environmental monitoring program provided additionalassurance <strong>of</strong> environmental safety; consequently <strong>the</strong>management program increased <strong>the</strong> use <strong>of</strong> Fusilade to allowtreatment <strong>of</strong> up to one hectare per day.More recently, two research projects are underway toinvestigate <strong>the</strong> impact <strong>of</strong> rice grass removal on benthiccommunities in <strong>the</strong> Rubicon/Port Sorell estuary and <strong>the</strong>potential effects <strong>of</strong> <strong>the</strong> release <strong>of</strong> sediments resulting fromlarge-scale rice grass removal in <strong>the</strong> Tamar River. Theseprojects will provide useful information to decision makerswhen reviewing management objectives for <strong>the</strong> tworemaining large infestations (i.e. Rubicon/Port Sorell andTamar River).Reduction <strong>of</strong> infestationsThe Management Strategy provided clear direction forprioritizing <strong>the</strong> reduction <strong>of</strong> infestations. The generalapproach was to focus on <strong>the</strong> smallest infestations (i.e.Derwent River, St Helens, Little Swanport estuary andBridport infestations) progressively building up to largerinfestations (i.e. Circular Head, Rubicon/Port Sorell andTamar River). A small, multi-disciplinary management teamwas used to reduce all targeted infestations to <strong>the</strong> pointwhere <strong>the</strong> efforts <strong>of</strong> local communities and industry becameimportant for locating remnant patches or isolated plants.Area-based Management Plans contained <strong>the</strong> specific details<strong>of</strong> how each infestation would be reduced (i.e. managementobjectives and responsibilities, control techniques andcontrol period). This was an important planning tool thatwas used to synchronize <strong>the</strong> efforts <strong>of</strong> <strong>the</strong> DPIWE, localcommunities and industry.The management team consisted <strong>of</strong> a core <strong>of</strong> threepeople, one management <strong>of</strong>ficer and two technical <strong>of</strong>ficers.A hovercraft and operator were contracted and <strong>the</strong> vesselfitted with a 50-liter spray tank, low spray equipment and100-meter remote-controlled, retractable hose. O<strong>the</strong>requipment used included all terrain vehicles fitted with lowpressurespray equipment and backpack low-pressure sprayunits. The primary control technique was application <strong>of</strong>Fusilade (application rate <strong>of</strong> 1000 litres/hectare <strong>of</strong> 1%Fusilade solution mixed with freshwater and <strong>the</strong> surfactantBS1000 at 0.2% vol/vol to maximize efficiency). A review<strong>of</strong> field-work data sheets showed <strong>the</strong> management teamtypically achieved efficacies <strong>of</strong> 90-99%. In 2002, <strong>the</strong>DPIWE decided to engage commercial weed managementcontractors to assist with treatment in <strong>the</strong> Circular Headregion. A comparison <strong>of</strong> costs and efficacies between <strong>the</strong>management team and commercial weed contractors isprovided in Table 2.Table 3 provides a summary <strong>of</strong> <strong>the</strong> infestations, <strong>the</strong>ircurrent management objectives and size in 2002 relative to1997. The area <strong>of</strong> infestation at Derwent River, St Helens,Little Swanport estuary and Bridport have all been reducedby 99% and are on target for achieving <strong>the</strong> managementobjective. The Circular Head infestation has been reduced byapproximately 50% and is also on target to achieve <strong>the</strong>management objective. The Rubicon/Port Sorell infestation,although having increased in total area, has beensuccessfully confined to <strong>the</strong> upper half <strong>of</strong> <strong>the</strong> estuary. TheTamar River infestation, <strong>the</strong> largest in Tasmania, appears tohave increased its area but is thought to be confined to <strong>the</strong>Table 2: Comparison <strong>of</strong> costs and efficacies between <strong>the</strong> management teamand commercial weed contractors for <strong>the</strong> application <strong>of</strong> Fusilade in <strong>the</strong>Circular Head region in 2002 (estimate <strong>of</strong> costs provided by DPIWE).Control CrewManagementTeamAveragearea treated(ha./day)Efficacy(%)CostUS $/ha0.5 ~95 2,500Contractors 1.0 >90 1,120- 245 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaestuary. Management options and objectives for <strong>the</strong>Rubicon/Port Sorell and Tamar River may be revised ifsignificant developments in large-scale control techniquesare forthcoming.CONCLUSIONSBetween 1998 and 2002 Tasmania’s rice grassManagement Strategy used <strong>the</strong> experiences <strong>of</strong> rice grassmanagement programs in o<strong>the</strong>r regions, particularly those inVictoria, Australia, and Washington State, USA, to identifypriority program risks and to develop its approach to <strong>the</strong>reduction <strong>of</strong> infestations. The management program placed ahigh degree <strong>of</strong> importance on stakeholder participation andengagement at all stages <strong>of</strong> <strong>the</strong> management program.Targeted research and environmental monitoring programshave also been strategically used in developing andprogressively scaling up <strong>the</strong> management program fromsmall-scale to medium-scale control efforts.The Management Strategy and area-based ManagementPlans set specific objectives for each <strong>of</strong> Tasmania’s infestedregions. In light <strong>of</strong> <strong>the</strong> progress made toward <strong>the</strong>seobjectives <strong>the</strong> implementation <strong>of</strong> <strong>the</strong> Management Strategybetween 1998 and 2002 was successful. This achievementwas acknowledged in 2001 when <strong>the</strong> Rice Grass AdvisoryGroup was awarded <strong>the</strong> Australian Water AssociationEnvironment Award for its contribution to <strong>the</strong> restoration <strong>of</strong>aquatic habitat. Factors that significantly contributed toprogram success include:• Development <strong>of</strong> a management strategy with clearobjectives and direction for management;• Establishment <strong>of</strong> <strong>the</strong> Rice Grass Advisory Groupinvolving key stakeholder interests and <strong>the</strong> appointment<strong>of</strong> <strong>the</strong> DPIWE as lead management agency;• Effective engagement <strong>of</strong> relevant regional and industrystakeholders at all stages <strong>of</strong> <strong>the</strong> management process;• Development <strong>of</strong> a management program that integratestargeted research and adequate environmentalmonitoring; and• A method <strong>of</strong> rice grass control that can be demonstratedto be environmentally responsible, safe, practicable andcost-effective.It is worth pointing out, however, that <strong>the</strong> battle against <strong>the</strong>spread <strong>of</strong> rice grass in Tasmania is far from over. There aresome important challenges still to be faced including:• A clear understanding <strong>of</strong> what is required to achieveeradication as opposed to just controlling rice grass, andensuring that those more stringent requirements are met(e.g. local communities and industries will play animportant role in eradication and <strong>the</strong>ir continued supportand commitment is vital);• Maintaining effective, practicable and cost-effectivearrangements to contain large infestations in <strong>the</strong> longterm;• Procurement <strong>of</strong> sufficient ongoing funding to meet <strong>the</strong>expense <strong>of</strong> reducing rice grass infestations, particularlyin <strong>the</strong> Circular Head region.Table 3: Summary <strong>of</strong> all infestations in Tasmania, including size andcurrent management objectives in 2003 relative to 1997.InfestationDerwentRiver*LittleSwanport*ManagementObjectiveInfestationarea 1997(ha)Infestationarea2002 (ha)eradicate 1 0.0001eradicate 10 0.0003St Helens* eradicate 1 0.0001Bridport* eradicate 5 0.0002Tamar RiverRubicon/Port SorellCircularHeadcontain toestuarycontain to upperhalf <strong>of</strong> estuary415 415109 141eradicate 50 25* Priority infestations treated in 1999, 2000 and 2002ACKNOWLEDGEMENTSThe authors would like to acknowledge <strong>the</strong> NaturalHeritage Trust Fisheries Action Program, <strong>the</strong> Department <strong>of</strong>Primary Industries, Water and Environment, TasmanianFishing Industry Council and participating oyster growersfor contributing <strong>the</strong> resources for developing andimplementing <strong>the</strong> Management Strategy. We would also liketo thank numerous o<strong>the</strong>r people who provided advice anddirection on control techniques, monitoring programs andcollaboration with regional communities. I would personallylike to thank all <strong>the</strong> field crew, including <strong>the</strong> oyster growerswho lived “in sync” with <strong>the</strong> tidal cycle to make it happenon <strong>the</strong> ground.REFERENCES:Davies, P.E. 1999. Use <strong>of</strong> Fusilade® to Control Rice Grass:Assessment <strong>of</strong> environmental risks. Unpublished report preparedfor <strong>the</strong> Department <strong>of</strong> Primary Industries, Water andEnvironment, Tasmania, Australia.DPIWE 2002. Strategy for <strong>the</strong> Management <strong>of</strong> Rice Grass(Spartina anglica) in Tasmania, Australia. Department <strong>of</strong>Primary Industries, Water and Environment, Tasmania,Australia.Hedge, P.T., Summers, D., Dittmann and Davies, P. 1999.Toxicological Effects <strong>of</strong> Fusilade® on Pacific Oysters,Crassostrea gigas. Unpublished technical report for <strong>the</strong>Department <strong>of</strong> Primary Industries, Water and Environment,Tasmania, Australia.Hedge, P.T. and Kriwoken, L.K. 2000. Evidence for effects <strong>of</strong>Spartina anglica invasion on benthic macr<strong>of</strong>auna in LittleSwanport estuary, Tasmania. Austral Ecology 25:150-159.Hedge, P.T., Kriwoken, L.K. and Patten. K. 2003. A Review <strong>of</strong>Spartina Management in Washington State, US. Journal <strong>of</strong>Aquatic Plant Management 41:82-89Kriwoken and Hedge 2000. Exotic species and estuaries: managingSpartina anglica in Tasmania, Australia. Ocean and CoastalManagement 43:573-584.- 246 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementCONTROLLING INVASIVE SPARTINA SPP.: THE NEW ZEALAND SUCCESS STORYG. MILLERRanger – Biodiversity, Department <strong>of</strong> Conservation, Southland, New ZealandSpartina is being successfully controlled in <strong>the</strong> estuaries <strong>of</strong> sou<strong>the</strong>rn New Zealand. With anestimated 800 hectares (ha) (2000 acres [ac]) <strong>of</strong> <strong>the</strong> New River estuary near Invercargill beingaffected by Spartina in 1988, this has now been reduced to less than 1 ha (2.5 ac). Within <strong>the</strong> nexttwo years, we expect Spartina to have been eradicated in <strong>the</strong>se estuaries. Spartina was originallyplanted in <strong>the</strong> Invercargill district in <strong>the</strong> 1930s to reclaim land in <strong>the</strong> estuary for <strong>the</strong> increasedindustrial activities in <strong>the</strong> area. It wasn’t until <strong>the</strong> 1970s that Spartina was first acknowledged as aserious problem causing significant impacts on <strong>the</strong> estuarine values, including wading birds andshellfish habitats, and thus declared an invasive weed to <strong>the</strong> area. Various control techniques weretrialled starting in <strong>the</strong> early 1970s producing some good results but were discontinued in <strong>the</strong> mid1980s. In 1988 <strong>the</strong> newly formed Department <strong>of</strong> Conservation took over <strong>the</strong> responsibility for <strong>the</strong>control <strong>of</strong> Spartina and trials began to find a method to control its spread.Trials with different chemicals show Haloxyfop (registered as Gallant nf ) to be <strong>the</strong> most successfulherbicide with up to 95% kill on first application. Trials on application techniques have seen <strong>the</strong>combination <strong>of</strong> helicopter boom spraying <strong>of</strong> meadows and follow-up with an eight-wheeledamphibious-tracked vehicle (Argo), fitted with tracks and spray unit, as <strong>the</strong> most efficient methods<strong>of</strong> applying this chemical. These methods are now being used extensively around New Zealand.The New Zealand Department <strong>of</strong> Conservation has beeninvolved with <strong>the</strong> control <strong>of</strong> Spartina in Southland for <strong>the</strong>past 16 years and has had some significant successes in <strong>the</strong>fight against this invasive weed. This document summarizes<strong>the</strong> lead-up to <strong>the</strong> department’s involvement, <strong>the</strong> trial workundertaken, <strong>the</strong> results achieved to <strong>the</strong> present and <strong>the</strong> plansfor <strong>the</strong> future.Planting <strong>of</strong> Spartina started in <strong>the</strong> New River Estuaryduring <strong>the</strong> early 1930s to help reclaim parts <strong>of</strong> <strong>the</strong> estuaryfor industrial development. The initial plantings were <strong>of</strong> <strong>the</strong>sterile hybrid Spartina x townsendii, but <strong>the</strong>se plantings didnot prove very successful. In 1947, plantings <strong>of</strong> <strong>the</strong> speciesSpartina anglica were undertaken. These plantingscontinued until 1954 and proved to be very successful. In <strong>the</strong>following decades, <strong>the</strong> area covered by Spartina continuedto increase in an uncontrolled manner and spread to o<strong>the</strong>restuary environments within Southland.In <strong>the</strong> 1960s and early 1970s, concern was first raisedon <strong>the</strong> increasing spread <strong>of</strong> <strong>the</strong> infestation and <strong>the</strong> effect itwas having on wading bird habitat. In 1972, a committeewas established to look at <strong>the</strong> Spartina issue, and withfunding from <strong>the</strong> Invercargill City Council, trials wereundertaken to try to control <strong>the</strong> spread <strong>of</strong> Spartina.Early work using <strong>the</strong> herbicide Tandex (activeingredient karbutilate) by spray application provedsuccessful, but later work was less so, possibly due to siltbuild-up on Spartina leaves reducing <strong>the</strong> herbicide’seffectiveness. A method <strong>of</strong> soil injection was developed;although successful, it was very difficult, time-intensivework. Successes achieved during <strong>the</strong>se early controlapplications were due mainly to <strong>the</strong> dedication, perseveranceand plain hard work <strong>of</strong> a few key people.Aerial trials with <strong>the</strong> herbicides Herbex and Phytazol Awere undertaken, but <strong>the</strong> consistency <strong>of</strong> <strong>the</strong> results wasdisappointing. Due to <strong>the</strong> restructuring <strong>of</strong> local authoritieswithin Southland in <strong>the</strong> mid 1980s, fewer resources wereallocated to <strong>the</strong> Spartina control program. Infestations <strong>of</strong>Spartina again grew in an uncontrolled manner and so some<strong>of</strong> <strong>the</strong> earlier successes were lost.By <strong>the</strong> end <strong>of</strong> <strong>the</strong> late 1980s, it was estimated thatinfestations <strong>of</strong> Spartina within New River Estuary affectedsome 800 hectares (ha), approximately 20% <strong>of</strong> <strong>the</strong> entireestuarine area. In 1987, <strong>the</strong> newly formed Department <strong>of</strong>Conservation took responsibility for <strong>the</strong> control <strong>of</strong> Spartinawithin Southland.A new interagency partnership was developed betweenInvercargill City Council, Environment Southland (a publicenvironmental monitoring organization), and <strong>the</strong> Department<strong>of</strong> Conservation. With this approach, efforts could again befocused on methods to control Spartina, not only in <strong>the</strong> NewRiver Estuary but also in all Southland estuaries. Thispartnership continues to this day and is one <strong>of</strong> <strong>the</strong> reasonsfor <strong>the</strong> success <strong>of</strong> <strong>the</strong> program. As some <strong>of</strong> <strong>the</strong> previouslyused herbicides were now unavailable, this led to a newseries <strong>of</strong> trials being undertaken with new herbicides andapplication methods.One <strong>of</strong> <strong>the</strong> first herbicides tried was glyphosate. Whilstthis herbicide proved fairly successful and was used for a- 247 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinanumber <strong>of</strong> years, its effectiveness was <strong>of</strong>ten reduced whensilt and dirt covered Spartina leaves, although kill rates <strong>of</strong>50% to 60% were still achieved. Fur<strong>the</strong>r trials wereundertaken using <strong>the</strong> herbicide Gallant? (Haloxyfop-Rmethyl).This grass-selective herbicide was being used by<strong>the</strong> Department <strong>of</strong> Conservation at that time for <strong>the</strong> control<strong>of</strong> marram grass (Ammophila arenaria) on sand dunesystems along <strong>the</strong> Southland coast. Gallant proved veryeffective at penetrating silt and dirt with at least 90% killrates being achieved with a single application. Thissubsequently became <strong>the</strong> herbicide <strong>of</strong> choice and its use hascontinued to this day.Whilst an effective herbicide was now available, earlyapplication methods left something to be desired, provingcostly in time and resources. Initial work with a highpressurespray gun and 200 meter (m) hose mounted on atruck was slow and difficult, needing a team <strong>of</strong> six people tohaul a hose across mudflats. Trials using a hovercraft in1995 proved successful in allowing quick access to scatteredSpartina patches that previously had to be approached onfoot. This machine was fitted with a small spray unit andproved very effective. However, this form <strong>of</strong> transport wasnot able to travel across <strong>the</strong> now vast Spartina meadows. Soeven though <strong>the</strong> hovercraft had made work easier, <strong>the</strong>re werestill shortcomings. Fur<strong>the</strong>r trials were undertaken using anArgo, an all-wheel-drive amphibious vehicle, with extremelysuccessful results. When fitted with a small spray unit, thisvehicle allows one person to undertake control work across<strong>the</strong> vast majority <strong>of</strong> areas within <strong>the</strong> estuary environment.When fitted with tracks, this vehicle is capable <strong>of</strong> travellingover waist-deep mud. Whilst <strong>the</strong> Argo was an extremelyefficient vehicle, an effective method <strong>of</strong> herbicideapplication to <strong>the</strong> vast Spartina meadows was still sought.Trials with a helicopter fitted with a boom and valsecolfoaming nozzles, flying at a height <strong>of</strong> 1-2 m above <strong>the</strong> tops<strong>of</strong> <strong>the</strong> plant and at very slow speeds, proved very effectivefor this task.As a result <strong>of</strong> our experience acquired during all <strong>of</strong><strong>the</strong>se trials, we now had <strong>the</strong> right tools in <strong>the</strong> fight againstSpartina. These tools were:• an herbicide that was extremely successful in <strong>the</strong> harshestuarine environment,• a helicopter that could attack <strong>the</strong> vast meadows, and• <strong>the</strong> Argo to control smaller areas on mud and to followup on <strong>the</strong> helicopter work.The correct use <strong>of</strong> <strong>the</strong> right herbicide by <strong>the</strong> mostappropriate method at <strong>the</strong> right time has had tremendousresults, which we were able to replicate in all estuarieswithin Southland.Sixteen years ago, <strong>the</strong> control and eradication <strong>of</strong>Spartina seemed an impossible task. However, previouslessons and <strong>the</strong> development <strong>of</strong> some <strong>of</strong> <strong>the</strong> early pioneeringwork has led to <strong>the</strong> elimination <strong>of</strong> Southland’s Spartinainfestation.To achieve eradication <strong>of</strong> Spartina, it is crucial for anactive surveillance program to be undertaken to continue tosearch for previously unrecorded infestations. At <strong>the</strong> presenttime, <strong>the</strong> Spartina meadows and patches have beeneradicated from all estuaries within Southland—sparselyscattered individual plants are all that remain.Where such plants are discovered, <strong>the</strong> appropriatecontrol action needs to be taken at <strong>the</strong> appropriate time. Theuse <strong>of</strong> <strong>the</strong> Argo is now <strong>the</strong> mainstay <strong>of</strong> our treatmentoperations because it is time and resource-effective.In addition to present eradication work, monitoring <strong>the</strong>effects <strong>of</strong> Spartina removal on <strong>the</strong> estuary will need tocontinue.ACKNOWLEDGMENTSThis has been a brief overview <strong>of</strong> <strong>the</strong> work that has beenundertaken over <strong>the</strong> past 30 years to enable estuaries withinSouthland to approach <strong>the</strong> eradication <strong>of</strong> non-nativeSpartina. This would not have been achievable without <strong>the</strong>continuing support and cooperation <strong>of</strong> our partnerorganisations: Environment Southland, Invercargill CityCouncil, and Dow AgroSciences.In addition, a number <strong>of</strong> individuals have given support,advice and assistance from <strong>the</strong> beginning <strong>of</strong> this program.My thanks go to all who have been a part <strong>of</strong> it.- 248 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementCOMPARISON OF CHEMICAL AND MECHANICAL CONTROL EFFORTS FOR INVASIVESPARTINA IN WILLAPA BAY,WASHINGTONK. PATTENWashington State University Long Beach Research and Extension Unit, 2907 Pioneer Road, Long Beach, WA 98631;pattenk@wsu.eduWillapa Bay, Washington, hosts <strong>the</strong> largest invasive Spartina population in North America. Stateand federal agencies have been conducting large-scale mechanical and chemical control efforts on<strong>the</strong> affected lands during <strong>the</strong> past decade with varying degrees <strong>of</strong> success. Assessments <strong>of</strong> <strong>the</strong> longtermefficacy, cost effectiveness and ecological risk <strong>of</strong> <strong>the</strong> various control tools are presented. Nolong-term control was achieved with multiple years <strong>of</strong> disking. Crushing Spartina was onlysuccessful when <strong>the</strong> plant was driven well below <strong>the</strong> surface sediment. This occurred on s<strong>of</strong>tsediment and a thin root mat. Winter tilling provided good control, but spring and summer tillingonly marginal control. In sites where seed stalks were tilled into <strong>the</strong> sediment during <strong>the</strong> winter,<strong>the</strong>re was a solid stand <strong>of</strong> seedlings <strong>the</strong> following spring. Spartina control with glyphosate at <strong>the</strong>high hand-sprayed rates (5–8% v/v) averaged approximately 50%, with permanent control takingseveral years <strong>of</strong> re-treatment. Control from late season (September/October) hand-sprayedglyphosate was poor. Control with broadcast application <strong>of</strong> glyphosate (8.4 kilograms activeingredient per hectare (kg ai/ha)) was highly variable. Although brown-down was usually observed,<strong>the</strong>re was no permanent control except under ideal conditions (clean leaves, clean spray water andseveral days <strong>of</strong> dry time), in which case up to 74% control was achieved. Based on same seasonobservations, large-scale control appears promising using broadcast (aerial and ground) and handsprayedimazapyr. Most variability with imazapyr occurred when hand-sprayed applications failedto achieve good canopy coverage. Best control, lowest long-term cost and least ecological riskoccurred with broadcast imazapyr application in June.Keywords: glyphosate, imazapyr, tilling, crushing, modeling, eradication, SpartinaINTRODUCTIONResearch to develop better control methods for invasiveSpartina has been on-going on a global scale for many years(Frid et al. 1999; Garnett et al. 1992; Kilbride et al. 1995;Major et al. 2003; Patten 2002; Patten and Stenvall 2002;Pritchards 1995; Shaw et al. 1995). How well <strong>the</strong>se variousmethods work when implemented by large-scale controlprograms has been poorly documented. In addition, <strong>the</strong>re islittle information about how successful <strong>the</strong>se programs are inachieving <strong>the</strong>ir control or eradication goals.Willapa Bay, Washington State, USA currently hostsone <strong>of</strong> <strong>the</strong> largest Spartina alterniflora infestations in <strong>the</strong>world. In 1993 <strong>the</strong> state legislators unanimously passed <strong>the</strong>Revised Code <strong>of</strong> Washington 17.26 which mandates thatSpartina shall be eradicated. To accomplish this goal, <strong>the</strong>rehave been large-scale control efforts on Spartina in WillapaBay since 1995. Those efforts have included an array <strong>of</strong>mechanical and chemical practices implemented by state,federal and private stakeholders. Treating Spartina,however, has not led to eradication. Control has not comeclose to keeping pace with <strong>the</strong> 15 to 20% annual rate <strong>of</strong>spread (Hedge et al. 2002), despite state and federal agencyexpenditures <strong>of</strong> $500,000-$1 million to treat thousands <strong>of</strong>acres <strong>of</strong> Spartina per year (WSDA 1999 to 2003). Asuccessful eradication strategy requires that <strong>the</strong> specificcontrol practices being utilized have adequate efficacy toaccomplish <strong>the</strong> task. Unfortunately, according to a study by<strong>the</strong> Government Accounting Office (GAO 2002), mostinvasive species management programs fail to objectivelyassess, improve and realign <strong>the</strong>ir control strategies torealistically correspond to <strong>the</strong>ir eradication goals. TheSpartina control effort in Willapa Bay is no exception to thisfailure <strong>of</strong> management strategies (Hedge et al. 2002).The objectives <strong>of</strong> this study were to monitor andevaluate <strong>the</strong> most commonly used Spartina control efforts inWillapa Bay and to determine if <strong>the</strong> efficacy <strong>of</strong> those effortsis sufficient to allow state and federal agencies toaccomplish <strong>the</strong>ir eradication goals.MATERIALS AND METHODSMajor mechanical and chemical control effortsconducted by Washington State Department <strong>of</strong> Agriculture(WSDA), Washington State Department <strong>of</strong> Fish andWildlife (WDFW), Washington State Department <strong>of</strong> NaturalResources (WDNR), Willapa National Wildlife Refuge(WNWR) and private tideland owners in Willapa Bay from2001 to 2004 were identified and selected for evaluation.- 249 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinacrushing 2 yrscrushing + glyphosate38 kg/ha 1 yrglyphosate 9 kg/ha 1 yr0 20 40 60 80 100% Spartina-free quadratscrushing 1 yrcrushing +glyphosate38 kg/ha 2 yrsglyphosate 2 yrsglyphosate 38 kg/ha 1 yrFig. 1 Box whisker graphs <strong>of</strong> percent Spartina-free quadrats from differentmechanical and chemical controls used in Willapa Bay from 2001 to 2003.Box whisker graphs only presented for treatments with five or more datapoints. (mean = dotted line, median = solid line, box = 25 th and 75 thpercentiles, whiskers = 5 th and 95 th percentiles, data point = outliers).Three mechanical control practices (tilling, crushing anddisking) and four chemical control practices (glyphosate attwo rates, crush and spray, and imazapyr) were assessed.Tilling was done by <strong>the</strong> WNWR in <strong>the</strong> winter on largemeadows that were previously mowed (2000/2001) oruntreated (2004). Crushing was done by WSDA, WDNR orWDFW using tracked amphibious vehicles. Disking wasdone with a tandem disk. Crushing timing and frequencyvaried by site while disking was done only in <strong>the</strong> winter.Sites usually received multiple crushing or disking eventsper year. Imazapyr and <strong>the</strong> low rate <strong>of</strong> glyphosate (9kilograms per hectare (kg/ha)) were applied by boom usingapproximately 500 liters per hectare (l/ha) spray volume.The high glyphosate rate was applied by hand using highpressure spray guns. A 5 to 8% v/v <strong>of</strong> product was appliedwith a spray volume <strong>of</strong> about 1,500 l/ha. The estimatedglyphosate rate for hand applications averaged 38 kg/ha.Spraying <strong>of</strong> crushed sites occurred once Spartina had regrownenough to provide adequate canopy coverage.Treatment sites ranged from 0.25 ha to 500 ha in size with<strong>the</strong> majority larger than 10 ha.Attempts were made to assure that <strong>the</strong> treatment sitesselected for monitoring were relatively free <strong>of</strong> confoundingfactors from previous control efforts at <strong>the</strong> sites, and thateach site had a discrete treatment that was identifiable.Monitoring was done along multiple transects in eachtreatment site. Spartina stem and seedling density countsfrom 0.25m 2 quadrats were taken along transects. Thenumber <strong>of</strong> quadrats and type <strong>of</strong> transect varied depending on<strong>the</strong> size and shape <strong>of</strong> each treatment area, with a range <strong>of</strong> 40to 2,600 quadrats per site. Methods are fully detailedelsewhere (Patten 2003, 2004). Stem density data wereclassified as follows:1) excellent Spartina control – 0 stems/0.25m 2 ;2) good to moderate control – 1 to 5 stems/0.25m 2 ;3) fair to poor control – 6 to 20 stems/0.25m 2 , and4) poor to no control – 21 stems/0.25m 2 . Mean stemdensity before treatment was typically >50 stems/0.25m 2 .Pooled and disaggregated data for each site were analyzedfor stem density and overall stem density frequencydistribution. Means and standard error values are presented.The percent <strong>of</strong> Spartina-free quadrats was used as aconservative method to assess treatment efficacy. This valueindicates a treatment’s ability to prevent vegetative recolonizationand to minimize <strong>the</strong> cost <strong>of</strong> re-treatment. Forexample, a thinned-out Spartina canopy with a mean density<strong>of</strong> 5 stems/0.25m 2 could represent a greater than 90%decrease in stem density, but would cost almost <strong>the</strong> same totreat as a solid canopy; without re-treatment this wouldbecome a solid infestation again within a year. For this paper,<strong>the</strong> percent <strong>of</strong> Spartina-free quadrats was also assumed toapproximate <strong>the</strong> extinction coefficient or <strong>the</strong> percent control.Years to achieve eradication for a given treatment, assumingminor re-infestation, was calculated as years = log 0.01/log (1-extinction coefficient). A discrete time, logistic growth model,with growth rate and outside seedling input parameters, wasused to predict <strong>the</strong> amount <strong>of</strong> Spartina meadow remainingafter four years <strong>of</strong> a given treatment.Parametric comparisons <strong>of</strong> treatment efficacy were notfeasible because treatments lacked true replication. That is,even sites with <strong>the</strong> same general treatment were treated bydifferent agencies under vastly different conditions.Therefore to make inferences about overall treatmenteffectiveness from a management perspective, treatmentvariation is presented in box whisker format (mean, median,5 th , 25 th , 75 th and 95 th percentiles, and outliers) for alltreatments where <strong>the</strong>re were more than five data sets.For comparative purposes, box whisker graphs <strong>of</strong>efficacy on Spartina across selected herbicides are shown;this research was conducted by <strong>the</strong> author between 1998 and2003 (Patten 2002; Patten & Stenvall 2002). Percent controldata were pooled across each experimental unit forglyphosate at 18 kg/ha rate and imazapyr at 1.7 kg/ha.RESULTSIndividual and pooled results are shown in Tables 1 and2, and Fig. 1. Control with <strong>the</strong> broadcast rate <strong>of</strong> glyphosate(9 kg/ha) was highly variable (14% to 56%) and averagedhalf <strong>the</strong> efficacy achieved at <strong>the</strong> much higher hand-sprayedrates (38 kg/ha) (Table 1 and Fig. 1). There was less- 250 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementTable 1. Summary <strong>of</strong> chemical control efficacy for Spartina in Willapa Bay from 1999 to 2003.Control methodYearGeneral locationin Willapa BayStem density Percent stem density frequency(#/0.25m 2 )±std. error <strong>of</strong> mean. 0 1 to 5 6 to 20 >20Glyphosate 9 kg/ha 2003 SE 21.0±3.0 20 13 22 44Glyphosate 9 kg/ha 2003 S 27.0±1.5 14 13 19 54Glyphosate 9 kg/ha 2003 SE 7.0±1.0 25 27 35 13Glyphosate 9 kg/ha 2002 S 16.8±2.8 25 15 25 35Glyphosate 9 kg/ha 2002 S 7.5±2.0 56 16 15 15Glyphosate 9 kg/ha 2002 S 13.0±1.3 19 15 31 35Glyphosate 9 kg/ha 2002 S 1.6±0.2 55 23 15 7Glyphosate 9 kg/ha 2002/03 S 6.0±0.7 59 16 16 9Glyphosate 38 kg/ha 2002 NE 7.0±1.0 75 7 5 18Glyphosate 38 kg/ha 2002 Mid Peninsula 1.7±0.3 60 30 10 0Glyphosate 38 kg/ha 2002 N. Peninsula 2.3±1.2 57 37 3 3Glyphosate 38 kg/ha 2002 N. Peninsula 5.0±1.8 53 27 12 8Glyphosate 38 kg/ha 2002 SE 2.7±1.1 74 14 7 5Glyphosate 38 kg/ha 2003 N 5.0±1.0 32 13 42 13Glyphosate 38 kg/ha 2003 Mid Peninsula 16.0±4.0 35 10 20 35Glyphosate 38 kg/ha 2003 Mid Peninsula 5.0±1.0 58 12 18 12Glyphosate 38 kg/ha 2003 SE 0.5 ±0.5 68 32 0 0Glyphosate 38 kg/ha 2002 SE 3.4±1.1 54 28 10 8Glyphosate 38 kg/ha 2002 N. Peninsula 6.8±2.1 60 20 15 0Glyphosate 38 kg/ha 2002/03 N 9.0±2.0 35 18 33 14Glyphosate 38 kg/ha 1999/02/03 N. Peninsula 1.3±0.5 83 10 8 0Glyphosate 38 kg/ha 2000/02 S. Peninsula 5.2±0.7 63 7 18 12Crushing+glyphosate 9 kg/ha 2003 E 8.0±2.0 47 13 25 15Crushing+glyphosate 9 kg/ha 2003 E 2.0±3.0 8 8 43 41Crushing+glyphosate 9 kg/ha 2002 SE 3.1±1.2 44 46 7 2Crushing+glyphosate 9 kg/ha 2002 SE 1.5±0.5 71 19 10 0Crushing+glyphosate 9 kg/ha 2002 SE 1.6±0.5 53 40 7 0Crushing+glyphosate 9 kg/ha 2002 SE 1.9±0.6 60 20 15 0Crushing+glyphosate 9 kg/ha 2002 N. Peninsula 6.8±1.7 53 40 7 0Crushing+glyphosate 9 kg/ha 2002 SE 2.8±1.0 73 15 15 7Crushing+glyphosate 9 kg/ha 2002/03 SE 0.5±0.5 85 15 10 0Crushing+glyphosate 9 kg/ha 2002/03 SE 2.0±2.0 100 0 0 0Crushing+glyphosate 9 kg/ha 2002/03 SE 6.0±1.0 51 18 22 9Crushing+glyphosate 9 kg/ha 2002/03 N. Peninsula 8.0±1.0 35 15 40 10Crushing+glyphosate 9 kg/ha 2002/03 SE 1.0±0.2 76 8 14 2Crushing+glyphosate 9 kg/ha 2002/03 SE 5.0±1.0 43 15 37 5Imazapyr 1.7 kg/ha 2003 E 1.0±0.7 82 15 0 3variability in control at <strong>the</strong> high application rate.Glyphosate applied at <strong>the</strong> same site in consecutive years didnot provide any marked improvement in overall control overa single year <strong>of</strong> treatment (Fig. 1). Crushing followed by <strong>the</strong>application <strong>of</strong> <strong>the</strong> high glyphosate rate did not provide anybetter control than glyphosate alone (Table 1 and Fig. 1).Similarly, multiple years <strong>of</strong> this treatment did not improvecontrol over that <strong>of</strong> a single year <strong>of</strong> treatment. Control withimazapyr was <strong>the</strong> highest achieved by any treatment (82%).Although <strong>the</strong>se results were obtained from a very limiteddata set, <strong>the</strong>y are very similar to those obtained over sevenyears <strong>of</strong> small plot research (Fig. 2). In contrast, <strong>the</strong> controllevel found at monitoring sites where resource agencies usedglyphosate was considerably less than that obtained fromresearch trials.- 251 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaWinter tilling provided <strong>the</strong> highest level <strong>of</strong> uniformcontrol across all sites <strong>of</strong> any mechanical effort (77%).Control achieved with crushing was highly variable (rangingfrom 0 to 92% Spartina-free quadrats), depending on type <strong>of</strong>sediment and location (Table 2). Highest control (92%)occurred on one site where <strong>the</strong> crushing pushed <strong>the</strong> crown <strong>of</strong><strong>the</strong> plant well below <strong>the</strong> surface <strong>of</strong> <strong>the</strong> sediment. No controlwas recorded at several sites with firm sediment. Averagecontrol was achieved from crushing on s<strong>of</strong>t and firmsediment at 52% and 26%, respectively (Table 2). Crushing<strong>the</strong> same site repeatedly over consecutive years did notprovide any marked improvement in control over a singleyear <strong>of</strong> treatment (Fig. 1). Control with disking wascomparable to control from a good crushing event (Table 2).Control results from multiple years <strong>of</strong> crushing or diskingsites have a level <strong>of</strong> uncertainty because only a few siteswere available for comparison.Seedling density data were also collected at all sites(data not shown). With two exceptions, density was low (20Winter tilling – silt 2004 SE 0.4± 0.1 75 25 0 0Winter tilling – silt 2001 S 1.5±0.3 77 13 8 2Winter tilling – silt 2001/02 S 3.2±0.6 67 19 10 4Crushing –sand 2002 N. Peninsula 6.7±1.0 34 31 6 1Crushing – sand 2002 N. Peninsula 4.0±0.6 47 29 21 4Crushing – sand 2002, 2003 N. Peninsula 5.0±2.0 55 27 8 10Crushing- sand 2002, 2003 N. Peninsula 3.0±1.0 60 15 22 3Crushing- sand 2001, 2002, 2003 N. Peninsula 32.0±3.0 5 3 25 67Crushing- s<strong>of</strong>t silt 2002, 2003 NE 37.0± 2.0 0 0 5 95Crushing- s<strong>of</strong>t silt 2002, 2003 NE 7.0± 2.0 45 8 33 5Crushing - s<strong>of</strong>t silt 2001, 2002, 2003 N 3.0±1.0 50 33 13 3Crushing - s<strong>of</strong>t silt 2002, 2003 N 5.0±1.0 55 17 20 8Crushing- firm silt 2003 N 10.0±2.0 54 20 20 6Crushing -firm silt 2001 SE 46.0±3.0 0 0 0 100Crushing- firm silt 2001 SE 25.0±3.0 22 16 29 33Crushing- firm silt 2001 N. Peninsula 41.0±7.0 0 4 17 80Crushing- s<strong>of</strong>t silt 2001 SE 6.1± 0.6 37 30 28 6Crushing- s<strong>of</strong>t silt 2002 SE 5.8± 1.0 26 31 38 5Crushing- s<strong>of</strong>t silt 2001 SE 0.2± 0.1 92 8 0 0Crushing- s<strong>of</strong>t silt 2001 SE 4.0± 0.5 53 15 30 2Disking + crushing 2001, 2002 SE 4.0± 0.7 71 19 10 0Disking- s<strong>of</strong>t silt 2001 N. Peninsula 3.8±1.2 53 30 10 7- 252 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and Managementglyphosate and crushing on firm surfaces extend <strong>the</strong> timeperiod well beyond 10 years. If a more realistic andconservative approach to modeling is used that includesgrowth and seedling inputs, <strong>the</strong>n only <strong>the</strong> use <strong>of</strong> imazapyrapproaches 100% control after four years <strong>of</strong> treatment.Treatment cost per hectare is only a minor component<strong>of</strong> <strong>the</strong> total cost <strong>of</strong> a control effort. Efficacy governs <strong>the</strong>majority <strong>of</strong> <strong>the</strong> total cost. For example, treating 100 hectares<strong>of</strong> Spartina with imazapyr or glyphosate at 38 kg/ha, wouldcost approximately $60,000/ha and $120,000/ha, respectivelyat current chemical prices. These values change dramaticallywhen <strong>the</strong> modeling results shown in Table 1 are factored in.To control 99% <strong>of</strong> Spartina using imazapyr, a total <strong>of</strong> 130 hawould need to be treated over four years at a cost <strong>of</strong>approximately $78,000/ha, whereas to achieve that controlwith glyphosate, a total <strong>of</strong> 208 ha would need to be treatedover seven years at a cost <strong>of</strong> approximately $250,000/ha.DISCUSSIONThe objectives <strong>of</strong> this study were to evaluate <strong>the</strong> Spartinacontrol efforts in Willapa Bay and to determine if <strong>the</strong> efficacy<strong>of</strong> those efforts was sufficient to accomplish eradication.Developing inferences about treatment efficacy based onmonitoring data from a few sites is problematic. Unlikeresearch plots, <strong>the</strong>se treatment sites lack replication andcomplete records on all <strong>the</strong> environmental and site conditions,and in general lack application precision. Traditional “inhouse”monitoring also lacks objectivity. These concerns wereminimized within this study by pooling data collected by animpartial party (WSU) across as many similar treatment sitesand times as possible and using a very conservative estimate<strong>of</strong> control (percent Spartina-free quadrats). The tremendousvariability across sites found in this monitoring effort is notatypical <strong>of</strong> that found in o<strong>the</strong>r Spartina control efforts (Patten2002). Due to <strong>the</strong> lack <strong>of</strong> background data, however, <strong>the</strong>variations in this study are difficult to account for. Detaileddata collection on <strong>the</strong> variables that affect Spartina controlefficacy, such as herbicide dry time, tidal conditions, canopyquality (dirtiness, height, intactness/health, growth stage, andclonal age), spray water quality, and sediment type (Patten2002), are usually not collected by field crews doing <strong>the</strong>control work.Several observations on specific treatments should benoted. Tilling, although a superlative method formechanically controlling Spartina, is costly, requiring anexpensive ($250,000) amphibious tiller and is slow (~0.25ha/hr). It has a limited window during <strong>the</strong> winter when itworks, and unless it is preceded by summer mowing, itresults in massive seedling density <strong>of</strong> more than 200seedlings/m 2 . Crushing is relatively less expensive(~$40,000 to $80,000 for <strong>the</strong> equipment) and is faster (1-2ha/hr) than tilling, but it requires multiple crushing eventsper year. The greatest success in crushing appears to beTable 3. Years to achieve 99% Spartina-free tideland using repeated annualtreatment <strong>of</strong> different control practices.TreatmentExtinctioncoefficient±std. errormean *Calculatedtime toreach 99%Spartinafreetideflat**(years)Amount <strong>of</strong> 100 hameadow Spartinaremaining at <strong>the</strong>start <strong>of</strong> year 5, after 4years <strong>of</strong> treatmentand total amount <strong>of</strong>area treated duringthat time***haremainingtotal hatreatedWinter tilling 0.73±0.03 3.6 1 145Crushing ons<strong>of</strong>tsediment0.52±0.14 6.3 11 211Crushing onsand or firmsilt sedimentGlyphosate9 kg/haGlyphosate38 kg/haCrushing +glyphosate38 kg/haImazapyr 1.7kg/ha0.26±0.09 15.4 73 3600.31±0.06 12.5 52 3230.57±0.04 5.4 7 1920.59±0.04 5.3 6 1850.82 2.7 0.3 127* Extinction coefficient equals <strong>the</strong> mean % Spartina-free quadrats for <strong>the</strong>given treatment.** Assumes that treatment is repeated annually, that <strong>the</strong> equation log0.01/log (1- extinction coefficient) approximates <strong>the</strong> system’s response,and that <strong>the</strong>re are no inputs from spread and seedlings.*** Discrete time, logistic growth model, with growth rate parameter0.15, carrying capacity 100 ha and additional seedling input from outsidesources equal to 0.1% <strong>of</strong> carrying capacity per year. Developed by Dr.Caz Taylor, UC Davis.limited to sites with certain sediment characteristics, such asareas with young Spartina on s<strong>of</strong>t sediment. Crushing wellestablishedSpartina meadows with a thick root mat orSpartina on sand provides marginal control. Disking is alsorelatively inexpensive and comparable to crushing, but wasproblematic in terms <strong>of</strong> sectioning and uprooting large mats<strong>of</strong> Spartina, which re-established in deeper tidal zones. Thebroadcast rate <strong>of</strong> glyphosate, although very inexpensive,only provided control under ideal conditions (>48 hour drytime and a clean, intact canopy). These conditions were rarein <strong>the</strong> field. The hand-sprayed rate <strong>of</strong> glyphosate (with orwithout previous crushing) provided fairly consistentcontrol, but accurate rates and cost analysis are difficult toassess. Tank mixes ranged from 5 to 8% v/v <strong>of</strong> product withspray volumes from 1,000 to 3,000 l/ha spray volume.Application from airboats is limited to a few hectares perday. Variability in efficacy from hand applications largelyreflected plants that were missed or only partially covered.Although large-site monitoring data is lacking, <strong>the</strong> broadcast- 253 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaapplication <strong>of</strong> imazapyr conducted in 2004 was relativelyinexpensive (~$600/ha), fast (>100 ha/day), and fairlyefficacious.Repeated control measures conducted year after year on<strong>the</strong> same site should result in an overall reduction <strong>of</strong>Spartina at <strong>the</strong> site. This overall reduction in Spartinaoccupation over time (two years) was not evident from <strong>the</strong>pooled monitoring data for ei<strong>the</strong>r chemical or mechanicalcontrol. These results are not inconsistent with those foundin Puget Sound (Hacker et al. 2001). They report that sitesrequired consistent control for four years to obtain an 86%decrease in Spartina. Most <strong>of</strong> <strong>the</strong> sites monitored in thisstudy were also large (>10 ha) and, according to Hacker etal. (2001), less likely to show decline in Spartina densityover time than small sites. Examples <strong>of</strong> this spatial scaleeffect were evident in this study. For three discrete sites thatwere smaller than five hectares and sprayed for twoconsecutive years, <strong>the</strong> percentage <strong>of</strong> Spartina-free quadratswent from 73% to 85%, 60% to 90% and 60% to 95%.Evidently on large sites, re-growth and seedling reinfestationcompensated for any additional control achievedin subsequent years. This concurs with results from <strong>the</strong>discrete time, logistic growth model. When parameters forgrowth rate are set at 0.15, seedling input at 0.01, andcontrol rates at 50%, 10 ha <strong>of</strong> Spartina within a 100-hamudflat would still have 50% <strong>of</strong> Spartina left in <strong>the</strong>beginning <strong>of</strong> <strong>the</strong> third year. On <strong>the</strong> o<strong>the</strong>r hand, when acontrol rate <strong>of</strong> 73% is used and <strong>the</strong> parameter for outsideseedling input is reduced to 0.005, <strong>the</strong> site is projected to be81% Spartina-free in <strong>the</strong> beginning <strong>of</strong> <strong>the</strong> third year (this issimilar to <strong>the</strong> 85% found at one <strong>of</strong> our monitoring sites).Three major conclusions can be reached from thismonitoring study. First, real world control data are highlyvariable between sites and years, and tend to show resultsthat are less effective than what would be expected underideal conditions. Efficacy can vary by at least 20% from <strong>the</strong>expected level. Second, with <strong>the</strong> exceptions <strong>of</strong> tilling andimazapyr, none <strong>of</strong> <strong>the</strong> treatments provided efficacyanywhere near <strong>the</strong> level <strong>of</strong> that which would be required toachieve eradication in a reasonable time frame. Based on <strong>the</strong>political landscape, it is not prudent to expect funding for alarge-scale control effort to go beyond a six to eight-yeartime period. Thus, unless <strong>the</strong> average control rate for a givenmethod can be expected to be greater than 75%, that methodhas minimal practical value. The tenfold expansion <strong>of</strong>Spartina in Willapa Bay during <strong>the</strong> last decade <strong>of</strong> controlefforts testifies to this maxim. <strong>Third</strong>, even if <strong>the</strong> control rateis 75%, as long as <strong>the</strong>re is significant seedling input fromoutside sources, eradication can not be achieved. The controlefforts should be mounted over a scale large enough tominimize threats from new seedlings. Similarly, unless <strong>the</strong>sites continue to be treated year after year, no appreciablegains toward eradication can be achieved.ACKNOWLEDGEMENT:Funding was provided by Willapa National WildlifeRefuge and Washington State Commission for PesticideRegistration. The discrete time, logistic growth model wasdeveloped and provided by Dr. Caz Taylor, UC Davis.REFERENCESFrid, C., W. Chandrasekara, and P. Davey, 1999. The restoration <strong>of</strong>mud flats invaded by common cordgrass (Spartina anglica)using mechanical disturbance and its effects on <strong>the</strong> macrobenthicfauna. Aquatic Conservation: Marine and FreshwaterEcosystems 9:47-61.Garnett, R.P., G. Hirons, C. Evans, and D. O’Connor. 1992. Thecontrol <strong>of</strong> Spartina (cord-grass) using glyphosate. Aspects <strong>of</strong>Applied Biology 29:359-364.Government Accounting Office. 2002. Invasive Species: Clearerfocus and greater commitment needed to effectively manage <strong>the</strong>problem. GAO-03-1; www.gao.gov/cgi-bin/getrpt?GAO-03-01.Hacker, S.D., D. Heimer, S. E. Hellquist, T.G. Reeder, B. Reeves,T.J. Riordon, and M.N. Dethier, 2001. A marine plant (Spartinaanglica) invades widely varying habitats: potential mechanisms<strong>of</strong> invasion and control. Biological Invasions 3: 211-217.Hedge P., L. Kriwoken and K. Patten, 2003. A review <strong>of</strong> Spartinamanagement in Washington State, US. Journal <strong>of</strong> Aquatic PlantManagement 41: 82-90.Kilbride, K.M., F.L. Paveglio and C.E. Grue, 1995. Control <strong>of</strong>smooth cordgrass with Rodeo in a southwestern Washingtonestuary. Wildlife Society Bulletin 23:53-524.Major, W.W., III, C.E. Grue, J.M. Grassley and L.L. Conquest,2003. Mechanical and chemical control <strong>of</strong> smooth cordgrass inWillapa Bay, Washington. Journal <strong>of</strong> Aquatic PlantManagement 41: 6-12.Patten, K. 2002. Smooth cordgrass control with imazapyr. WeedTechnology 16: 826-832.Patten, K. 2003. The efficacy <strong>of</strong> mechanical treatment efforts in2001 on <strong>the</strong> control <strong>of</strong> Spartina in Willapa Bay in 2002. Progressreport submitted to <strong>the</strong> Willapa National Wildlife Refuge.(Available from <strong>the</strong> author via email [pattenk@wsu.edu].)Patten, K. 2004. The efficacy <strong>of</strong> chemical and mechanical treatmentefforts in 2002 on <strong>the</strong> control <strong>of</strong> Spartina in Willapa Bay in 2003.Progress report submitted to <strong>the</strong> Willapa National WildlifeRefuge. (Available from <strong>the</strong> author via email[pattenk@wsu.edu].)Patten, K. and C. Stenvall. 2002. Control <strong>of</strong> smooth cordgrass(Spartina alterniflora): a comparison between variousmechanical and chemical control methods for efficacy, cost andaquatic toxicity. Proc. <strong>of</strong> <strong>the</strong> 11th <strong>International</strong> <strong>Conference</strong> onAquatic Invasive Species. pp. 340-350.Pritchard, G. H. (1995). Herbicide trials on Spartina. In: J.E. Rash,R.C. Williamson, and S.J. Taylor, eds. How green is yourmudflat? <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> Australasian <strong>Conference</strong> onSpartina Control. Department <strong>of</strong> Conservation and NaturalResources, Yarram, Victoria, pp. 66.Shaw, W. B. and D.S. Gosling. 1995. Spartina control in NewZealand—an overview. In: J.E. Rash, R.C. Williamson, and S.J.Taylor, eds. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> Australasian <strong>Conference</strong> onSpartina Control Melbourne, Australia: Victorian GovernmentPublication. pp. 43-60.WSDA. 1999-2003, Spartina Eradication Program—LegislativeReports, http://agr.wa.gov/PlantsInsects/Weeds/Spartina.- 254 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementFRAGMENT PROPAGULES OF SPARTINA ALTERNIFLORA AND POTENTIAL EASTERNPACIFIC DISPERSALV. H. MORGAN 1 AND M. SYTSMA 21, 2Portland State University, Dept <strong>of</strong> Environmental Science and Management, Center for Lakes & Reservoirs, P.O. Box 751,Portland, OR 97207-0751;1vhoward@pdx.eduCommonly used mechanical control methods for Spartina alterniflora involve varying levels <strong>of</strong>disturbance to rhizomes and roots. We examined <strong>the</strong> viability <strong>of</strong> rhizome fragments and <strong>the</strong>irpotential role in dispersal. Production <strong>of</strong> rhizome fragments by rototilling in Willapa Bay,Washington, USA was studied. The top 10 centimeters (cm) <strong>of</strong> <strong>the</strong> sediment contained an average <strong>of</strong>310 fragments per square meter (m 2 ). Median rhizome length was 3.7 cm. Eighty-seven percent <strong>of</strong><strong>the</strong> rhizome fragments had at least one vegetative shoot attached. Survivorship <strong>of</strong> S. alterniflorarhizome fragments from Willapa Bay and San Francisco Bay populations was investigated using athree-way factorial design. Treatments included two fragment sizes, approximating those found inWillapa Bay, immersed in ei<strong>the</strong>r freshwater, 15 parts per thousand (ppt) saltwater, or 35 pptsaltwater for 3, 8 or 15 days. Fragments were <strong>the</strong>n individually planted and grown in greenhouseponds for four months. Rhizome survivorship was low (8.6% or less) in all 35 ppt treatments.Survivorship was 37.3% and 87.5% in 15 ppt and freshwater treatments, respectively. Largerhizomes had higher survivorship than small rhizomes at all salinities, and <strong>the</strong> length <strong>of</strong> time <strong>the</strong>rhizome fragments were immersed prior to planting had a variable effect on survivorship. Resultssuggest rototilling for control <strong>of</strong> Spartina may spread <strong>the</strong> infestation within an estuary but isunlikely to result in spread to o<strong>the</strong>r estuaries by ocean transport. Thus, tilling should be used withcaution in estuaries with small, isolated populations <strong>of</strong> Spartina.Although ocean transport <strong>of</strong> rhizome fragments appears to be a small risk, ocean transport <strong>of</strong> wrackand viable S. alterniflora seed is likely. A drift card study was begun in late September 2004 tobetter understand potential dispersal from invaded west coast estuaries. Monthly releases <strong>of</strong> cardsfrom Humboldt and San Francisco bays in California, as well as Willapa Bay, Washington will aididentification <strong>of</strong> wrack deposition sites. Data from <strong>the</strong> first two months <strong>of</strong> this year-long studyindicate that long distance dispersal <strong>of</strong> up to 270 kilometers (km) over a four-week period can occur.Keywords: Spartina alterniflora, rhizome fragment, propagule dispersal, drift card simulationINTRODUCTIONEffective invasive plant management considers potentialvectors <strong>of</strong> propagules as well as how to minimize propaguleproduction. In <strong>the</strong> case <strong>of</strong> Spartina alterniflora, earlydetection and treatment efficacy are high priorities for manystakeholders wanting to preserve historic habitat, indigenousspecies, and o<strong>the</strong>r beneficial uses <strong>of</strong> mudflats and native saltmarshes in <strong>the</strong> Pacific Northwest. Within <strong>the</strong> core infestationsites, thousands <strong>of</strong> hectares have already been colonizedincluding an estimated 790 net hectares (ha) (1,960 acres[ac]) in San Francisco Bay, California (Zaremba andMcGowan 2004) and 3,200 net ha (8,000 ac) in WillapaBay, Washington. Additionally, thousands <strong>of</strong> hectares in 31Pacific estuaries are at risk for future colonization by one ormore invasive Spartina spp. (Daehler & Strong 1996; Pfau<strong>the</strong>t al. 2003). In Oregon alone, approximately 13,622 ha(33,660 ac) <strong>of</strong> intertidal mudflats and aquatic beds arevulnerable to invasion (Pfauth et al. 2003). Understandingboth <strong>the</strong> potential risks and <strong>the</strong> efficacy <strong>of</strong> any controlmethod is critical to refining management choices and earlydetection efforts.Efficacy and cost data have been evaluated for a widearray <strong>of</strong> chemical and mechanical treatments, (Patten 2002;Hedge et al. 2003; Pfauth et al. 2003); however <strong>the</strong> riskassessments <strong>of</strong> non-target effects have focused morenarrowly on chemical controls. Mechanical treatments suchas rototilling, disking, crushing, pulverizing and digginghave been problematic due <strong>the</strong>ir slow pace, variableefficacy, and high cost per area treated (Patten 2002). Yetrototilling and disking are still used in some situations t<strong>of</strong>acilitate <strong>the</strong> decomposition <strong>of</strong> below-ground biomass aftersuccessful chemical treatment, which allows for more rapidrestoration to usable shorebird habitat (Patten & Stenvall2002); <strong>the</strong>se mechanical methods have also been employedwhere landowners oppose chemical treatment options.- 255 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaCordgrasses are capable <strong>of</strong> reproducing by vegetativefragments (Landin 1990; Stiller & Denton 1995; Daehler &Strong 1996; Sayce et al. 1997; Patten & Stenvall 2002).Disturbances to Spartina’s extensive below-groundstructure, such as those caused by rototilling, couldpotentially produce rhizome fragments. While invasivecordgrasses are known for <strong>the</strong> resiliency <strong>of</strong> <strong>the</strong>ir rhizomes(Reeder & Hacker 2004; Patten 2003), <strong>the</strong> viability <strong>of</strong>mechanically produced rhizome fragments and <strong>the</strong>ir role indispersal has not been closely examined (i.e., evidence isanecdotal) (Randall & Milne unpublished; Pfauth et al.2003).Research has focused on <strong>the</strong> sexual reproductivecapacity <strong>of</strong> Spartina (Broome et al. 1974; Sayce 1988;Daehler 1996; Plyer & Proseus 1996; Sayce & Dumbauld1997; Daehler 1999; Davis et al. 2004), ra<strong>the</strong>r than asexualproduction, since this is considered to be <strong>the</strong> primary source<strong>of</strong> new clones (Stiller and Denton 1995; Sayce et al. 1997).Seed as well as rhizome fragments could disperse Spartinalocally or across long distances if carried by tides and oceancurrents (Daehler and Strong 1996; Stenvall and Patten2002, Pfauth et al. 2003). Repeated reports <strong>of</strong> Spartinafragments washing ashore near Ft. Stevens (near Astoria,Oregon) suggest transport <strong>of</strong> wrack from nearby WillapaBay, Washington (Grevstad and Graves pers. comm.;Howard et al. unpublished report 2004). Huiskes et al.(1995) collected seeds <strong>of</strong> S. anglica in floating and standingnets in a tidal salt marsh in <strong>the</strong> Ne<strong>the</strong>rlands. Eighty-eightpercent <strong>of</strong> <strong>the</strong> seeds collected were captured in floating nets,indicating that tidal transport <strong>of</strong> seed was primarily on <strong>the</strong>water surface ra<strong>the</strong>r than along <strong>the</strong> sediment. In an earlierstudy in <strong>the</strong> same location, Koutsaal et al. (1987) releaseddyed sunflower seeds on outgoing and incoming tides totrack tidal movement <strong>of</strong> seeds in <strong>the</strong> salt marsh. Seeds werefound as far as 45 km away within one week <strong>of</strong> release. Thefinal location <strong>of</strong> seeds was determined by <strong>the</strong> wind velocityand direction as well as by tidal currents.Oregon’s Spartina Response Plan (Pfauth et al. 2003)was developed to prevent <strong>the</strong> introduction and spread <strong>of</strong> anySpartina species in Oregon. Areas requiring fur<strong>the</strong>r researchwere identified , including an investigation <strong>of</strong> <strong>the</strong> likelihood<strong>of</strong> <strong>the</strong> ability <strong>of</strong> Spartina fragments to resprout and anexamination <strong>of</strong> potential transport <strong>of</strong> propagules via oceancurrents.Preliminary results from three studies addressing <strong>the</strong>seresearch needs are presented here. First, a field study wasperformed to assess <strong>the</strong> production <strong>of</strong> Spartina fragments byrototillng. Secondly, a greenhouse experiment examined <strong>the</strong>ability <strong>of</strong> rhizome fragments to resprout. <strong>Third</strong>ly,preliminary data from a propagule dispersal study arepresented.MATERIALS AND METHODSField Study <strong>of</strong> Rototilling EffectsSamples were collected on January 16, 2004 along <strong>the</strong>south shore <strong>of</strong> <strong>the</strong> Naselle River, which flows into <strong>the</strong>sou<strong>the</strong>astern end <strong>of</strong> Willapa Bay, Washington. Staff <strong>of</strong>Willapa National Wildlife Refuge (NWR) was mechanicallytreating <strong>the</strong> site between high tides with a rototillingattachment towed by a Wilco amphibious vehicle. TheWilco operator made single passes within a solid meadow <strong>of</strong>Spartina alterniflora and tilled to a depth <strong>of</strong> approximately15 cm. Immediately following rototilling, three quadrats(0.25 m 2 ) were randomly chosen approximately 30 m apartand excavated to a depth <strong>of</strong> 10 cm. Excavated material wasrinsed clean and all fragments were measured for culmlength, number <strong>of</strong> culms, and rhizome diameter and length.Fragments were divided into two rhizome sizes, small andlarge, by <strong>the</strong> median value for rhizome length. The meanvalue <strong>of</strong> each <strong>of</strong> <strong>the</strong>se rhizome class sizes was <strong>the</strong>n roundedto <strong>the</strong> nearest half centimeter and used as <strong>the</strong> experimentalrhizome sizes for <strong>the</strong> greenhouse study <strong>of</strong> fragment viability.Greenhouse StudyA 2x3x3 factorial design was used to evaluatesurvivorship <strong>of</strong> S. alterniflora fragments. Factors were initialrhizome size (large or small), immersion duration (3, 8 or 15days), and salinity (freshwater, 15 ppt, or 35 ppt). Samplesfrom two populations were compared. Rhizome fragmentsfrom San Francisco were collected on March 26, 2004 from<strong>the</strong> shoreline <strong>of</strong> Elsie Roemer Bird Sanctuary on AlamedaIsland (San Francisco Bay, California). Two to threesamples were dug from each <strong>of</strong> ten clones. Samples fromWillapa Bay were collected on April 5, 2004 from fourriverbank locations (two along <strong>the</strong> Naselle River, one on <strong>the</strong>Niawiakum River, and one on <strong>the</strong> Palix River). Seven to tensamples were dug from each location. Sampling locationshad not been subjected to any previous chemical ormechanical treatment.Samples were returned to Portland State University andrinsed clean <strong>of</strong> all mud and organic matter within two days<strong>of</strong> field collection. Within 20 minutes <strong>of</strong> rinsing, fragmentswere cut to fit one <strong>of</strong> two rhizome class sizes (large ~7.5 cmor small ~2.5 cm). Fragments were <strong>the</strong>n placed in openplastic tubs containing water at 0 ppt, 15 ppt or 35 ppt(Instant Ocean ® aquarium salts). Tubs were maintainedunder ambient greenhouse conditions. Salinityconcentrations were monitored daily and adjusted as needed.After floating for a period <strong>of</strong> 3, 8 or 15 days (referred to asimmersion duration), each fragment was measured todetermine rhizome length, rhizome diameter, number <strong>of</strong>attached culms and culm length. The 8-day immersionduration was eliminated from <strong>the</strong> San Francisco treatmentdesign due to limited plant material. Fragments were <strong>the</strong>nindividually potted in six-inch diameter pots with a sterile- 256 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and Managementpotting medium and each pot placed into a wet bed (1.83 mx 2.44 m wood-framed beds lined with three layers <strong>of</strong> 6 mmclear plastic) containing ei<strong>the</strong>r 0 ppt, 15 ppt or 35 ppt salinewater to a depth <strong>of</strong> 10 cm. A total <strong>of</strong> six wet beds werecreated (two at each salinity level) with three utilized for <strong>the</strong>San Francisco fragments and three for <strong>the</strong> Willapa Bayfragments. Salinity <strong>of</strong> <strong>the</strong> wet beds was monitored every oneto three days and adjusted with saline or fresh water asneeded.A total <strong>of</strong> 234 fragments, all having at least one culm,were potted from <strong>the</strong> San Francisco samples. A total <strong>of</strong> 353fragments having at least one culm were potted from <strong>the</strong>Willapa Bay samples. An additional 116 fragments with noculms attached were created from <strong>the</strong>se samples. Thepurpose <strong>of</strong> <strong>the</strong>se fragments was to test if survival wasdependent upon <strong>the</strong> presence <strong>of</strong> at least one culm assuggested by Randall and Milne (unpublished). Small andlarge rhizome fragments were immersed in <strong>the</strong> saline baths(0 ppt, 15 ppt or 35 ppt) for ei<strong>the</strong>r 4 or 16 days. Theimmersion duration treatments for <strong>the</strong>se culmless fragmentsTable 1: S. alterniflora rhizome fragment metrics following single-pass,winter rototilling effects in Willapa Bay.MeasureMean ± SDNumber <strong>of</strong> fragments per 0.25 m 2 x 0.1m deep 77.7 ± 13.7Percentage with ≥ 1 culm 87.8 ± 3.86Rhizome length (cm) 4.96 ± 3.38Rhizome diameter (cm) 0.58 ± 0.29Culm length (cm) 5.37 ± 3.58Culms per fragment 1.31 ± 0.89were extended by 1 day to allow adequate time for planting<strong>of</strong> <strong>the</strong> culm treatment groups. They were <strong>the</strong>n individuallypotted in <strong>the</strong> same sterile potting medium and placed in <strong>the</strong>wet beds.Pots were randomly positioned within <strong>the</strong> wet beds. All6.50.8Rhizome Length (cm)6.05.55.04.5Rhizome diameter (cm)0.70.60.54.00.412Quadrat312Quadrat381.6Culm Length (cm)765# <strong>of</strong> culms1.51.41.31.241.112Quadrat31.012Quadrat3Fig. 1: Interval plots by quadrat (with 95% C.I.) <strong>of</strong> rhizome length, rhizome diameter, culm length and number <strong>of</strong> culms per rhizome fragment foundimmediately following single pass, winter rototilling in Willapa Bay, WA.- 257 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaFrequency60504030202.47.5a% survivalSan Francisco1008060402000 30 60 90 120days post-planting100plants were exposed to ambient light and temperatureconditions for 132 days after planting. Survival, culm lengthand <strong>the</strong> number <strong>of</strong> culms were recorded at 30, 51, 74, 95,116 and 132 days after planting. Survival was defined as <strong>the</strong>presence <strong>of</strong> at least one green culm. Roots, rhizomes, culmsand inflorescences were separated 132 days after plantingand <strong>the</strong>ir fresh weight (fw) recorded. Dry weight (dw) wasobtained after drying <strong>the</strong> roots, culms and inflorescences to aconstant weight in <strong>the</strong> greenhouse and oven drying <strong>the</strong>rhizomes at 65-70°C for 48 hours.Propagule Dispersal StudyMonthly releases <strong>of</strong> buoyant, biodegradable woodendrift cards began in September 2004 from <strong>the</strong> mouths <strong>of</strong>Willapa Bay in Washington and Humboldt and SanFrancisco bays in California. A total <strong>of</strong> 600 cards arereleased each month—200 each per bay. Releases occurredwithin two hours after high tide to ensure an outgoingcurrent. Each batch <strong>of</strong> cards was printed with a unique codedenoting <strong>the</strong> location, month and year <strong>of</strong> <strong>the</strong> release as wellas reporting instructions and contact information. Velocityestimates were made under <strong>the</strong> assumption that <strong>the</strong> recoverydate was <strong>the</strong> same as date <strong>the</strong> card washed ashore and that<strong>the</strong> card followed a straight line <strong>of</strong> travel.RESULTS36Field Study <strong>of</strong> Rototilling EffectsThe mean fragment density was 310 (± 54.8)/m 2 within10 cm <strong>of</strong> <strong>the</strong> surface. Of <strong>the</strong>se, 87.7% had at least one culmattached (Table 1). No plant material o<strong>the</strong>r than S.alterniflora was present in any <strong>of</strong> <strong>the</strong> plots.One-way Analysis <strong>of</strong> Variance (ANOVA) (α=0.05)comparing quadrats were performed on rhizome length andculm length (log(x+1) transformed) and rhizome diameter91215Rhizome Length (cm)Fig. 2 Rhizome sizes found immediately following single-pass winterrototilling in Willapa Bay, WA.b% survivalWillapa Bay1008060402000 30 60 90 120days post-plantingFig. 3. Percent S. alterniflora fragment survival over time for (a) SanFrancisco and (b) Willapa Bay plants. Treatment groups are delineated bysalinity (- - - - - for 0 ppt, ––––– for 15 ppt, and –– - - –– for 35 ppt), andrhizome size (plain line = small, = large).((x+1) 1/2 transformed) and number <strong>of</strong> culms per fragment.The quadrats did not vary significantly in rhizome length(p=0.240) or <strong>the</strong> number <strong>of</strong> culms per rhizome fragment(p=0.322). There were significant differences betweenquadrats with regard to culm length (p

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and Managementab% survival% survival100806040200100806040200San Francisco3 day 15 day 3 day 15 daysmallWillapa Baylarge3 day 8 day 15 day 3 day 8 day 15 daysmalllargeFig. 4: Percent fragment survival 132 days after planting <strong>of</strong> S. alterniflorafrom a) San Francisco and b) Willapa Bay. White = 0 ppt, striped = 15 ppt,black = 35 ppt. All fragments represented here were planted with at leastone attached culm.The proportion <strong>of</strong> surviving fragments per treatmentgroup stabilized by <strong>the</strong> end <strong>of</strong> <strong>the</strong> growing period (Fig. 3).For San Francisco (SF) plants, survival 132 days afterplanting ranged from 45-100% in <strong>the</strong> freshwater treatments,55-90.0% in 15 ppt water and 0 to 5.56% in 35 ppt water.For WB plants, survival ranged from 26.3-81.8% in <strong>the</strong>freshwater treatments, 27.8-88.9% in 15 ppt water and0-11.1% in 35 ppt water.For nearly all SF groups, rhizome fragments immersedfor 3 days prior to planting showed lower rates <strong>of</strong> survivalthan those immersed for 15 days (Figure 4). The samepattern emerged with <strong>the</strong> WB 15 ppt fragments where <strong>the</strong>three-day immersion groups showed much lower survivalthan <strong>the</strong> eight-day immersion groups. Large rhizomefragments consistently had higher viability than smallrhizome fragments. For SF plants, 80.3% <strong>of</strong> large fragmentssurvived compared to only 62.5% <strong>of</strong> <strong>the</strong> small fragments.For WB plants, <strong>the</strong> difference was more pronounced with74.8% <strong>of</strong> large and 36.2% <strong>of</strong> small fragments surviving. SFand WB populations were compared using a two-tailed test<strong>of</strong> two proportions; <strong>the</strong>re were significant differencesbetween <strong>the</strong> proportions surviving in both freshwater (72.5%vs. 55.5%, respectively, p=0.012, α=0.05) and 15 ppt water(70.1% vs. 56.0%, p=0.043, α=0.05). In both <strong>of</strong> <strong>the</strong>secomparisons, SF fragments had higher survivorship. Therewas no notable difference between survival rates for <strong>the</strong> twoTable 2: Summary data for two months <strong>of</strong> drift card releases from three S.alterniflora infested bays.Release Date Willapa HumboldtSeptember 2004SanFranciscoRecovery Rate 57.5% 21.5% 30.5%Quantity <strong>of</strong> recovered cardsNorthSouthInside bayMax distance traveled (km)NorthSouthOctober 20047934295351031275651219304520Recovery Rate 29.5% 0.5% 24.0%Quantity <strong>of</strong> recovered cardsNorthSouthInside bayMax distance traveled (km)NorthSouth563022335locations (3.9% vs. 5.1%) in <strong>the</strong> high salinity treatment (p=0.691, α=0.05).PROPAGULE DISPERSAL STUDYDrift card return rates have been over 20% for five <strong>of</strong><strong>the</strong> six releases performed as <strong>of</strong> November 11, 2004 (Table2). Cards have consistently been found both to <strong>the</strong> north andsouth <strong>of</strong> each release location. In four <strong>of</strong> <strong>the</strong> six releasesperformed to date, cards have been found inside <strong>the</strong>estuaries. The majority <strong>of</strong> cards are staying within 25 km <strong>of</strong><strong>the</strong> release locations, although a few have traveled longerdistances.In Willapa Bay, over 69% <strong>of</strong> <strong>the</strong> September releasecards and 95% <strong>of</strong> <strong>the</strong> October release cards were found to<strong>the</strong> north <strong>of</strong> <strong>the</strong> bay. Maximum northward velocities forSeptember and October releases were 6.9 and 11.2centimeters per second (cm/s) respectively, while maximumsouthward velocities reached 6.7 and 3.4 cm/s.In Humboldt Bay, approximately 72% <strong>of</strong> <strong>the</strong> Septembercards were carried south. Maximum velocity for Septemberrelease was approximately 3.4 cm/s both to <strong>the</strong> north andsouth. Only one card, found after six days approximately sixkm north <strong>of</strong> <strong>the</strong> Humboldt Bay entrance, was recovered fromHumboldt’s October release.1006na64117530- 259 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaThe highest number <strong>of</strong> cards found within an estuaryoccurred in <strong>the</strong> case <strong>of</strong> <strong>the</strong> San Francisco September releasewhen thirty cards were recovered on <strong>the</strong> eastern edge <strong>of</strong> <strong>the</strong>bay, mainly near <strong>the</strong> Berkeley and Albany shoreline.Observed winds at <strong>the</strong> time <strong>of</strong> that release were from <strong>the</strong>west at approximately 7.7 – 10.3 m/s. Recoveries <strong>of</strong> SanFrancisco cards not blown back into <strong>the</strong> bay after <strong>the</strong>September release showed 61% were carried south.Maximum velocities for this release were approximately 2.4cm/s to both <strong>the</strong> north and south. Eighty-seven percent <strong>of</strong>October cards were found to <strong>the</strong> south. Maximum velocitieswere 5.4 cm/s to <strong>the</strong> north and 6.9 cm/s to <strong>the</strong> south.DISCUSSIONRototilling appears to have fairly uniform cutting actionon solid meadows; it produces rhizome fragments <strong>of</strong>consistent size and with similar numbers <strong>of</strong> attached culms.A high percentage (87.8%) <strong>of</strong> <strong>the</strong> S. alterniflora fragmentsproduced by rototilling in Willapa Bay had at least oneattached culm. The observed difference in culm length andrhizome diameter between quadrats was likely a result <strong>of</strong>variations in <strong>the</strong> age <strong>of</strong> coalesced clones, ra<strong>the</strong>r thanvariable tearing action by <strong>the</strong> tilling blades. Assuminguniformity in <strong>the</strong> production <strong>of</strong> rhizome fragments(approximately 312 fragments/m 2 within <strong>the</strong> top 10 cm) wecould make a conservative estimate that 0.5% <strong>of</strong> fragmentsmight be loosened by wave or tidal action, becomingsuspended in <strong>the</strong> water column. Based on those assumptions,as many as 15,600 fragments might be distributed in <strong>the</strong>open water for every hectare rototilled (~6,300 per acre).Viability <strong>of</strong> rhizomes after floating at <strong>the</strong> water surfaceseems to be primarily dependent on <strong>the</strong> presence <strong>of</strong> at leastone culm. Randall and Milne (unpublished) found thatrhizome fragments 2.5 to 15 cm long with no attached culmshad 100% mortality regardless <strong>of</strong> position on <strong>the</strong> mud-flatsubstrate or beneath it at various depths. The presence <strong>of</strong>attached culms should allow <strong>the</strong> fragment to respire and<strong>the</strong>reby increase its chance <strong>of</strong> survival. In fact, repeatedmowing to remove vegetative shoots caused a reduction <strong>of</strong>oxygen to <strong>the</strong> root system and was initially utilized as acontrol method for Spartina (Ebasco Environmental 1993;Hedge et al. 1997). Of <strong>the</strong> 116 rhizomes planted withoutattached culms during <strong>the</strong> greenhouse study, none survived.Compared to fragments planted with culms, this findingseems to support <strong>the</strong> <strong>the</strong>ory that at least one vegetative shootis needed for survival <strong>of</strong> vegetative propagules.For fragments having culms, salinity and initial rhizomesize determined survival. The 35 ppt treatment resulted innotably poorer survival rates compared to <strong>the</strong> lower salinitytreatments. Differences in 0 ppt and 15 ppt treatmentsappeared largely due to initial rhizome size. Larger rhizomeswould logically have higher chances <strong>of</strong> survival since <strong>the</strong>ywould be likely to have more nodes, a greater number <strong>of</strong>established roots and more non-structural carbohydrates t<strong>of</strong>uel new growth.The length <strong>of</strong> immersion was a less importantdeterminant <strong>of</strong> survival and establishment or rhizomefragments than salinity or rhizome size. Longer immersiondurations may increase viability, although this effect was notconsistent. Three to four months <strong>of</strong> wet, cool conditions mayhelp break seed dormancy and increase germination rates byleaching a germination inhibitor. Similarly, <strong>the</strong> conditionsthat fragments are exposed to while floating in open watermay retard growth <strong>of</strong> pathogens, encourage shoot productionor elongation or o<strong>the</strong>rwise increase chances <strong>of</strong> survival.Repeated monitoring <strong>of</strong> treated sites in Willapa Bay hasshown that mechanical treatments such as rototilling anddisking have higher efficacy during <strong>the</strong> period <strong>of</strong> Decemberthrough February (Patten & Stenvall 2002). All <strong>of</strong> <strong>the</strong> plantsused for this study were collected four to five weeks after <strong>the</strong>normal rototilling period in Willapa Bay. Increased culmlength, as well as higher air, soil and water temperatures at<strong>the</strong> time <strong>of</strong> collection, may have increased survival rates.Additionally, <strong>the</strong> vigorous action <strong>of</strong> rototilling producesmore ragged edges and somewhat damaged culms than werereproduced in <strong>the</strong> greenhouse. This might also elevate rates<strong>of</strong> survival shown here.Preliminary results from <strong>the</strong> first two months <strong>of</strong> <strong>the</strong> driftcard study suggest propagule deposition from infestedestuaries lessens with increased distance. Flow over <strong>the</strong>coastal shelf is predominantly poleward in <strong>the</strong> winter andearly spring, with mean current velocities <strong>of</strong> 20 cm/s.Summertime flow is typically southward with meanvelocities <strong>of</strong> 10 cm/s. Dispersal patterns seen from <strong>the</strong>sepreliminary findings may be due to a seasonal transitionperiod between <strong>the</strong>se predominant currents. Recoverypatterns may also reflect wind forcing and local eddies from<strong>the</strong> mouths <strong>of</strong> <strong>the</strong> release estuaries. Frequent recoveries inbeaches along Long Beach peninsula, where Spartina wrackis commonly found in <strong>the</strong> fall, suggests that <strong>the</strong> cardssimulate wrack dispersal with some accuracy.CONCLUSIONSThe estimate <strong>of</strong> 15,600 fragments per hectare may seeminconsequential when compared to estimates <strong>of</strong> seedgermination rates for S. alterniflora which range from nineto nineteen million seeds per hectare (3.7 million to 7.7million seeds per acre) (Callaway 1990; Daehler and Strong1994). However, understanding <strong>the</strong> risks associated with allcontrol methods is necessary when site-specific treatmentdecisions are made. Rototilling or o<strong>the</strong>r mechanicaldisturbance that produces fragments larger than 2.5 cmshould be used with caution in areas with fresh tomoderately brackish (mesohaline) waters. Mechanicaldisturbance following some o<strong>the</strong>r treatment method, such asherbicide application, may pose less <strong>of</strong> a risk <strong>of</strong> starting new- 260 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and Managementclones. If <strong>the</strong> infestation is isolated and/or <strong>the</strong> population isnot setting seed, caution may be warranted in usingrototilling or similar mechanical treatments since it couldproduce viable propagules in areas with few o<strong>the</strong>r loci <strong>of</strong>dispersal.A greater understanding <strong>of</strong> <strong>the</strong> dispersal patterns frominfested bays should help to identify <strong>the</strong> risk <strong>of</strong> seed orvegetative propagule transport. This data, combined withknown characteristics <strong>of</strong> susceptible habitat, will helpidentify natural deposition sites <strong>of</strong> invasive Spartina spp.ACKNOWLEDGEMENTSThanks to Debra Ayres, University <strong>of</strong> California DavisDepartment <strong>of</strong> Evolution & Ecology, and Erik Grijalva, SanFrancisco Estuary Invasive Spartina Project for <strong>the</strong>irassistance in California; Terri Butler and Jonathan Bates <strong>of</strong>Willapa National Wildlife Refuge, Fritzi Grevstad and KimPatten <strong>of</strong> Washington State University for <strong>the</strong>ir suggestionsand field assistance in Washington. For volunteering torelease drift cards, we thank David Heimer, Les Holcomband Travis Haring <strong>of</strong> Washington State Department <strong>of</strong> Fish& Wildlife; Bill Pinnix, U.S. Fish and Wildlife Service inArcata, CA and Kristen Ward, National Park Service,Golden Gate Recreation Area. Additionally, <strong>the</strong> staff andmany students at <strong>the</strong> Center for Lakes & Reservoirs,Portland State University have been immensely helpfulsuggestions and field/greenhouse work.REFERENCESBroome, S.W., W.W. Woodhouse, and E.D. Seneca. 1974. Propagation<strong>of</strong> smooth cordgrass, Spartina alterniflora, from seed inNorth Carolina. Chesapeake Science 15(4):214-221.Daehler, C.C. 1999. Inbreeding depression in smooth cordgrass(Spartina alterniflora, Poaceae) Invading San Francisco Bay.American Journal <strong>of</strong> Botany 86(1):131-139.Daehler, C.C. and D.R. Strong. 1996. Status, prediction and prevention<strong>of</strong> introduced cordgrass Spartina spp. invasions in Pacificestuaries, USA. Biological Conservation 78:51-58.Davis, H.G., C.M. Taylor, J.C. Civille and D.S. Strong. 2004. AnAllee effect at <strong>the</strong> front <strong>of</strong> a plant invasion: Spartina in a Pacificestuary. Journal <strong>of</strong> Ecology 92:321-327.Hedge, P., L.K. Kriwoken, and K. Patten. 2003. A review <strong>of</strong>Spartina Management in Washington State, US. Journal <strong>of</strong>Aquatic Plant Management 41:82-90.Howard, V., D. Isaacson, and M. Sytsma. 2004.: Detection <strong>of</strong>Spartina spp. on <strong>the</strong> Oregon Coast in 2003. Oregon Department<strong>of</strong> Agriculture Report (unpublished).Huiskes A.H.L., B.P. Koutstaal, P.M.J. Herman, W.G. Beeftink,M.M. Markusse and W. De Munck. 1995. Seed dispersal <strong>of</strong> halophytesin tidal salt marshes. Journal <strong>of</strong> Ecology 83:559–567Koutstaal, B.P., M.M. Markusse, and W. de Munck. 1987. Aspects<strong>of</strong> seed dispersal by tidal movements. In: Huiskes, A.H.L.,C.W.P.M. Blom and J. Rozema, eds. Vegetation Between Landand Sea. Boston: Dr. W. Junk Publishers.Landin, M. 1990. Growth habits and o<strong>the</strong>r considerations <strong>of</strong>smooth cordgrass, Spartina alterniflora Loisel. In: Mumford Jr.,T.F., P. Peyton, J.R. Sayce and S. Harbell, eds. Spartina WorkshopRecord. Seattle, WA: pp. 15-20.Patten, K. and C. Stenvall. 2002. Nothin’ could be fina’ than <strong>the</strong>killin’ o Spartina.” Agrichemical and Environmental News,Washington State Cooperative Extension Newsletter. 196.Patten, K. 2002. The efficacy <strong>of</strong> mechanical treatment efforts in2001 on <strong>the</strong> control <strong>of</strong> Spartina in Willapa Bay in 2002. Progressreport to <strong>the</strong> Willapa National Wildlife Refuge.Patten, K. 2003. Persistence and non-target impact <strong>of</strong> imazapyrassociated with smooth cordgrass control in an estuary. Journal<strong>of</strong> Aquatic Plant Management 41:1-6.Pfauth, M., M. Sytsma and D. Isaacson. 2003. Oregon SpartinaResponse Plan. Portland State University, Oregon.Plyler, D.B. and T.E. Proseus. 1996. A comparison <strong>of</strong> <strong>the</strong> seeddormancy characteristics <strong>of</strong> Spartina patens and Spartina alterniflora(Poaceae). American Journal <strong>of</strong> Botany 83(1):11-14.Randall, P. and D. Milne. 1996. Loose Spartina alterniflora rhizomesdon’t sprout new plants. Unpublished report. EvergreenState College, Washington.Reeder, T.G., and S.D. Hacker. 2004. Factors contributing to <strong>the</strong>removal <strong>of</strong> a marine grass invader (Spartina anglica) and subsequentpotential for habitat restoration. Estuaries 27(2):244-252.Sayce, K. 1988. Introduced cordgrass, Spartina alternifloraLoisel., in salt marshes and tidelands <strong>of</strong> Willapa Bay, Washington.Illwaco, WA: U.S. Fish and Wildlife Service.Sayce, K., B. Dumbauld, and J. Hidy. 1997. Seed dispersal in drift<strong>of</strong> Spartina alterniflora. Presented at <strong>the</strong> Second <strong>International</strong>Spartina <strong>Conference</strong>, Olympia, WA.Stiller, J.W. and A.L. Denton. 1995. One hundred years <strong>of</strong> Spartinaalterniflora (Poaceae) in Willapa Bay, Washington: random amplifiedpolymorphic DNA analysis <strong>of</strong> an invasive population.Molecular Ecology 4:355-363.Zaremba, K. and M.F. McGowan. 2004. San Francisco EstuaryInvasive Spartina Project Monitoring Report for 2003. Oakland:California State Coastal Conservancy.- 261 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementCOVERING THE SPARTINA THREAT: AN ALTERNATIVE CONTROL METHOD FOR NON-NATIVE SPARTINA PATENS IN A WEST COAST SALT MARSHD.L. PICKERING2499 North Bank Rd., Otis, Oregon 97368; dpickering@tnc.orgOn Oregon’s central coast, <strong>the</strong> Nature Conservancy’s Cox Island Preserve harbors <strong>the</strong> introducedsalt marsh grass Spartina patens (saltmeadow cordgrass), <strong>the</strong> only known infestation in <strong>the</strong> state.The 187-acre preserve lies in <strong>the</strong> Siuslaw River, seven miles inland from <strong>the</strong> Pacific Coast. Spartinapatens is native to <strong>the</strong> east coast <strong>of</strong> <strong>the</strong> United States, found from Newfoundland to Texas. It wasprobably introduced to <strong>the</strong> west coast in <strong>the</strong> early 1900s. Spartina. patens invades mid-marshcommunities at elevations ranging from 1.83 to 2.05 meters (m) above mean low water (Frenkel andBoss 1988). Initially, it spreads primarily by rhizomes and forms circular, monotypic stands whichcrowd out native plants and eliminate wildlife habitat. These Spartina patches accumulate sedimentand litter at a faster rate than <strong>the</strong> surrounding marsh vegetation, thus altering <strong>the</strong> natural succession<strong>of</strong> <strong>the</strong> site. To restore Cox Island and prevent S. patens from spreading to o<strong>the</strong>r estuaries, we fieldtestedmethods that have shown promise at controlling S. patens elsewhere. Covering with heavydutylandscaping fabric anchored by spikes was <strong>the</strong> most effective control method <strong>of</strong> those we tried.Leaving <strong>the</strong> fabric on for two years kills <strong>the</strong> Spartina. After removal, native salt-marsh vegetationre-colonizes on its own. To date, 0.81 hectares (ha) (2 acres [ac]) have been restored and ano<strong>the</strong>r1.62 ha (4 ac) are now covered. This methodology may not be feasible for large-scale control effortson well-established infestations, but it is a very viable option for control efforts at <strong>the</strong> most effectivetime to eliminate invasive non-native species, during <strong>the</strong> early stages <strong>of</strong> an infestation.Keywords: Spartina patens, non-chemical control, salt marsh management, geotextile coveringTHE SPARTINA THREATThe only known Oregon occurrence <strong>of</strong> <strong>the</strong> introducedsalt marsh grass Spartina patens (saltmeadow cordgrass)occurs on <strong>the</strong> Nature Conservancy’s 187-acre Cox IslandPreserve located in <strong>the</strong> Siuslaw River. Spartina patens isnative to <strong>the</strong> east coast <strong>of</strong> <strong>the</strong> United States, found fromNewfoundland to Texas. It was probably introduced to <strong>the</strong>west coast around <strong>the</strong> turn <strong>of</strong> <strong>the</strong> 20 th century.Spartina. patens invades mid-marsh communities atelevations ranging from 1.83 to 2.05 meters (m) above meanlow water (Frenkel and Boss 1988). Initially, it spreadsprimarily by rhizomes and forms circular, monotypic standswhich crowd out native plants and eliminate wildlife habitat.These Spartina patches accumulate sediment and litter at afaster rate than <strong>the</strong> surrounding marsh vegetation, thusaltering <strong>the</strong> natural succession <strong>of</strong> <strong>the</strong> site.At Cox Island, S. patens apparently established inundisturbed vegetation before 1939. Since <strong>the</strong>n, it spread tocover approximately 1.1 hectares (ha) (2.7 acres [ac]) by1996. If uncontrolled, it could continue to spread until allavailable habitat is occupied (Frenkel and Boss 1988).EARLY DETECTION PARTNERSHIPO<strong>the</strong>r invasive species <strong>of</strong> Spartina from <strong>the</strong> east coasthave become established in Washington and California. Forexample, S. alterniflora has infested Willapa Bay inWashington where it is eliminating an important feedingarea for migratory waterfowl by invading mudflats andforming a Spartina monoculture (Aberle 1993).Studies indicate that 13 Oregon estuaries are at risk <strong>of</strong>invasion by non-native Spartina species (Daehler and Strong1996). Seeds or o<strong>the</strong>r propagules could be dispersed to newareas by migratory waterfowl, dredging operations, shellfishharvesting or movement <strong>of</strong> materials between oysterproducingareas. Early detection is essential to prevent newinfestations in Oregon (Pfauth et al. 2003).The Nature Conservancy has partnered with <strong>the</strong> OregonDepartment <strong>of</strong> Agriculture (ODA) and <strong>the</strong> SiuslawWatershed Council to control Spartina on <strong>the</strong> Cox IslandPreserve and to detect any new infestations in <strong>the</strong> state. In2003, ODA developed a Spartina Response Plan to protectOregon estuaries from Spartina invasions (Pfauth et al.2003). Nature Conservancy and Watershed Councilmembers have volunteered to survey salt marsh areas in twoestuaries for invasive Spartina species. Two patches <strong>of</strong> S.patens were found in <strong>the</strong> Siuslaw Estuary on propertyadjacent to Cox Island and were controlled. No o<strong>the</strong>rSpartina was found.RESTORING COX ISLANDTo restore Cox Island and prevent S. patens fromspreading to o<strong>the</strong>r estuaries in Oregon, <strong>the</strong> NatureConservancy field-tested several control methods from 1996- 263 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinato 1998 that have shown promise at o<strong>the</strong>r sites, includingcovering with heavy-duty landscape fabric, repeatedmowing, artificial inundation and removal by manualdigging.Results <strong>of</strong> <strong>the</strong>se tests indicated that covering <strong>the</strong>Spartina with landscaping fabric (Mirafi 500 or Amoco2002) anchored by gutter spikes pushed into <strong>the</strong> sedimentwas <strong>the</strong> most effective and least environmentally detrimentalmethod, <strong>of</strong> those that were tried, for controlling this invasivespecies at Cox Island. Leaving <strong>the</strong> fabric on for two yearseffectively kills <strong>the</strong> Spartina, leaving behind bare ground.After removal <strong>of</strong> <strong>the</strong> fabric, native salt-marsh vegetation recolonizes<strong>the</strong>se bare patches on its own so active planting <strong>of</strong>native vegetation is not necessary.The landscaping fabric we use (Amoco 2002) is a tough,woven black plastic fabric (geotextile) that resists UV lightand holds up well in a saltwater environment. We havesuccessfully used pieces <strong>of</strong> this fabric for six consecutiveyears (on three different patches for two years each) before itbecame too thin to block light. Since <strong>the</strong> fabric is woven, cutedges need to be folded under so <strong>the</strong> wind does not unravel<strong>the</strong> edges. This fabric is available from ACF West Inc. inPortland, Oregon (1-800-878-5115 or 503-771-5115). Itscost ranged from $315/roll (picked up) to $385/roll (locallydelivered). These rolls contain 6,300 square feet (ft 2) <strong>of</strong>fabric and measure 18 ft x 350 ft.To install <strong>the</strong> fabric, we identify <strong>the</strong> spatial limits <strong>of</strong> <strong>the</strong>patch, mow <strong>the</strong> vegetation around <strong>the</strong> boundary <strong>of</strong> <strong>the</strong> patchso <strong>the</strong> fabric lays flat, cut a piece <strong>of</strong> fabric <strong>of</strong> sufficient sizeto extend well beyond <strong>the</strong> border <strong>of</strong> <strong>the</strong> patch, and pin inplace with gutter spikes every two to three feet along <strong>the</strong>edges while pulling <strong>the</strong> fabric taut. The fabric should extenda minimum <strong>of</strong> two feet beyond <strong>the</strong> edge <strong>of</strong> patches to helpprevent any rhizomes from growing out beyond <strong>the</strong>covering. We fold <strong>the</strong> edge <strong>of</strong> <strong>the</strong> fabric under and push aspike with a washer (to keep <strong>the</strong> head <strong>of</strong> <strong>the</strong> spike fromgoing through <strong>the</strong> fabric) through <strong>the</strong> fabric and pound itinto <strong>the</strong> substrate. If <strong>the</strong> ground is too s<strong>of</strong>t in places to get agood grip with <strong>the</strong> gutter spikes, we use nine-inch nailsinstead. (These also tend to last longer in <strong>the</strong> saltwaterenvironment and do not need washers.). We angle <strong>the</strong> spikestoward <strong>the</strong> center <strong>of</strong> <strong>the</strong> covering as we drive <strong>the</strong>m in to helpprevent <strong>the</strong> tidal currents and wind from pulling up on <strong>the</strong>edges. If additional anchoring is needed, for example wheretwo pieces <strong>of</strong> fabric overlap, two spikes can be driven in atopposite angles in <strong>the</strong> same spot.In <strong>the</strong> fall <strong>of</strong> 1998, we began efforts to control <strong>the</strong> smalloutlier patches <strong>of</strong> Spartina before <strong>the</strong>y could turn into largepatches. This strategy <strong>of</strong> beginning with small outliers hasproven to be <strong>the</strong> most effective way to contain a non-nativespecies invasion (Moody and Mack 1988). As <strong>of</strong> 2004, wehave controlled all <strong>of</strong> <strong>the</strong> outliers and are making progresson <strong>the</strong> main infestation area. The Spartina in this area hasgrown from three patches in 1939 to a Spartina meadowtoday. We are covering <strong>the</strong> perimeters <strong>of</strong> <strong>the</strong>se meadows tohalt <strong>the</strong>ir spread and will work towards <strong>the</strong>ir interiors insuccessive years. Each year we mow flowering patches thathave not yet been covered to prevent seed set. The patchesare mowed from mid- to late-August with gas-poweredstring trimmers and are cut well above <strong>the</strong> ground surface toavoid tearing out rhizomes that might spread <strong>the</strong> infestation.Overall, we have successfully restored about 0.81 ha (2ac) <strong>of</strong> former Spartina patches to native salt marsh. At leastano<strong>the</strong>r 1.62 ha (4 ac) are currently covered with fabric. Weestimate that 0.43 ha (1.1 ac) <strong>of</strong> large Spartina patchesremain uncontrolled on <strong>the</strong> island, which we hope to coverby <strong>the</strong> end <strong>of</strong> 2005. Controlling this invasive species on CoxIsland will not only restore native salt marsh to <strong>the</strong> island,but will also help to prevent Spartina's spread to o<strong>the</strong>r parts<strong>of</strong> <strong>the</strong> Siuslaw Estuary and to o<strong>the</strong>r estuaries in Oregon andWashington.I believe this weed control method shows promise foruse on o<strong>the</strong>r invasive species as well (we are currently tryingit on upland grasses and reed canarygrass [Phalarisarundinacea]). While it may not be practical for wellestablished invasive species populations that cover morethan 4-8 ha (~10-20 ac), it is a viable non-chemical controlmethod for outlier patches and small pioneering infestationson a site.ACKNOWLEDGEMENTSWe are very grateful to <strong>the</strong> following for providingfunding for this project: NOAA/TNC Community-basedRestoration Program; Oregon Watershed EnhancementBoard; Oregon Department <strong>of</strong> Agriculture's Oregon StateWeed Board, Noxious Weed Control Grant Program; andU.S. Fish and Wildlife Service, North American WetlandsConservation Act, Small Grants Program.REFERENCESAberle, B.L. 1993. The biology and control <strong>of</strong> introduced Spartina(cordgrass) worldwide and recommendations for its control inWashington. Master's Thesis, Olympia, WA: The EvergreenState College.Daehler, C.C. and D.R. Strong. 1996. Status, prediction, and prevention<strong>of</strong> introduced cordgrass Spartina spp. invasions in Pacificestuaries, USA. Biological Conservation 78:51-58.Frenkel, R.E. and T.R. Boss. 1988. Introduction, establishment andspread <strong>of</strong> Spartina patens on Cox Island, Siuslaw Estuary, Oregon.Wetlands 8:33-49.Moody, M.E. and R.N. Mack. 1988. Controlling <strong>the</strong> spread <strong>of</strong> plantinvasions: <strong>the</strong> importance <strong>of</strong> nascent foci. Journal <strong>of</strong> AppliedEcology 25:1009-1021.Pfauth, M., M. Sytsma, and D. Isaacson. 2003. Oregon SpartinaResponse Plan. Portland State University, Center for Lakes andReservoirs report prepared for Oregon Department <strong>of</strong> Agriculture.- 264 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementCOMMUNITY SPARTINA EDUCATION AND STEWARDSHIP PROJECTK. O’CONNELLRestoration Ecologist, People For Puget Sound, 911 Western Ave., Ste. 580, Seattle, WA 98104; koconnell@pugetsound.orgPeople For Puget Sound’s Community Spartina Education and Stewardship Project is a communitybasedSpartina control program targeted at small groups <strong>of</strong> private property owners and select publicbeaches. The goals <strong>of</strong> <strong>the</strong> program are to: 1) educate and mobilize shoreline property ownersthrough stewardship to eradicate Spartina from <strong>the</strong>ir beaches, and 2) educate and involve all citizensin <strong>the</strong> Puget Sound-wide Spartina control effort through participation in large, organized communitySpartina-removal dig events. This project includes a strong component <strong>of</strong> monitoring and long-termstewardship driven by property owners and partnering organizations/agencies.Keywords: Community, stewardship, educationINTRODUCTION:In Washington State, Spartina is an invasive, salttolerantweed that threatens our natural shorelineecosystems. It aggressively displaces native habitat, and leftuntreated could cause irreversible damage to <strong>the</strong> nearshoreenvironment <strong>of</strong> Puget Sound. State agencies andorganizations involved in Spartina control agree thateducation <strong>of</strong> <strong>the</strong> public is key to <strong>the</strong> weed’s eradication fromPuget Sound. The goal <strong>of</strong> People For Puget Sound’sCommunity Spartina Education and Stewardship Program,launched in 2004, is to enhance community education,involvement and stewardship in <strong>the</strong> Puget Sound bytargeting areas <strong>of</strong> small Spartina infestations, generallyunder one acre, in sites with multiple private shorelineproperty owners.Research indicates that if eradication <strong>of</strong> Spartina is tooccur in an effective manner, individual infestations must betreated simultaneously and in a coordinated effort. Inaddition, <strong>the</strong> monitoring <strong>of</strong> sites post-eradication isnecessary to ensure that re-invasion <strong>of</strong> Spartina does notoccur. This education and stewardship project will act as aninvestment in <strong>the</strong> future <strong>of</strong> Spartina-free beaches in PugetSound, by focusing on community involvement andhighlighting <strong>the</strong> importance <strong>of</strong> stewardship. The projectenhances our own current programs, o<strong>the</strong>r local stewardshipprograms, and builds <strong>the</strong> key link between <strong>the</strong> agenciescommitted to Spartina removal and <strong>the</strong> citizens who dependon <strong>the</strong> health <strong>of</strong> <strong>the</strong> sound.The project works in close partnership with variousregional, state, local and tribal agencies in Puget Sound tomaximize <strong>the</strong> effectiveness <strong>of</strong> our efforts. Washington StateDepartment <strong>of</strong> Agriculture (WSDA), Skagit and IslandCounty Noxious Weed control boards, <strong>the</strong> Northwest StraitsCommission, Washington State University “BeachWatchers” program, and tribal communities assist inbuilding connections with <strong>the</strong> wider community and instaging large public dig events to raise awareness <strong>of</strong> <strong>the</strong>Spartina problem.METHODSThree to five priority shoreline communities aretargeted for outreach each year <strong>of</strong> this project. The projectmanager works with <strong>the</strong> state agencies, especially WSDA, toprioritize sites for outreach, and works with o<strong>the</strong>r localpartners to identify an interested neighbor in <strong>the</strong> target areato act as a community steward. The manager <strong>the</strong>n assists <strong>the</strong>steward in hosting a ‘Spartina Social’ with neighbors toeducate <strong>the</strong>m on Spartina’s environmental impacts, controlmethods, eradication strategies, monitoring and stewardship,as well as roles community members can play in <strong>the</strong> project.Generally two to six additional neighbors are also interestedin becoming stewards for <strong>the</strong> community following thisinitial meeting.Community stewards are trained to conduct baselinesurveys and monitoring at <strong>the</strong> site twice a year, and areequipped and supported in hopes that <strong>the</strong>y will continuemonitoring for up to 15+ years. We opted to use <strong>the</strong> protocolfor photo-point monitoring and photo-plot monitoringoutlined by Portland State University’s EnvironmentalSciences and Resources Program Student WatershedResearch Project (www.swrp.org). Photo-point monitoringinvolves taking landscape photos <strong>of</strong> <strong>the</strong> area <strong>of</strong> concern tomonitor overall changes at <strong>the</strong> site each year, including <strong>the</strong>presence <strong>of</strong> Spartina. In addition, one or two photo plotswere established in each site to capture an area <strong>of</strong> one squaremeter (1 m 2 ) <strong>of</strong> Spartina to monitor <strong>the</strong> percent coverage peryear in those plots. Baseline monitoring was conducted ineach community between August and October 2004.Community stewards will continue monitoring each year inlate May and again post-control (September or October).Community stewards were provided with all necessarymonitoring materials. Each community received a photomonitoringinstruction manual, including monitoringguidelines, data entry sheets, photo marker tags, monitoringequipment, a copy <strong>of</strong> People For Puget Sound’s SoundStewardship native and invasive plant ID guide, and contactinformation for <strong>the</strong> Island County Noxious Weed ControlBoard. In addition to <strong>the</strong> manual, stewards received a 1 m 2- 265 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaPVC photo plot, rebar stakes for marking photo points, and acompass for calculating directional bearings for photos. Thecommunities provided <strong>the</strong>ir own digital cameras for <strong>the</strong>monitoring.RESULTS AND DISCUSSIONPeople For Puget Sound’s project manager establishedan initial steward in each community in July and August2004 through our 2003 outreach efforts and with <strong>the</strong>assistance <strong>of</strong> <strong>the</strong> Washington State University ExtensionBeach Watchers Program. To date, <strong>the</strong> project has conductedoutreach to <strong>the</strong> communities <strong>of</strong> Juniper Beach and EagleTree Estates on Camano Island, and Harrington Lagoon onWhidbey Island. A neighbor from a fourth community,Cavelero Country Club on Camano Island, attended ano<strong>the</strong>rcommunity’s meeting and has been working with <strong>the</strong> projectmanager to get her community involved in <strong>the</strong> project.The project manager met with <strong>the</strong> community stewardsbetween August and October to conduct Spartina surveys <strong>of</strong><strong>the</strong> neighborhood’s shoreline, and training on monitoringtechniques. Baseline visual and photographic surveys wereconducted at this time, along with photo monitoring. Theproject manager also began training <strong>the</strong> stewards in nativeplant identification, focusing on plants commonly confusedwith Spartina.People For Puget Sound and its partners also hostedthree large, public Spartina dig events from June throughAugust 2004. These events focused on manual removal inareas where chemical control is not desired. Volunteers weretrained in manual removal techniques including digging andseedling removal with an emphasis on removing allrhizomatous roots to prevent regrowth.The project manager worked with partners to educate<strong>the</strong> general public about Spartina through <strong>the</strong>se communitydig events. The 6th Annual Skagit Dig Days was held onJune 19th at <strong>the</strong> Swinomish Casino lagoon. This event drew38 volunteers for four hours <strong>of</strong> manual removal. Twoadditional dig events were held in Island County: OakHarbor on Whidbey Island had 14 volunteers for three hourson August 14, and Iverson Spit on Camano Island drew 22volunteers for four hours on August 28.Monitoring surveys <strong>of</strong> approximately 20 hectares (48acres) <strong>of</strong> area affected by Spartina were conducted betweenMay 1 and October 31 This number was estimated from datain WSDA’s 2003 Spartina Eradication Program Report to<strong>the</strong> Legislature, PUB 805-110 (available online athttp://agr.wa.gov/PlantsInsects/Weeds/Spartina/default.htm).It is projected that this same area will be monitored oneadditional time during <strong>the</strong> course <strong>of</strong> this grant period, endingMay 2005. The three dig events resulted in an estimated 0.6hectares (1.5 acres) <strong>of</strong> Spartina removed from <strong>the</strong> nearshore.Outreach to <strong>the</strong> private communities was not intended toresult in direct control <strong>of</strong> Spartina, since <strong>the</strong> Island CountyNoxious Weed Control Progrram is currently managingcontrol efforts throughout <strong>the</strong> county. We expect that WSDAwill report <strong>the</strong> results <strong>of</strong> acres treated at <strong>the</strong>se sites in its2004 report to <strong>the</strong> Legislature.Thus far this project has trained over 70 citizens inproper identification and manual removal methods forinvasive Spartina through <strong>the</strong> large community dig events.We have educated over 60 private shoreline property ownerson Spartina’s environmental impacts, control methods,identification, and options for community-based eradicationthrough community meetings and presentations. The projectmanager has trained 11 stewards from <strong>the</strong> three communitiesin proper monitoring techniques. The stewards have agreedto conduct photo monitoring <strong>of</strong> <strong>the</strong>ir community beachestwice a year, conduct visual monitoring throughout <strong>the</strong> year,report all monitoring activities to <strong>the</strong> project manager,provide copies <strong>of</strong> all monitoring photos, work with <strong>the</strong>project manager and Island County Noxious Weed ControlBoard to organize control efforts, and commit to long-termstewardship <strong>of</strong> <strong>the</strong>ir community beaches for hopefully 20years or more.We have received very positive feedback from <strong>the</strong>public for this project and we feel that this first year hasbeen a great success. The project will continue to conductoutreach to three to five additional communities each year,while maintaining <strong>the</strong> necessary level <strong>of</strong> involvement incurrent communities’ activities. The project manager islooking into expanding <strong>the</strong> project in upcoming years inseveral areas. There is a distinct need to conduct yearly GPSsurveys <strong>of</strong> each community’s infestation and to acquire GIScapabilities that would allow for <strong>the</strong> production <strong>of</strong> maps ando<strong>the</strong>r visuals.The project manager is also investigating options for anonline database that can be accessed by our staff, communitystewards and neighbors, and our partnering agencies andorganizations. The project manager will also be workingwith WSDA, NOAA, and Restore America’s Estuaries todevelop a Spartina stewardship manual and a regionalSpartina information and identification flyer.ACKNOWLEDGEMENTSNational Oceanic and Atmospheric Administration(NOAA), Restore America’s Estuaries (RAE), SkagitCounty Marine Resources Committee, Island County MarineResources Committee and Island County Noxious WeedControl Board have provided funding support for thisproject. I would also like to thank Kyle Murphy at WSDAfor his support and assistance in this project. Susan Horton<strong>of</strong> <strong>the</strong> Island County Noxious Weeds Control Program andSusan Moreno <strong>of</strong> <strong>the</strong> Swinomish Tribe Planning Departmentprovided hours <strong>of</strong> <strong>the</strong>ir time in <strong>the</strong> planning and execution <strong>of</strong><strong>the</strong> three dig events. Dot Irvin and John Custer <strong>of</strong> <strong>the</strong> WSUExtension Beach Watchers Program assisted in contactingpotential stewards in new communities. Lastly, a verygenerous thank you to Adrianna Ericson who provided atleast 15 hours a week during <strong>the</strong> summer months to assist infield work and steward training.- 266 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementBIOLOGICAL CONTROL OF SPARTINAF. S. GREVSTAD 1 ,M.S.WECKER 1 , AND D. R. STRONG 21 Olympic Natural Resources Center, University <strong>of</strong> Washington, P.O. Box 1628, Forks, WA 98331;grevstad@u.washington.edu2 Department <strong>of</strong> Evolotion and Ecology, University <strong>of</strong> California, Davis, CA 95616Biological control using introduced natural enemies can be an effective approach to <strong>the</strong> long termcontrol <strong>of</strong> widespread weeds. A biological control program against Spartina spp. is underway inWashington State, where more than 10,000 hectares (ha) <strong>of</strong> intertidal mudflat are affected bySpartina alterniflora and Spartina anglica. Releases <strong>of</strong> <strong>the</strong> planthopper Prokelisia marginata havebeen made into Willapa Bay each year since 2000 and into Puget Sound since 2003. Prior tointroducing this insect, rigorous host specificity testing and a review by <strong>the</strong> Technical AdvisoryGroup on Biological Control <strong>of</strong> Weeds confirmed that <strong>the</strong> risk to non-target plants was minute.Populations <strong>of</strong> <strong>the</strong> biocontrol agent were initially slow to establish and grow. However, earlyproblems with high winter mortality have been remedied through a combination <strong>of</strong> improved releasesite selection and <strong>the</strong> use <strong>of</strong> cold-hardy east coast biotypes. At least two populations in Willapa Bayare well established and expanding. At a localized scale, we have measured 50 percent reductions <strong>of</strong>Spartina biomass and 90 percent reduction in viable seed set due to P. marginata. The full extent <strong>of</strong><strong>the</strong> impact will only be known with time.While <strong>the</strong> use <strong>of</strong> biological control in California may pose a risk to <strong>the</strong> closely related nativeSpartina foliosa, it would be an excellent option in o<strong>the</strong>r o<strong>the</strong>r parts <strong>of</strong> <strong>the</strong> world where Spartina hasinvaded and where <strong>the</strong>re are no closely related native Spartina species. In addition to P. marginata,o<strong>the</strong>r candidate biocontrol agents from <strong>the</strong> Atlantic Coast are currently being investigated.Keywords: Biological control, Spartina alterniflora, Spartina anglica, Prokelisia marginata,Willapa Bay, Puget SoundINTRODUCTIONClassical biological control <strong>of</strong> a pest or weed involves<strong>the</strong> introduction <strong>of</strong> a natural enemy (biocontrol agent) fromano<strong>the</strong>r geographic region. The goal is to establish apermanent population <strong>of</strong> <strong>the</strong> biocontrol agent that willprovide long-term control. During <strong>the</strong> last century, close to1,000 biological control introductions for weeds were madethroughout <strong>the</strong> world (Julien and Griffiths 1998). Modernweed biocontrol projects in <strong>the</strong> United States proceed onlyafter extensive testing <strong>of</strong> <strong>the</strong> natural enemy followed by areview by <strong>the</strong> federal Technical Advisory Group onBiological Control <strong>of</strong> Weeds to ensure that it will not harmo<strong>the</strong>r organisms. Weed biocontrol has proven to be a safeand <strong>of</strong>ten very effective method <strong>of</strong> long term control <strong>of</strong>widespread invasive plants (Cruttwell-McFadyen 1998).Biological control has both advantages anddisadvantages over traditional control. Unlike most chemicaland mechanical approaches, biological control is highlyspecific to <strong>the</strong> target weed. Since biological control agentshave been chosen for <strong>the</strong>ir host specificity, <strong>the</strong>y will notharm o<strong>the</strong>r plant species intermixed with <strong>the</strong> weed.Biological control is economical over large areas. Onceestablished, <strong>the</strong> biological control agent will reproduce,spread to new sites, and continue to damage <strong>the</strong> plant withlittle or no additional input. Biological controls have notoxic residues or health hazards as do some herbicides.When it is successful, biological control can provide apermanent solution to a weed problem, although it usuallymaintains a very low level <strong>of</strong> infestation ra<strong>the</strong>r than bringingabout full eradication.The disadvantages <strong>of</strong> biological control include <strong>the</strong>large amount <strong>of</strong> pre-release research that is required.Moreover, even with careful selection <strong>of</strong> host-specificagents, <strong>the</strong>re will always be some small risk to non-targetorganisms from ei<strong>the</strong>r direct or indirect interactions.Biological control is <strong>of</strong>ten slow in its action, taking severalyears and even up to a decade for an impact to be seen.Finally, <strong>the</strong> complexities <strong>of</strong> ecological interactions mean that<strong>the</strong> effectiveness <strong>of</strong> biological control is difficult to predictahead <strong>of</strong> time. The probability <strong>of</strong> successful controlincreases when multiple biocontrol agents are used (Deno<strong>the</strong>t al. 2003).A biological control program for Spartina alternifloraand Spartina anglica in Washington State was developedduring <strong>the</strong> late 1990s. The biocontrol agent Prokelisiamarginata, a sapsucking planthopper, was introducedbeginning in 2000. In this paper, we provide backgroundinformation about this biological control program. Then wepresent results <strong>of</strong> a comparison <strong>of</strong> <strong>the</strong> performance <strong>of</strong> four- 267 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinapopulations <strong>of</strong> P. marginata imported from differentgeographic locations. Finally, we present possible futuredirections for <strong>the</strong> biocontrol program including <strong>the</strong> screening<strong>of</strong> additional agents from S. alterniflora’s native range.BACKGROUND ON SPARTINA BIOCONTROL INWASHINGTON STATETo date only one biological control agent has beenintroduced into Washington for control <strong>of</strong> Spartina spp. Thedelphacid planthopper, Prokelisia marginata, wasintroduced from California into Willapa Bay beginning in2000 and into north Puget Sound in 2003. The introductionswere made only after extensive testing demonstrated its highlevel <strong>of</strong> host specificity (Grevstad et al. 2003) and afterruling out <strong>the</strong> possibility that it could vector a disease (Daviset al. 2002). The project was reviewed and approved by <strong>the</strong>Technical Advisory Group on Biological Control <strong>of</strong> Weedsand permitted by <strong>the</strong> Washington State Department <strong>of</strong>Agriculture and <strong>the</strong> U.S. Department <strong>of</strong> Agriculture’sAnimal and Plant Health Inspection Service (USDA-APHIS).P. marginata is native to <strong>the</strong> Atlantic and Gulf Coasts <strong>of</strong>North America and also occurs in San Francisco Bay,California. Genetic analyses (R. Denno and D. Hawthorne,University <strong>of</strong> Maryland, pers. comm.) indicate that P.marginata was probably introduced to California from <strong>the</strong>East Coast in recent decades. The absence <strong>of</strong> Prokelisia spp.in an early 1970’s survey <strong>of</strong> insects on Spartina foliosa inSan Francisco Bay also supports a recent introduction to <strong>the</strong>West Coast (Cameron 1972). P. marginata was selected as apromising biocontrol agent because <strong>of</strong> its narrow host rangeand its known potency against S. alterniflora and S. anglica(Daehler and Strong 1997; Wu et al. 1999). P. marginataadults and nymphs feed by sucking <strong>the</strong> sap from <strong>the</strong> plant,draining its energy supply. Spartina is also damaged by <strong>the</strong>scars that arise on <strong>the</strong> leaf surface where adult females inserteggs. If high enough densities <strong>of</strong> P. marginata are attained,feeding and oviposition scars cause <strong>the</strong> leaves to turn brownand eventually kill <strong>the</strong> plant.The Spartina biocontrol program is unique in being <strong>the</strong>first classical biocontrol program to target a grass, althougho<strong>the</strong>rs are being considered (Tewksbury et al. 2002; Wittand McConnachie 2004). It is also <strong>the</strong> first application <strong>of</strong>classical weed biocontrol in a marine intertidal environment.This project differs from most classical biocontrol projectsin that <strong>the</strong> targeted weed is invasive in <strong>the</strong> same country towhich it is native (although a different region). Thebiocontrol agents are likewise transferred between statesra<strong>the</strong>r than between countries.To <strong>the</strong> advantage <strong>of</strong> a biological control program,invasive Spartina in Washington appears to have lostresistance to herbivory since its introduction. In greenhouseexperiments (Daehler and Strong 1997; Wu et al. 1999;Garcia-Rossi et al. 2003), plants from <strong>the</strong> invasivepopulations in Washington suffer much greater biomassreduction and mortality from P. marginata than plants fromnative locations. Herbivore exclusion and additionexperiments in <strong>the</strong> field also demonstrate this difference inresponse (compare Daehler and Strong 1996 with Grevstadet al. 2003). The vulnerability <strong>of</strong> <strong>the</strong> Washingtonpopulations may be due to an evolved loss <strong>of</strong> resistance in<strong>the</strong> absence <strong>of</strong> herbivores (Garcia-Rossi et al. 2003). Themechanism <strong>of</strong> Spartina vulnerability is unknown, but it maybe related to <strong>the</strong> structural breakdown <strong>of</strong> vascular cells as aresult <strong>of</strong> piercing by <strong>the</strong> planthopper during feeding andoviposition (Wu et al. 1999). The possibility that <strong>the</strong>vulnerability is due to a disease vectored by <strong>the</strong> planthopperwas ruled out by Davis et al. (2002).Over <strong>the</strong> past few years, P. marginata has been releasedat 40 locations in Willapa Bay and Puget Sound. Resultshave been encouraging, but not without setbacks. Followingrelease, P. marginata populations typically grow explosivelyduring <strong>the</strong>ir first summer and cause visible damage to <strong>the</strong>plants by fall (Grevstad et al. 2003). Local densities in somesites have exceeded 50,000 insects per m 2 . A 50% reductionin local biomass was measured in an early field cageexperiment (Grevstad et al. 2003). A 90% reduction in fieldseed viability was found in localized areas where P.marginata density was greater than 30 per stem (Grevstad,unpublished data). However, low survival <strong>of</strong> nymphs over<strong>the</strong> winter has prevented <strong>the</strong>se populations from building to<strong>the</strong> high densities over large areas required to have largescaleimpacts on <strong>the</strong> Spartina population. In spite <strong>of</strong> <strong>the</strong>summer boom, P. marginata populations are typically muchsmaller <strong>the</strong> following spring than <strong>the</strong>y were at <strong>the</strong> time <strong>of</strong>release. Some <strong>of</strong> <strong>the</strong>se populations eventually build updensities, but many have gone extinct and o<strong>the</strong>rs aredwindling or growing only slowly.Determining <strong>the</strong> best geographic source <strong>of</strong> P. marginataIn selecting a geographic source <strong>of</strong> a biocontrol agent, itis important to consider <strong>the</strong> ways that herbivore populationsmay be locally adapted. Classical biological controlprograms <strong>of</strong>ten seek an agent source population from alocation that has a climate similar to <strong>the</strong> region where it willbe introduced. However, in <strong>the</strong> case <strong>of</strong> <strong>the</strong> Spartinabiocontrol program, a close match to <strong>the</strong> Willapa Bayclimate does not exist. The San Francisco Bay area has asimilarly moderate climate but temperatures areapproximately 5ºC warmer at all times <strong>of</strong> year. East Coastlocations have more extreme seasonality and no location canmatch both winter and summer temperatures. A nor<strong>the</strong>asternlocation such as Rhode Island has <strong>the</strong> best match during <strong>the</strong>summer months, but a mid-Atlantic location, such asVirginia, has <strong>the</strong> best match during <strong>the</strong> winter.- 268 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementIn addition to climate and host plant adaptations,seasonal adaptations affecting <strong>the</strong> phenology <strong>of</strong> <strong>the</strong>biocontrol agent are also likely to vary among potentialagent source populations. Many insects use photoperiod as acue for synchronizing life history events with seasonalchange in environmental conditions. When an insect ismoved from one geographic location to ano<strong>the</strong>r, itsphenology may not be synchronized to <strong>the</strong> new seasonalschedule. In Willapa Bay, <strong>the</strong> California population <strong>of</strong> P.marginata has been observed to emerge in late February, atime that may be too early for nymph survival in <strong>the</strong> coolerand longer winters <strong>of</strong> coastal Washington. Based on a match<strong>of</strong> <strong>the</strong> timing <strong>of</strong> arrival <strong>of</strong> warm temperatures, a RhodeIsland source may be <strong>the</strong> best match. However, <strong>the</strong>possibility that late emergence could reduce <strong>the</strong> number <strong>of</strong>generations produced each year makes this outcomeuncertain.Ano<strong>the</strong>r consideration in choosing a source is potentialvariation in ability to compensate for plant defenses. Apopulation from a location far<strong>the</strong>r south, such as Georgia,while poorly adapted climatically, could be better adaptedfor overcoming plant defenses to herbivory, which areknown to be greater in sou<strong>the</strong>rn Spartina plants (seePennings et al. 2005 and Katz et al. 2005 in <strong>the</strong>seproceedings). A “new association” between a sou<strong>the</strong>rnherbivore and nor<strong>the</strong>rn plant could make biocontrol moreeffective.In <strong>the</strong> spring <strong>of</strong> 2004, after obtaining permits fromUSDA-APHIS and <strong>the</strong> Washington State Department <strong>of</strong>Agriculture, four populations P. marginata were introducedinto Willapa Bay. The source populations were (1) Sausalito,San Francisco Bay, California, (2) Grayville, Rhode Island,(3) Quinby, Virginia, and (4) Jekyll Island, Georgia. Allpopulations passed through a period <strong>of</strong> quarantine in alaboratory in Davis, California before being imported intoWashington for rearing in a greenhouse.At each <strong>of</strong> five sites in Willapa Bay, we set up four 4x4-meter (m) release plots located at similar tidal elevationswithin <strong>the</strong> same Spartina meadow, but separated by at least100 m. Each plot was randomly assigned one <strong>of</strong> <strong>the</strong> foursource populations. Five thousand insects <strong>of</strong> <strong>the</strong> assignedpopulation were released into each plot by placing heavilyinfested Spartina stems clipped from five rearing plantsuniformly throughout <strong>the</strong> plot. A 100 m distance betweenrelease plots ensures that populations will remain separatelong enough to compare <strong>the</strong>ir phenology and performance.P. marginata were sampled in late September using aninsect vacuum converted from a leaf blower. Eightuniformly spaced sample points 21-cm in diameter(corresponding to <strong>the</strong> vacuum tube’s diameter) within eachrelease area were vacuumed. Adults and nymphs werecounted separately and adults were scored for wing form;P. marginata per m 2Proportion in adult stagePercent macropterous18000160001400012000100000.300.250.200.150.100.050.0080006000400020001008060402000ABGA VA CA RISource <strong>of</strong> populationGA VA CA RICSource <strong>of</strong> populationGA VA CA RISource <strong>of</strong> populationFig. 1. (A) Densities <strong>of</strong> Prokelisia marginata measured at <strong>the</strong> end <strong>of</strong> Septemberfollowing introduction <strong>of</strong> 5,000 individuals from each <strong>of</strong> four geographicsources. (B) The proportion <strong>of</strong> each source population in <strong>the</strong> adult stage at <strong>the</strong>end <strong>of</strong> September. (C) The percentage <strong>of</strong> adults in each population that weremacropterous (long-winged). Each bar represents <strong>the</strong> mean and standard errorfor five replicate populations from each source.- 269 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinaei<strong>the</strong>r brachypterous (rudimentary or abnormally small) ormacropterous (long or large).We found striking differences among <strong>the</strong> fourpopulations in <strong>the</strong> densities attained by <strong>the</strong> end <strong>of</strong> summer(Fig. 1A; ANOVA F=9.227, P=0.002). The best performinggeographic source, in terms <strong>of</strong> <strong>the</strong> densities <strong>of</strong> P. marginataobtained by end <strong>of</strong> summer, was Rhode Island, which is also<strong>the</strong> location with <strong>the</strong> best summer temperature match toWillapa Bay. In post-hoc tests, Rhode Island populationssignificantly differed from Georgia and Virginia, but did notsignificantly differ from California. California did notsignificantly differ from Georgia and Virginia.The populations also differed in <strong>the</strong> proportion <strong>of</strong> <strong>the</strong>population that was in <strong>the</strong> adult stage at <strong>the</strong> end <strong>of</strong>September (Fig. 1B; ANOVA F=3.637, P=0.030). Californiapopulations had four to five times <strong>the</strong> proportion <strong>of</strong> adultsfound in <strong>the</strong> o<strong>the</strong>r populations. This may indicate differencesin phenology although fur<strong>the</strong>r data is needed.Finally, we found differences among <strong>the</strong> fourpopulations in <strong>the</strong> proportion <strong>of</strong> individuals that weremacropterous (Fig. 4C; ANOVA F=8.71, P

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and Managementlarvae <strong>of</strong> <strong>the</strong>se flies develop inside young S. alterniflorashoots, feeding on meristem tissue and developing leaves.The result is death <strong>of</strong> <strong>the</strong> shoot tip and no flower productionin nearly 100% <strong>of</strong> <strong>the</strong> stems that are infested. Both speciesoccur from Florida to Maine. The potency <strong>of</strong> <strong>the</strong>se flies liesin <strong>the</strong> fact that a single larva can kill a shoot. Bycomparison, it takes approximately 200 planthoppers to killa shoot (Daehler and Strong 1997). In our surveys, we foundseveral sites where <strong>the</strong> rate <strong>of</strong> shoot death due to this insectwas greater than 50%.O<strong>the</strong>r promising candidates include <strong>the</strong> sapsuckinginsects, Trigonotylus uhleri, Haliaspis spartinae, andProkelisia dolas. All three <strong>of</strong> <strong>the</strong>se species have <strong>the</strong>advantage that <strong>the</strong>y already occur in California, where <strong>the</strong>native Spartina foliosa occurs. Thus, introduction <strong>of</strong> <strong>the</strong>seinsects to Washington does not pose a risk to S. foliosa if<strong>the</strong>y were to disperse from Washington to California. T.uhleri is a mirid bug that feeds primarily on <strong>the</strong> tips <strong>of</strong> <strong>the</strong>leaves. H. spartinae is a scale insect that occurs in lowdensities on <strong>the</strong> Atlantic Coast but can be found at very highdensities in San Francisco Bay. P. dolas is a close relative <strong>of</strong>our current biocontrol agent. P. dolas has been shown to bemore tolerant <strong>of</strong> low host quality than P. marginata (Dennoet al. 2000), which suggests that it could be a potentbiocontrol agent. However, this species has been shown tooutcompete P. marginata (Denno et al. 2000). This could bedetrimental to <strong>the</strong> overall biocontrol program because unlikeP. marginata, P. dolas is unable to exploit <strong>the</strong> majority <strong>of</strong>Spartina occuring in mid to lower tidal elevations.UPDATE ON BIOCONTROL PROGRAMMuch has happened with regard to control <strong>of</strong> Spartinain Willapa Bay and Puget Sound since this report was originallyprepared in 2004. In particular, <strong>the</strong> state and federalagencies involved in <strong>the</strong> control program increased capacityand improved <strong>the</strong> effectiveness <strong>of</strong> <strong>the</strong> herbicide control treatmentsand eventually sprayed all Spartina-infested areas inWashington including biocontrol sites in an effort to eradicate<strong>the</strong> plant. As <strong>of</strong> 2007, prior to herbicide treatment <strong>of</strong><strong>the</strong> last remaining biocontrol site, <strong>the</strong> P. marginata populationwas increasing and spreading in <strong>the</strong> north end <strong>of</strong> WillapaBay with measured densities <strong>of</strong> more than 10,000 perm 2 , sufficient to cause browning <strong>of</strong> <strong>the</strong> plants over a twoacrearea. The full impact <strong>of</strong> biological control may never beknown. If P. marginata is capable <strong>of</strong> persisting on <strong>the</strong> sparseshoots that remain, it may help suppress any reinvasion <strong>of</strong>Spartina if traditional control methods are discontinued.CONCLUSIONSThe biological control agent, Prokelisia marginata, hasa demonstrated capacity for reducing Spartina biomass andseed set in Washington. In <strong>the</strong> first few years <strong>of</strong> <strong>the</strong>biocontrol program, poor overwintering survival <strong>of</strong> nymphskept populations from growing from year to year. With <strong>the</strong>introduction <strong>of</strong> P. marginata ecotypes from <strong>the</strong> East Coast,we are confident that we have provided <strong>the</strong> best opportunityfor P. marginata to be successful. P. marginata from RhodeIsland outperformed P. marginata from o<strong>the</strong>r locations in<strong>the</strong> first summer after release. Impacts to Spartina in <strong>the</strong> lab(Daehler and Strong 1997) and in field cages (Grevstad et al.2003) have been clearly demonstrated. Impacts on a muchlarger scale were beginning to be seen in 2007 prior toherbicide treatment <strong>of</strong> biocontol sites.The success <strong>of</strong> a biological control program for Spartinacould be fur<strong>the</strong>r improved with <strong>the</strong> screening andintroduction <strong>of</strong> additional biocontrol agents. Severalpromising candidate agents have been selected from adiverse community <strong>of</strong> insects and o<strong>the</strong>r organisms that use S.alterniflora as a host in <strong>the</strong>ir native range. Biological controlcould be a valuable tool in o<strong>the</strong>r parts <strong>of</strong> <strong>the</strong> world whereSpartina has invaded, especially where complete eradicationis too expensive or not feasible. Even where traditionalcontrol programs are underway, biological control cancontribute to Spartina reduction in an integrated weedmanagement approach, or it can serve as a backup in <strong>the</strong>case that complete eradication is not achieved.Unfortunately, San Francisco Bay is not a good targetlocation for biological control because <strong>of</strong> <strong>the</strong> risk to <strong>the</strong>native S. foliosa. However, it would be an excellent option inChina, New Zealand, and Australia where <strong>the</strong>re are no nativeSpartina species.ACKNOWLDEGMENTSWe would like to acknowledge <strong>the</strong> following agenciesand sponsors who have made <strong>the</strong> Spartina biocontrolprogram possible: National Sea Grant Program, U.S. Fishand Wildlife Service, U.S. Department <strong>of</strong> Fish and Wildlife,Coastal Resource Alliance, Washington State Department <strong>of</strong>Agriculture, Washington State Department <strong>of</strong> NaturalResources, and <strong>the</strong> Willapa National Wildlife Refuge. Wewould also like to thank <strong>the</strong> following individuals who havecontributed to <strong>the</strong> program in a wide variety <strong>of</strong> ways: BobDenno, Dino Garcia-Rossi, Robin Switzer, Carol O’Casey,Deanna McQuarrie, Hea<strong>the</strong>r Davis, Allison Fisher, KimPatten, Kathleen Sayce, Janie Civille, Todd Brownlee,Wendy Brown, Charlie Stenvall, Teri Butler, Dave Heimer,Kyle Murphy, Sue Moreno, Dick Casagrande, David Viola,and Lisa Tewksbury.REFERENCESCruttwell-McFayden, R.E. 1998. Biological control <strong>of</strong> weeds. AnnualReview <strong>of</strong> Entomology 43:369-393.Davis, L.V. and I.E. Gray. 1966. Zonal and seasonal distribution <strong>of</strong>insects in North Carolina salt marshes. Ecological Monographs36:275-295.Davis, H.G., D.R. Strong and D. Garcia-Rossi. 2002. The use <strong>of</strong>molecular assays to identify plant pathogenic organisms vectoredby biological control agents. BioControl 47: 487-497.- 271 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaDaehler, C.C. and D.R. Strong. 1995. Impact <strong>of</strong> high herbivoredensities on introduced smooth cordgrass, Spartina alterniflora,invading San Francisco Bay, California. Estuaries 18:409-417.Daehler, C.C. and D.R. Strong. 1997. Reduced herbivore resistancein introduced smooth cordgrass (S. alterniflora) after a century <strong>of</strong>herbivore-free growth. Oecologia 110:99-108.Denno, R.F. 1977. Comparison <strong>of</strong> <strong>the</strong> assemblages <strong>of</strong> sap-feedinginsects (Homoptera-Hemiptera) inhabiting two structurally differentsalt marsh grasses in <strong>the</strong> genus Spartina. EnvironmentalEntomology 6:359-372.Denno, R.F., L.W. Douglass, and D. Jacobs. 1985. Crowding andhost plant nutrition, environmental determinants <strong>of</strong> wing-form inProkelisia marginata. Ecology 66:1588-1596.Denno, R.F., M.A. Peterson, C. Gratton, J. Cheng, G.A. Langellotto,A.F. Huberty, and D.L. Finke. 2000. Feeding-inducedchanges in plant quality mediate interspecific competition betweensap-feeding herbivores. Ecology 81:1814-1827.Denoth, M., L. Frid, and J.H. Myers. 2002. Multiple agents in biologicalcontrol: improving <strong>the</strong> odds? Biological Control24:20-30.Garcia-Rossi, D., N. Rank, and D.R. Strong. Potential for selfdefeatingbiocontrol? variation in herbivore vulnerability amonginvasive Spartina genotypes. Ecological Applications 13:1640-1649.Grevstad, F.S., D.R. Strong, D. Garcia-Rossi, R.W. Switzer, andM.S. Wecker. 2003. Biological control <strong>of</strong> Spartina alterniflora inWillapa Bay, Washington using <strong>the</strong> planthopper Prokelisia marginata:agent specificity and early results. Biological Control27:32-42.Grevstad, F.S., M.S. Wecker, and R.W. Switzer. 2004 Habitattrade<strong>of</strong>fs in <strong>the</strong> summer and winter performance <strong>of</strong> <strong>the</strong> planthopperProkelisia marginata introduced against <strong>the</strong> intertidal grassSpartina alterniflora in Willapa Bay, WA. In: J. Cullen, ed. <strong>Proceedings</strong><strong>of</strong> <strong>the</strong> XI <strong>International</strong> Symposium on Biological Control<strong>of</strong> Weeds, April 27 – May 3, 2003. CSIRO, Canberra, Australia.Harris, P. 1981. Stress as a strategy in <strong>the</strong> biological control <strong>of</strong>weeds. In: Papavisa, G.C., ed. Biological Control in Crop Production,Beltsville Agricultural Research Center (BARC). Allanhead,Osman and Co., Totowa, New Jersey, pp. 333-340.McCoy, E.D. and J.R. Rey. 1981. Terrestrial arthropods <strong>of</strong> <strong>the</strong>northwest Florida salt marshes: Coleoptera. Florida Entomologist64(2):405-411.Montague, C.L., S.M. Bunker, E.B. Haines, M.L. Pace, and R.L.Wetzel. 1981. Aquatic Macroconsumers. In: Pomeroy, L.R. andR.G. Wiegert, eds. The ecology <strong>of</strong> a saltmarsh. Springer-Verlag,New York, pp. 69-85.Myers, J.H. 1985. How many insects are necessary for successfulbiocontrol <strong>of</strong> weeds? In: Delfosse, E.S., ed. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong>VI <strong>International</strong> Symposium on <strong>the</strong> Biological Control <strong>of</strong> Weeds.Vancouver, Canada, pp. 19-25.Newton, N.H. 1984. Stem-Boring Insect Larvae <strong>of</strong> Spartina alternifloraLoisel. (Poaceae): Distribution, Biology and Influenceon Seed Production. Ph.D. Dissertation, North Carolina StateUniversity, Raleigh.Nowierski, R.M., Z. Zeng, D. Schroeder, A. Gassmann, B.C. Fitzgerald,and M. Crist<strong>of</strong>aro. 2002. Habitat associations <strong>of</strong> Euphorbiaand Aphthona species from Europe: Development <strong>of</strong> predictivemodels for natural enemy release with ordination analysis.Biological Control 23:1-17.Pemberton, R.W. 2000. Predictable risk to native plants in weedbiological control. Oecologia 125: 489-494.Pringle, A.W. 1993. Spartina anglica colonization and physicaleffects in <strong>the</strong> Tamar Estuary, Tasmania 1971-91. Papers and<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> Royal Society <strong>of</strong> Tasmania 127:1-10Stiling, P.D. and D.R. Strong. 1983. Weak competition among stemborers, by means <strong>of</strong> murder. Ecology 64:770-778.Tauber, M.J., C. A. Tauber, and S. Masaki. 1986. Seasonal Adaptations<strong>of</strong> Insects. Oxford University Press.Tewksbury, L., R. Casagrande, B. Blossey, P. Häfliger, and M.Schwärzlander. 2002. Potential for biological control <strong>of</strong> Phragmitesaustralis in North America. Biological Control 23:191-212.Vince, S.W., I. Valiela, and J.M. Teal. 1981. An experimentalstudy <strong>of</strong> <strong>the</strong> structure <strong>of</strong> herbivorous insect communities in a saltmarsh. Ecology 62:1662-1678.Witt, A.B.R. and A.J. McConnachie. 2004. The potential for classicalbiological control <strong>of</strong> invasive grass species with special referenceto invasive Sporobolus spp. (Poaceae) in Australia. In:Cullen, J. ed. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> XI <strong>International</strong> Symposium onBiological Control <strong>of</strong> Weeds. April 27 – May 3, 2003. CSIRO,Canberra, Australia.Wu, M., S. Hacker, D. Ayres, and D.R. Strong. 1999. Potential <strong>of</strong>Prokelisia spp. as Biological Control Agents <strong>of</strong> EnglishCordgrass, Spartina anglica. Biological Control 16(3):267-273.- 272 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementECOLOGICAL INVESTIGATIONS OF NATURAL ENEMIES FOR AN INTERSTATE BIOLOGICALCONTROL PROGRAM AGAINST SPARTINA GRASSESD. VIOLA 1,2 ,L.TEWKSBURY 1 AND R. CASAGRANDE 11Department <strong>of</strong> Plant Sciences, University <strong>of</strong> Rhode Island, Kingston, RI 028812dviola@myrealbox.comSpartina alterniflora (smooth cordgrass) is a dominant member <strong>of</strong> salt marsh communities in itsnative range along <strong>the</strong> East Coast <strong>of</strong> <strong>the</strong> United States as well as an introduced invasive in severalWest Coast intertidal areas. In 2000, a biological control agent, Prokelisia marginata, was releasedin Willapa Bay, Washington, USA where S. alterniflora was rapidly spreading. Despiteestablishment <strong>of</strong> this agent, <strong>the</strong> invasive population was not brought under control. Although <strong>the</strong>natural enemies <strong>of</strong> S. alterniflora have been previously catalogued, little has been written regarding<strong>the</strong>ir ecology, with <strong>the</strong> exception <strong>of</strong> P. marginata. The purpose <strong>of</strong> this study was to provideadditional information on <strong>the</strong> ecology <strong>of</strong> insect herbivores <strong>of</strong> S. alterniflora in Rhode Island insupport <strong>of</strong> a larger effort to develop biological control agents for <strong>the</strong> West Coast.An intensive survey <strong>of</strong> insect species found on S. alterniflora in Rhode Island was conducted todescribe <strong>the</strong> composition <strong>of</strong> <strong>the</strong> insect herbivore community. Vacuum samples taken at both highand low marsh locations facilitated <strong>the</strong> characterization <strong>of</strong> species assemblages. Our results suggestthat <strong>the</strong>re is little variation in insect herbivore species richness among Rhode Island salt marshes.Damage due to Chaetopsis spp. is easily observed in <strong>the</strong> field, but <strong>the</strong> frequency <strong>of</strong> damaged stemsvaries greatly among sites. Additional research is needed to determine <strong>the</strong> factors responsible forthis variation.Keywords: Spartina alterniflora, Chaetopsis, biological control, natural enemies, invasive speciesINTRODUCTIONSpartina grasses have become serious invasive pests inWest Coast intertidal regions since <strong>the</strong>ir introduction in <strong>the</strong>late 1800s. Many Pacific intertidal areas are predominantlyexposed mudflats at low tide, and Spartina is rapidlycolonizing <strong>the</strong>se areas. Its effects range from <strong>the</strong> reduction<strong>of</strong> valuable feeding habitat for migratory shorebirds to <strong>the</strong>possible collapse <strong>of</strong> some areas’ shellfishing industries.Traditional control methods such as mowing, tilling, andherbicide application were not yet effective or economical at<strong>the</strong> writing <strong>of</strong> this article, and a long-term solution was stillbeing sought. In 2000, a biological control agent, <strong>the</strong>planthopper Prokelisia marginata Van Duzee (Homoptera:Delphacidae), was released in Willapa Bay, Washington, tocontrol Spartina alterniflora Loisel. (Grevstad et al. 2003),but <strong>the</strong> invasive population was not brought under controldespite establishment <strong>of</strong> this control agent.Past studies documenting <strong>the</strong> insect fauna in Spartinamarshes (Davis and Gray 1966; Denno 1977; Vince et al.1981; Newton 1984) have demonstrated <strong>the</strong> diversity <strong>of</strong> <strong>the</strong>associated insect communities, but relatively few havefocused on New England marshes. The purpose <strong>of</strong> this studywas to ga<strong>the</strong>r information on <strong>the</strong> ecology <strong>of</strong> insec<strong>the</strong>rbivores <strong>of</strong> S. alterniflora in Rhode Island in support <strong>of</strong> alarger effort to develop biological control agents for releasein Willapa Bay.METHODSThirteen Rhode Island salt marshes (Fig. 1) wereFig. 1: Vacuum samples were collected at sites marked with a circle.Vacuum and stem samples were collected at sites marked with a star.- 273 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaTable 1: Vacuum Survey Results: Number <strong>of</strong> salt marshes, <strong>of</strong> 13 sampled,inhabited by known Spartina herbivores by zone. Feeding Guilds: (LC)Leaf-chewer; (SS) Sap-sucker; (SB) Stem-borer.SpeciesFeedingGuildLowMarshZoneHighMarshOrchelimum spp. LC 2 5Prokelisia dolus SS 3 13Prokelisia marginata SS 10 10Trigonotylus uhleri SS 11 10Chaetopsis aenea SB 11 11Chaetopsis apicalis SB 12 10sampled once from mid July to late September using aHomelite ® gas-powered leaf blower/vacuum with a finemesh net fitted inside <strong>the</strong> intake tube. Samples were taken bywalking along transects (approximately 30 meters (m) atboth low and high marsh elevations) and sweeping <strong>the</strong>vacuum over S. alterniflora plants from base to tip. Sampleswere frozen prior to being sorted. Specimens <strong>of</strong> interest were<strong>the</strong>n mounted and identified.At a subset <strong>of</strong> five marshes, <strong>the</strong> frequency <strong>of</strong> damagedue to specific species or feeding guild was measured. Thetotal number <strong>of</strong> stems in a square meter quadrat was countedinitially. Fifty stems chosen at random were harvested aswell as every stem showing apical leaf damage. Fourreplicates were sampled at each site. Stem height anddiameter were recorded before <strong>the</strong> stem was scored forexternal feeding damage and <strong>the</strong>n dissected for stem-boringlarvae. Data were analyzed using Micros<strong>of</strong>t Excel andAnalyse-it ® supplementary s<strong>of</strong>tware.RESULTSSpecimens collected in vacuum samples representedfour genera <strong>of</strong> Spartina herbivores in three feeding guilds(Table 1; Fig. 2). Of <strong>the</strong> three guilds, sap-sucking and stemboringspecies were collected at all but a few sites, while <strong>the</strong>sole leaf-chewing genus, Orchelimum Serville (Orthoptera:Tettigoniidae), was recovered rarely. Prokelisia dolusWilson was also anomalous in that it was found much morefrequently at high marsh than at low marsh. Additionalspecies observed during stem dissections were not recoveredat all in vacuum samples.Harvested stem analysis revealed disparities in <strong>the</strong>infestation rates across species or feeding guild (Fig. 3).Number <strong>of</strong> Species8765432Leaf-chewerStem-borerSap-sucker10FoglandEmilie RueckerTouissetNagPondMarsh MeadowsSheffield CoveFox HillRome PointNarrow RiverGalileeSuccotashWest BeachLathropFig. 2: Composition <strong>of</strong> Spartina alterniflora herbivore feeding guilds based on vacuum samples collected at thirteen marshes.- 274 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementInfestation Rate ± SE1.0000.8000.6000.4000.200Emilie RueckerFox HillGreenwich CoveSuccotashTouisset0.000ChaetopsisLanguria taedataCalamomyia alternifloraeHaliaspisProkelisia / TrigonotylusOrchelimumFig. 3: Infestation rates <strong>of</strong> Spartina herbivores at five marshes. Feeding damage caused by Prokelisia spp. and that caused byTrigonotylus uhleri were not distinguishable.There were significant differences among sites in <strong>the</strong> rates <strong>of</strong>infestation by Chaetopsis spp. Loew (Diptera: Otitidae) (p =0.0007), Languria taedata LeConte (Coleoptera:Languriidae) (p = 0.0025), Calamomyia alterniflorae Gagne(Diptera: Cecidomyiidae) (p < 0.0001), Haliaspis spp.Takagi (Homoptera: Diaspididae) (p < 0.05), and <strong>of</strong> leafchewingdamage (p < 0.0001). Infestation by Chaetopsisspp. was associated with an average stem height reduction <strong>of</strong>24.8%, although <strong>the</strong>se results were not significant.DISCUSSIONThe results <strong>of</strong> vacuum sampling suggest that <strong>the</strong>re islittle variation in S. alterniflora herbivore communitycomposition among Rhode Island marshes. The apparentrarity <strong>of</strong> Orchelimum spp. and <strong>the</strong> absence <strong>of</strong> o<strong>the</strong>r speciesin vacuum samples was likely due to limitations <strong>of</strong> <strong>the</strong>sampling methods. Vacuum sampling used in this surveymay have provided more opportunity for evasion than sweepnetting used in previous studies (Davis and Gray 1966;Vince et al. 1981). Seasonal variation in communitystructure is also <strong>of</strong> importance as adult forms <strong>of</strong> many stemboringspecies are only present in limited intervals (Newton1984). To alleviate <strong>the</strong> inherent biases <strong>of</strong> this samplingmethod, surveys <strong>of</strong> community composition should include awider array <strong>of</strong> sampling methods and increased frequency <strong>of</strong>sampling throughout <strong>the</strong> season.Analysis <strong>of</strong> feeding damage to S. alterniflora stemsprovides sufficient evidence to warrant fur<strong>the</strong>r investigation<strong>of</strong> some species as possible biological control agents.Although sap-sucking insects caused <strong>the</strong> highest frequency<strong>of</strong> damage, <strong>the</strong>y did not cause <strong>the</strong> severity <strong>of</strong> damageassociated with <strong>the</strong> infestation by Chaetopsis aeneaWiedemann and C. apicalis Johnson. The larvae <strong>of</strong> <strong>the</strong>seflies destroy <strong>the</strong> shoot apical meristem <strong>of</strong> all infested stems,effectively stopping growth and preventing flowering. Thereexists, however, significant variation in <strong>the</strong> proportion <strong>of</strong>stems damaged by Chaetopsis spp. both among sites andwithin sites. This variation may be explained by differencesin variables such as plant nutrition, plant resistance, tidalfluctuation, and water salinity. Newton (1984) demonstratedthat salinity was strongly correlated to <strong>the</strong> infestation rates <strong>of</strong>several stem borers. Denno et al. (2000) investigated- 275 -

Chapter 4: Spartina Control and Management<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartinachanges in intra-guild competition among sap-suckingspecies due to variations in plant resistance. Until <strong>the</strong> degree<strong>of</strong> niche overlap between Chaetopsis species is betterunderstood, this sort <strong>of</strong> interaction ought to be consideredpossible.Future studies should focus on ga<strong>the</strong>ring additionalinformation on <strong>the</strong> potential effectiveness <strong>of</strong> Chaetopsis spp.as biological control agents. A major component <strong>of</strong> thisresearch should seek to explain <strong>the</strong> variation in Chaetopsisinfestation rates in an attempt to understand what factors arelimiting <strong>the</strong> species where it is found at lower densities. Itwill be important to quantify <strong>the</strong> effect <strong>of</strong> Chaetopsisinfestation on <strong>the</strong> productivity <strong>of</strong> S. alterniflora genets andto learn which ecological traits distinguish C. aenea from C.apicalis. Lab rearing techniques using a modified protocolfrom Allen & Foote (1992) currently in development at <strong>the</strong>University <strong>of</strong> Rhode Island should aid future work.ACKNOWLEDGEMENTSThis research was supported by <strong>the</strong> Coastal FellowsProgram and <strong>the</strong> Biological Control Laboratory at <strong>the</strong>University <strong>of</strong> Rhode Island.REFERENCESAllen, E.J. and B.A. Foote. 1992. Biology and immature stages <strong>of</strong>Chaetopsis massyla (Diptera: Otitidae), a secondary invader <strong>of</strong>herbaceous stems <strong>of</strong> wetland monocots. <strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> EntomologicalSociety <strong>of</strong> Washington 94: 320-328.Davis, L.V. and I.E. Gray. 1966. Zonal and seasonal distribution <strong>of</strong>insects in North Carolina salt marshes. Ecological Monographs36: 275-295.Denno, R.F. 1977. Comparison <strong>of</strong> <strong>the</strong> assemblages <strong>of</strong> sap feedinginsects (Homoptera-Hemiptera) inhabiting two structurallydifferent salt marsh grasses in <strong>the</strong> genus Spartina. EnvironmentalEntomology 6: 359-372.Denno, R.F., M.A. Peterson, C. Gratton, J. Cheng, G.A. Langellotto,A.F. Huberty, and D.L. Finke. 2000. Feeding-inducedchanges in plant quality mediate interspecific competition betweensap-feeding herbivores. Ecology 81: 1814-1827.Grevstad, F.S., D.R. Strong, D. Garcia-Rossi, R.W. Switzer andM.S. Wecker. 2003. Biological control <strong>of</strong> Spartina alterniflora inWillapa Bay, Washington using <strong>the</strong> planthopper Prokelisiamarginata: agent specificity and early results. Biological Control27: 32-42.Newton, N.H. 1984. Stem-Boring Insect Larvae <strong>of</strong> Spartina alternifloraLoisel (Poaceae): Distribution, Biology and Influenceon Seed Production. Ph.D. Dissertation, North Carolina StateUniversity, Raleigh.Vince S.W., I. Valiela and J.M. Teal. 1981. An experimental study<strong>of</strong> <strong>the</strong> structure <strong>of</strong> herbivorous insect communities in a saltmarsh. Ecology 62: 1662-1678.- 276 -

<strong>Proceedings</strong> <strong>of</strong> <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive SpartinaChapter 4: Spartina Control and ManagementPOTENTIAL FOR SEDIMENT-APPLIED ACETIC ACID FOR CONTROL OF SPARTINAALTERNIFLORAL.W.J. ANDERSONUSDA-Agricultural Research Service, Exotic and Invasive Weed Research, One Shields Ave., Davis, CA 95616;lwanderson@ucdavis.eduNote: This paper, presented at <strong>the</strong> <strong>Third</strong> <strong>International</strong> <strong>Conference</strong> on Invasive Spartina, was subsequently published in 2007. Thecitation and abstract <strong>of</strong> <strong>the</strong> published paper is provided below.CitationAnderson, L.W.J. 2007. Potential for Sediment-Applied Acetic Acid for Control <strong>of</strong> Spartina alterniflora.J. Aquat. Plant Manage. 45:100-105 [http://www.apms.org/japm/vol45/v45p100.pdf]AbstractSmooth cordgrass (Spartina alterniflora), a tall grass native to <strong>the</strong> east coast, has invaded Willapa Bay,Washington, and <strong>the</strong> San Francisco Bay, California. Management with glyphosate and imazapyr can beeffective, but in <strong>the</strong> San Francisco populations, applications in several sites are confined to short periodsin <strong>the</strong> fall in order to protect nesting habitats <strong>of</strong> Clapper rails (Rallus longirostris). Use <strong>of</strong> efficacioussoil-active herbicides could mitigate this restriction. Acetic acid, a readily degraded natural product, hasbeen shown to kill sediment-borne propagules <strong>of</strong> aquatic plants such as Hydrilla verticillata andStuckenia pectinatus. Effects <strong>of</strong> acetic acid on sediment-free rhizomes <strong>of</strong> S. alterniflora were examined.Exposure <strong>of</strong> 0.1, 1.0, 1.5% vol/vol acetic acid for a few hours to several hours resulted in increasedconductivity in distilled water compared to unexposed controls, indicating loss <strong>of</strong> cellular integrity andleakage <strong>of</strong> electrolytes. Regrowth from exposed rhizomes was significantly inhibited at higher (1.0% and1.5%) concentrations applied for 2 or 4 hr. When rhizomes that had been directly exposed to 1.5% aceticacid were transferred to outdoor conditions in Albany, CA, both new shoot number and average plan<strong>the</strong>ight were reduced by over 90% at nine months post-treatment. The exposure to all concentrations <strong>of</strong>acetic acid for 4 hr also led to reduced frequency <strong>of</strong> inflorescence production, thus potentiallydiminishing <strong>the</strong> dispersal capacity <strong>of</strong> <strong>the</strong> treated plants. Field trials are needed to determine if judiciousdrenching <strong>of</strong> sediments with acetic acid (e.g., at low tide) may have utility as an alternative to foliarapplied herbicides such as imazapyr and glyphosate.Keywords: soil-active herbicide, electrolyte, pore-water, HPLC, seawater, smooth cordgrass, vinegar.- 277 -

Proceedings of the Third International Conference on Invasive ...

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?