This thesis has been submitted in fulfilment of the requirements for a postgraduate degree (e.g. PhD, MPhil, DClinPsychol) at the University of Edinburgh. Please note the following terms and conditions of use: This work is protected by copyright and other intellectual property rights, which are retained by the thesis author, unless otherwise stated. A copy can be downloaded for personal non-commercial research or study, without prior permission or charge. This thesis cannot be reproduced or quoted extensively from without first obtaining permission in writing from the author. The content must not be changed in any way or sold commercially in any format or medium without the formal permission of the author. When referring to this work, full bibliographic details including the author, title, awarding institution and date of the thesis must be given. The biogeographic affinities of the Sri Lankan flora A thesis submitted for the degree of Doctor of Philosophy Lakmini Darshika Kumarage Institute of Molecular Plant Sciences, University of Edinburgh Royal Botanic Gardens, Edinburgh January 2016 Abstract The island of Sri Lanka’s exceptional biodiversity and enigmatic biogeography begs investigation, as the island is key in understanding the evolution of the Asian tropical flora. Since the Jurassic, Sri Lanka has been subjected to remarkable tectonic changes, thus its flora could have been influenced by that of a number of nearby landmasses, as well giving Sri Lanka the potential to have played a wider role in the assemblage of floras elsewhere. Firstly, as Sri Lanka originated as a fragment of the supercontinent Gondwana, part of its flora may contain Gondwanan relict lineages. There is also the potential for immigration from Laurasia after the Deccan Plate collided with it 45-50 Mya. Further, Sri Lanka may harbour floristic elements from nearby land masses such as Africa and Southeast Asia as a result of long distance dispersals, and in situ speciation has the potential to have played an important role in enhancing the endemic Sri Lankan flora. I tested the relative contributions of the above hypotheses for the possible origins of the Sri Lankan flora using three representative families, Begoniaceae, Sapotaceae and Zingiberaceae. These families represent both herbaceous and woody elements, and have high diversity across the tropics. Dated molecular phylogenies were constructed for each family. I used recent analytical developments in geographic range evolution modelling and ancestral area reconstruction, incorporating a parameter J to test for founder event speciation. A fine scale area coding was used in order to obtain a better picture of the biogeography of continental Asia. Amongst all the models compared, a dispersal-extinction cladogenesis model incorporating founder event speciation proved to be the best fit for the data for all three families. The dates of origin for Sri Lankan lineages considerably post-date the Gondwanan break up, instead suggesting a geologically more recent entry followed by diversification of endemics within the island. The majority of Sri Lankan lineages have an origin in the Sunda Shelf (53%). Persistence of warm temperate and perhumid climate conditions in southwestern Sri Lanka resembling those of Peninsular Malaysia and Sumatra could have facilitated suitable habitats for these massive dispersals from the Sunda Shelf region. Some trans-oceanic long distance dispersals from Africa (11%) are also evidenced, again these are too young to accept a hypothesis of dispersal during i the Deccan Plate’s migration close to the African coast during the late Cretaceous, but occurred later during the Miocene. Further, some lineages of Laurasian origin (20%) are evidenced in the Zingiberaceae with ancestral areas of China and Indochina, which is congruent with a post collision invasion. Among the families tested, dispersals have occurred stochastically, one during the Eocene, six during the Oligocene, seven during the Miocene, two during the Pliocene and one during the Pleistocene. The highest number of dispersals occurred during the Miocene when a warm climate was prevailing during the Miocene thermal maximum. My results confirm that in situ speciation is an important contributor to the Sri Lankan flora. More rapid radiation of endemics has occurred during Pliocene-Pleistocene; two endemics in Begoniaceae, ten endemics in Sapotaceae and ten endemics in Zingiberaceae have evolved in situ during this period. Sri Lanka will have been subjected to expansion and contraction of climatic and vegetation zones within the island during glacial and interglacial periods, potentially resulting in allopatric speciation. As a conclusion, long distance dispersals have played a prominent role in the evolution of the Sri Lankan flora. The young ages challenge the vicariant paradigm for the origin and current disjunct distributions of the world’s tropical lineages and provide strong evidence for a youthful tropics at the species level. The thesis contains six chapters; first two are introductory chapters, then there are three analytical chapters, one for each family, and finally a summary chapter is provided. Each analytical chapter is written as a stand-alone scientific publication, thus there is some repetition of relevant content in each. ii iii Acknowledgements Firtslty I would like to give my sincere thanks and gratitude to Dr Mark Hughes, Dr James Richardson and Dr Richard Milne for their immense support and advices throughout my project from the beginning to the end. They were always very helpful and have always made the time to discuss my ideas, to answer my questions. A special thank goes to Dr Mark Hughes who was with me in my ups and downs, helping me to face difficult situations and finding solutions. He kept me on trackevery time, encouraged me and was so kind to visit Sri Lanka twice to help me in the thesis work. Without his huge support and guidance this would not have been possible. Also I like to thank Dr Sumudu Rubasinghe, University of Peradeniya, Sri Lanka for helping me in applying for permits in palnt collections in Sri Lanka. Funding for this research came from the Darwin Trust of Edinburgh which covered my tuition fee, lab costs and living expenses. I am immensely grateful to the founder of the schoralship late Prof Kenneth Murray and interview panel for selecting me to offer the shoclarship which provided opportunity for me to pursue a PhD in University of Edinburgh. Funding for field work in Sri Lanka was facilitated by the Davis expedition fund, I am thankful to them to offering me the funds to carry out the field work successfully. Also a big thanks goes to the forest Department and the wildlife department of Sri Lanka for giving the permission to collect plant specimens and Upali, Tharanga, Indika and many other people who helped me in the field. A special thank you goes to Royal Botanic Gardens, Edinburgh which provided me the opportunity to join the Tropical Group and working space in the molecular biology laboratories. Also I am grateful to all members at RBGE for their kindness, smiles, and support throughout the period. More specially, Dr Michelle Hart, Dr Laura Forrest, Ruth Hollands for their help and advice in lab and always willingness to help and Deborah Vaile for her support in the library. I am grateful to everyone who has helped with the collection of plant material, provided silica dried samples, DNA isolations and sequence data during the course of my research. Notable mentions go to Dr Daniel Thomas, Dr Jane Droop for providing the DNA sequence alignments and Dr Mark Newman for his support in Zingiberaceae iv specimen identification and sharing his expertise. Also I am thankful to Dr Sangeetha Rajbhandary, Dr Ching-I-Peng and Koh Nakamura for providing me the Himalayan Begonia DNA samples. Further, I am grateful to all the academic and non academic staff members of the Open University of Sri Lanka for their support in various forms for me to carry out my research very well. Thanks also go to all my collegues and my fellow PhD students for their moral support and humour and for making the PhD room more pleasant. I am thankful to all my family members for their immense support and encouragement in all my achievements. And last but not least I am grateful to Ayon, being so good, allowing me to work freely and Kumari for looking after him very well.More specially thanks goes to Sahan, who looked after Ayon very well giving freedom for me to work on my research. Without his love, encouragement, understanding and support, none of this would have been possible. v TABLE OF CONTENTS CHAPTER 1: Introduction to Biogeography………………..…………….. 1 What is Biogeography?................................................................................................. 1 1.2 The early history of biogeography as a discipline………………………….………..…………. 1 1.3 A review of historical biogeographic methods………………………………………….………. 3 1.3.1 Paradigm shifts key to historical biogeographic analysis………………..…………. 3 1.3.2 The beginnings of a modern synthesis of analytical historical biogeography 8 1.3.3 Approaches to analytical historical biogeography……………………………………. 9 1.4 The current state of the field…………………………………………………………………………….. 15 1.5 Summary…………………………………………………………………………………………………………… 15 CHAPTER 2: Introduction to Sri Lankan biogeography………………. 17 2.1 Current Geology and climate of Sri Lanka………………………….………………………………. 17 2.2 Palaeoclimate, geological changes………………………………….…………………………………. 19 2.3 Flora of Sri Lanka…………………………………………………………………………………..………….. 20 2.4 Possible origins of Sri Lankan Flora……………………………………………………………………. 23 2.5 Study groups……………………………………………………………………………………………………… 25 2.5.1 Family Begoniaceae……………………………………………………………………………………. 25 2.5.2 Family Sapotaceae……………………………………………………………………………………… 26 2.5.3 Family Zingiberaceae………………………………………………………………………………….. 27 2.6 Distribution of Begoniaceae, Sapotaceae and Zingiberaceae in Sri Lanka…………. 29 2.7 Field collections in Sri Lanka…………………………………………………………………….………… 34 2.8 Species portrait: Rediscovery of Amomum nemorale (endemic)…………………..…… 36 2.9 Aims of the doctoral research and structure of the thesis………………………………….. 37 vi CHAPTER 3: The early evolution and mode of range expansion in Asian Begonia……………………………………………………………………………… 46 CHAPTER 4: Biogeographic history of Sri Lankan Sapotaceae………. 84 CHAPTER 5: Pantropical biogeography of Zingiberaceae……………… 127 CHAPTER 6: Summary and Conclusions……………….……………………… 165 vii APPENDICES Appendix 1: List of taxa used in the biogeographic analysis of Begoniaceae and Voucher numbers. 188 Appendix 2: Presence/ Absence coding of each taxa of Begoniaceae in the biogeographic analysis in Biogeobears. 198 Appendix 3: List of taxa used in the biogeographic analysis of Sapotaceae and Voucher numbers. 206 Appendix 4: Presence/ Absence coding of each taxa of Sapotaceae in the biogeographic analysis in Biogeobears. 214 Appendix 5: List of taxa used in the biogeographic analysis of Zingiberaceae, Voucher number and GenBank accession number of gene sequences. 220 Appendix 6: Presence/ Absence coding of each taxa of Zingiberaceae in the biogeographic analysis in Biogeobears. 226 Appendix 7: Example script used in Biogeobears in ancestral area reconstructions. 231 Appendix 8: DNA sequence alignments and Biogeobears result output for Begoniaceae, Sapotaceae, Zingiberaceae. 256 viii LIST OF FIGURES Figure 1.1 The processors assumed by different biological methods. …...………… 13 Figure 2.1 Geographic regions in Sri Lanka………………………………………… 18 Figure 2.2 Break-up of Gondwana showing vicariance and collision times between continental fragments. …………………………………………………… 20 Relationships between Gondwanan fragments and time of phylogenetic splits based on paleogeographic data ………………..………………….. 25 Figure 2.4 Species distribution of Begonia in Sri Lanka…………………………….. 29 Figure 2.5 Distribution of Palaquium species in Sri Lanka…………...…………….. 29 Figure 2.6 Distribution of Isonandra species in Sri Lanka…………..……………… 30 Figure 2.7 Distribution of Madhuca species in Sri Lanka…………………………… 30 Figure 2.8 Distribution of Mimusops species in Sri Lanka………………………….. 31 Figure 2.9 Distribution of Manilkara hexandra in Sri Lanka………………………... 31 Figure 2.10 Distribution of Xantolis tomentosa in Sri Lanka…………………………. 32 Figure 2.11 Distribution of Alpnia fax, Alpinia abundiflora and Cyphostigma pulchellum species in Sri Lanka………..………………….. 32 Figure 2.12 Distribution of Amomum species in Sri Lanka……….………………….. 33 Figure 2.13 Distribution of Curcuma and Zingiber species in Sri Lanka…….………. 33 Figure 2.14 Localities of field collections in Sri Lanka………………………………. 34 Figure 2.15 Habit and inflorescence of Amomum nemorale……………………....… 36 Figure 3.1 Phylogenetic representations showing the history of four hypothetical routes of entry of Begonia to Asia from Africa, and dispersal within Asia 52 Figure 3.2 Bayesian majority rule consensus tree resulted from non coding region of plastid DNA: Begoniaceae. …………………………………………... 62 Maximum-clade-credibility chronogram of a beast analysis of the non coding region Begonia data set showing optimal range reconstructions under the DEC+J model in the package Biogeobears ...………………..... 63 Maximum clade credibility chronogram of the non coding region 64 Figure 2.3 Figure 3.3 Figure 3.4 ix Begonia dataset showing 95% highest posterior density date ranges Figure 4.1 Figure 4.2 Figure 4.3 Figure 4.4 Figure 5.1 Figure 5.2 Figure 5.3 Figure 5.4 Bayesian majority rule consensus tree based on nuclear ITS data of Sapotaceae. Bayesian posterior probability (PP) support values are indicated next to the nodes……………….………………………………. 105 Maximum clade credibility chronogram of a Beast analysis of the ITS region Sapotaceae ……………………………………………………... 106 Maximum clade credibility chronogram of Sapotaceae ITS dataset showing 95% highest posterior density data ranges…..………………… 107 Maximum clade credibility chronogram of a Beast analysis of the Sapotaceae chloroplast dataset ………………………………………… 108 Bayesian majority rule consensus tree based on nuclear ITS data Zingiberaceae…………………………………………………… 140 Maximum-clade-credibility chronogram of a relaxed molecular clock analysis of the ITS data set Zingiberaceae…………………………… 141 Maximum clade credibility chronogram of the Zingiberaceae ITS dataset showing 95% highest posterior density data ranges…………………… 142 Maximum-clade-credibility chronogram of a relaxed molecular clock analysis of the matK data setZingiberaceae……………………………. 143 Figure 5.5 Histogram representing the incongruence between ITS and matK phylogenies……………………………………………………..………. Figure 6.1 Different geographic regions in world and collision times between continental fragments..…………………………………………………. 174 Figure 6.2 Present positions of continental plates and geographic boundaries…… 174 Figure 6.3 Reconstructed geographic origins for Sri Lankan flora and their relative contributions……………………………………………………………. 177 Figure 6.4 Percentage of dispersals in terms of geological epochs………………… 178 Figure 6.5 A plot of the minimum ages of Sri Lankan clades from the present study and other published studies on plants (Hortonia) and animal groups (toads) ……………………………………………………………………. 180 x 144 LIST OF TABLES Table 1.1 Characters of different ancestral area reconstruction methods .......... 14 Table 2.1 List of species for each study group, ecological status and distribution. 27 Table 2.2 Species collected in each family during the field work in Sri Lanka.….. 35 Table 3.1 Voucher information for newly generated sequences for the study Begoniaceae………………………………………………………………………………….……. 55 Table 3.2 Primers used in the study for amplification of plastid DNA……………..…. 56 Table 3.3 Dataset descriptive statistics Begoniaceae …………………………………..…… 59 Table 3.4 d (dispersal), e (extinction), j (j value, founder-event speciation) LnL (log likelihood) for each of the geographic range evolution models compared in Biogeobears for Begoniaceae …………………………….………… Table 3.5 61 Posterior probabilities, divergence ages and ancestral area probabilities in BGB for Begoniaceae ………………..…………….…………………. 66 Table 4.1 Voucher specimens of the newly generated ITS sequences Sapotaceae 90 Table 4.2 Voucher information for newly generated chloroplast sequences (trnH-psbA, trnC-trnD, ndhF) ……………………….…………………….……………….. Table 4.3 91 Nuclear (ITS) and chloroplast primer sequences used in Sapotaceae DNA sequence generation ……………………………………………………………….…. 96 Table 4.4 Descriptive statistics of ITS and plastid data for Sapotaceae …………..….. 100 Table 4.5 Posterior probabilities, divergence ages and ancestral area probabilities in BGB for Sapotaceae …………….…………………………….…..…. Table 4.6 102 d (dispersal), e (extinction), j (j value, founder-event speciation) LnL (log likelihood) for each of the geographic range evolution models compared in Biogeobears for Sapotaceae ……………………………...…………. xi 104 Table 5.1 Division of the family Zingiberaceae ………………………………………………… Table 5.2 Voucher information for the newly generated sequences for Zingiberaceae………………………………………………………………………….………. Table 5.3 127 131 Nuclear (ITS) and chloroplast primer sequences used in Zingiberaceae DNA sequence generation ………………………………………………….……….…. 133 Table 5.4 Descriptive statistics of nuclear ITS and matK datasetsZingiberaceae... 136 Table 5.5 Posterioir probabilities, Divergance ages and Ancestral Area probabilities in BGB for Zingiberaceae ……………………………….….……… Table 5.6 145 d (dispersal), e (extinction), j (J value, founder-event speciation) LnL (log likelihood) for each of the models compared in Biogeobears for Zingiberaceae ………………………………………………………………………….……. Table 6.1 Species collected in Begoniaceae, Sapotaceae and Zingiberaceae during the field work in Sri Lanka. …………………….…………………………… Table 6.2 165 Newly generated sequences for the present study for Begoniaceae, Sapotaceae and Zingiberaceae ……………………………………………………… Table 6.3 146 167 A summary for reconstructed geographic origins for the tested families, and the number of dispersals during different geological epochs …………………………………………………………………………………………. Table 6.4 176 Hypothesized rafting times across the Atlantic Ocean, Caribbean Sea and Southeast Indian Ocean during different periods throughout the Tertiary…………………………………………………………………………………………. xii 182 CHAPTER 1: Introduction to Biogeography 1.1 What is Biogeography? The Earth is a diverse planet inhabited by millions of species found in many different habitats such as forests, deserts, deep seas and even some extreme environments like thermal vents. Since the beginning of natural philosophy, we have tried to understand and explain these different distributions of organisms over the earth’s surface and the possible causes for such distributions. These endeavours developed into a new science Biogeography. Biogeography is the study of the distribution of organisms in space and time. Some organisms inhabit wide ranges all over the world and are termed cosmopolitan, while others are restricted to a specific area and are termed endemic. Biogeography is a vast field of biology that requires information from other disciplines like systematics, molecular biology, evolutionary biology, ecology and geology. Recent advances in the field have developed new analytical methods to understand the causes of differing distributions of species over the earth surface. 1.2 The early history of biogeography as a discipline Aristotle was the first scientist to ask central questions about biology, and also the first to present a view of a dynamic and changing world. However the consensus view until the mid eighteenth century was that the earth was largely static, in combination with a belief in God’s creation of living organisms. Carolus Linnaeus (1707-1778) considered living organisms to be immutable, and his explanation of their varying distributions involved a centre of origin on a paradisical mountain after the biblical deluge. He accepted that species originated on the slopes of Mountain Ararat and they were adapted to live on various habitats at different elevations. Once the flood receded they were capable of migrating downward and spreading throughout the continents. Georges-Louis Buffon’s (1707-1788) idea of the geographic origin of life was in contrast to that of Linnaeus, hypothesized to be near the North Pole during a period when the climate was much warmer. When the environment cooled, species migrated south and the species capable of inhabiting newly accessible environments survived, 1 while others died. His observation that environmentally similar but isolated regions have distinct assemblages of mammals and birds became called Buffon’s law (Buffon, 1761, 1776) that is accepted as the first principle of biogeography (Lomolino et al. 1998). Sir Joseph Banks (1743-1820) was a naturalist who accompanied Captain James Cook on his three year voyage around the world which permitted him to collect around 3600 plant specimens, including 1000 new to science. The discoveries by Banks confirmed Buffons Law of different regions having different biotas, but in contrast also found some cosmopolitan species showing wide distributions. His discoveries shed new light on the complexity of the living world (Lyte 1981). Johann Reinhold Forster (1729-1798) came up with new ideas in the field of biogeography which lead to the application of Buffon’s law to plants, birds and mammals not only in the tropics but also throughout the globe. He was the first scientist to publish systematic descriptions of the world’s biotic regions based on plant assemblages and his explorations on patterns of species diversity are considered as the foundation of ecology. He observed that the species diversity decreases towards the poles from the equator and that the species diversity of an island is lower than the mainland and depends on available resources (Forster 1996). The German botanist Karl Ludwig Willdenow (1765-1812) came up with contrasting ideas to the single-site origin of life. He proposed that there had been several refuges for plants on mountains, each with a distinct flora that eventually spread throughout the globe giving rise to new species communities (Willdenow 1811). The 19th century is considered as a golden age for biogeography with an expansion of contributions from great naturalists and botanists. Alexander von Humboldt (17691859) is considered as the father of phytogeography, and one of the great naturalists of his time who had new views on floristic distributions. Through surveys conducted in the Andes he discovered distinct floristic elevational gradients that he named ‘floristic belts’ (von Humboldt 1805, 1808). Later, Augustin P de Candolle (1778-1841) made an immense contribution to the field and made further explanations to Johann Reinhold Forster’s views by incorporating 2 island age, volcanism, isolation and climate as factors determining an island’s biodiversity. His view of species competing for natural resources in the environment for their survival and coexistence can be considered as the basis for the theory of evolution and ecology (de Candolle 1820, 1855). Charles Lyell (1797-1875) was a great 19th paleobotanist who used the fossil record to draw inferences regarding ancient climates. His invaluable findings on the nature of earth’s surface led him to be regarded as the father of geology. He was amongst the first to propose that the earth was ancient, and his findings revealed that the earth’s surface and biotic component is not a static system, but dynamic (Lyell 1830). The scene was set in the latter half of the 19th century for contributions from scientists such as Charles Darwin, Alfred Russel Wallace, Joseph Hooker and Philip Sclater to come up with novel interpretations of the living world and to provide the basis for a modern synthesis of biogeography. 1.3 A review of historical biogeographic methods 1.3.1 Paradigmshifts key to historical biogeographic analysis Theory of evolution Historically species were believed to be immutable and the continents and oceans were believed to be fixed and stable. Before biogeography could mature as a discipline, a modern, realistic view of geology and evolution was necessary. The great naturalist Charles Darwin made an enormous contribution to the field of biogeography through his theory of evolution, based largely on observations made during his five year voyage on HMS Beagle. He reasoned that the occurrence of different forms of tortoises and finches on the different islands in the Galapagos archipelago were the result of geographic isolation and natural selection, leading him to propose the theory of evolution (Darwin 1859). A British naturalist, Alfred Russel Wallace (1823-1913) came up independently with the theory of evolution through natural selection from the observations made during extensive field work done in Amazone river basin and Malay Archipelago (Wallace 3 1860, 1863, 1869, 1876). His ideas prompted Darwin to publish his ideas in “On the Origin of Species”. The laws of inheritance were synthesized by Gregor Johann Mendel (1822-1884) based on his experiments using pea plants, where he explained how traits could be passed from one generation to the other. Mendel’s population genetic theory combined with Darwin’s theory of natural selection lead to the modern synthesis of evolutionary theory, the current paradigm in evolutionary biology. Plate tectonics The true age of the earth and its tectonic history only became apparent and widely accepted during the latter half of the twentieth century. Sir Francis Bacon is accepted as the first person to have views of moving continents, but his ideas were never exposed for discussions (Lomolino et al. 1998). The German scientist Alfred L. Wegener developed the theory of continental drift based on the observation of congruent patterns of coastlines and coal deposits across the Atlantic Ocean. According to his hypothesis, during the early history of earth, all the landmasses were connected to each other forming a super-continent called Pangaea that began to separate during Mesozoic. Wegener’s hypothesis (Wegener 1912, 1966) matured into the theory of plate tectonics, further developed by Oliver (1968). The earth is now understood to be a complex, dynamic structure composed of continental crust, mantel and core. The forces generated from the intense heat produced beneath the earth’s crust result in lateral plate movements and seafloor spreading. Plate tectonics revolutionised the field of biogeography through providing a potential mechanism for explaining disjunct distributions (Lomolino et al. 1998). Phylogenetics Phylogenetics is the representation of the relationships and history of evolution of organisms in the form of a bifurcating tree. The German biologist, Willi Hennig (19131976) is considered to be the founder of phylogenetic systematic and to begin with it was largely dependent on morphological characters (Hennig 1966). Cladistics revolutionaised phylogenetic analysis, where it builds on the fact that the members of a group who share a common evolutionary history, and are more "closely related" to 4 members of the same group than to other organisms. In cladistics, any group of organisms arerelated by descent from a common ancestor, and they exhibit a bifurcating pattern of cladogenesis. Common shared characters are called as Synplesiomorphic while changes in characters occurs in lineages results in shared derived characters so called Synapomorphies. The advent of advanced technologies such as PCR and DNA sequencing allowed the development of molecular phylogenetics which revolutionised the field providing more strongly supported trees. Both nuclear and plastid DNA has been extensively utilized in the field of molecular phylogenetics to infer relationships between organisms at different taxonomic levels. Non coding region of the chloroplast genome are of interest since they evolve more rapidly than coding regions and mutations are accumulated exhibiting sufficient variation to be phylogenetically informative (Borsch & Quandt 2009). Chloroplast markers provide a useful tool at higher taxonomic levels, while internal transcribed spacer (ITS) regions of nuclear ribosomal DNA has been proved to be useful in inferring lower phylogenetic relationships such as interspecific variations due to its rapidly evolving nature (Kim et al. 1999; Baldwin et al. 1997) providing more phylogenetically informative characters. Further, ITS provide a better resolution for relatively young or rapidly radiated genera where chloroplast markers tend to fail to achieve resolution. Phylogenetic trees constructed using sequence data can be either distance based or character based. Distance based methods utilize a distance matrix which is obtaining by calculating genetic distance between each pair of sequences. Character based methods are of three inference types, ie Maximum Parsimony (MP), Maximum Likelihood (ML) or Bayesian methods (BI). Both likelihood and Bayesian methods are model based while parsimony assumes all possible mutations are equally likely. Parsimony has been widely used for phylogenetic studies due to its simplicity and the availability of software such as PAUP (Swofford 1993). The tree with minimum number of character state changes is considered the most parsimonious and the best hypothesis for the evolution of the data. When constructing deeper-time phylogenies, or data which presents a mixture of long and short branches, likelihood or Bayesian methods are preferable to avoid errors due to long branch attraction (Yang & Rannala 5 2012). In ML, a likelihood score is calculated under a substitution model and averaged over all possible states. ML calculations are computationally intensive (Guindon & Gascuel 2003; Yang & Rannala 2012), but software has been developed to cope with large datasets such as GARLI (Zwickl 2006) and RAxML (Stamatakis 2006). In Bayesian inference, posterior probabilities are calculated depending on a chosen model of DNA sequence evolution. Due to calculation of posterior probability values, further analysis to calculate bootstrap values of support for monophyletic groups is not required, making the analysis more convenient (Yang & Rannala 2012). Molecular dating The mutations accumulated in organisms over time provide a useful tool in inferring divergence times with the use of fossil data, so-called Molecular dating. Molecular dating has been widely used in the field of biogeography to test biogeographic hypothesis and to determine the possible geological scenarios for plant distributions. The use of molecular clocks in phylogenetic studies of plants began in the 1990’s (Renner 2005) and since then developments in theory and the availability of user friendly software has led to the publication of hundreds of plant phylogenetic studies dated using fossil data. In order to date a phylogeny, the genetic distance between sequences or taxa is calculated and the substitution rate is calculated by dividing the genetic distance by the age of an appropriate fossil. Finally, the substitution rate is used to obtain the absolute ages of taxa. When a fossil record from the ingroup is unavailable for calibration, a fossil from a more distant outgroup is used in a first analysis to date a node and then secondary calibration is done to bridge them to obtain absolute ages of desired taxa (Renner 2005). In an ideal case, a clocklike rate of molecular evolution is assumed, thus the substitution rate of all the branches of the phylogenetic tree assumes remains same (Rutschmann 2006; Drummond et al. 2006). However, in reality substitution rates vary over the branches because of varying generation times and mutation rates, resulting in departure from clock like evolution (Drummond et al. 2006). 6 Alternatively, divergence time is estimated by incorporating rate heterogeneity rather than trying to account for it, under the assumption of autocorrelation by estimating branch length information and modelling divergence times and rates (Rutschmann 2006). The penalised likelihood approach (PL) has been widely used, which assumes the substitution rates of two lineages either side of a node are auto correlated (Renner 2005; Rutschmann 2006). An alternative assumption is to use “local clocks”, where different substitution rates are applied to different parts of the tree by assuming that the age of calibration nodes are known (Renner 2005). The Bayesian approach to autocorrelation uses MCMC procedure to obtain posterior distribution of rates and times (Rutschmann 2006). Even though the relaxed molecular clock is useful in many cases, the need for a userdefined fully resolved tree topology in initially developed methods could be potentially disadvantageous since unresolved tree topologies could result in several equally plausible trees. Also due to alteration of posterior probabilities under relaxed clock assumption, the best tree of relaxed clock model might not be the same of strict molecular clock (Drummond et al. 2006). Errors in fossil ages, fossil misidentifications, sampling errors and tree topologies could result in erroneous divergence times. Use of large sequences, use of bootstrap values to calculate errors associated with branch lengths and assumption of strict clock model when data are approximately clock like could be used in order to obtain more reliable divergence time estimates (Renner 2005). BEAST 1.7 (Drummond & Rambaut 2007) Bayesian Evolutionary Analysis by Sampling Trees is a widely used software which implements a Markov chain Monte Carlo (MCMC) algorithms for Bayesian phylogenetic inference to date phylogenies (Drummond et al. 2012). BEAUti is the graphical user interface for BEAST which allows beast to generate dated phylogenies under given models of sequence evolution and fossil calibrations (Drummond & Ho 2007). BEAST simultaneously generates both phylogeny and molecular dates, hence allowing each to influence the other during the analysis. 7 1.3.2 The beginnings of a modern synthesis of analytical historical biogeography Dispersalism Dispersalism was considered for a long time as the major driving force for disjunctions, which states that higher taxa originated at a centre of origin; organisms then disperse and colonize new areas depending on their capabilities for survival and the physical conditions of the new environment, with speciation resulting in new species. Dispersal is always a possibility and strong evidence for dispersal is normally only provided when a barrier is older than the disjunction (Morrone & Crisci 1995). There are some striking examples of long dispersal events, such as the occurrence ofAndira, Drepanocarpus, Hernandia, Hymenaea, Sacoglottis in both South America and Africa, which can only be explained by using dispersal since the lineages are not old enough to support vicariance (Givnish & Renner 2004). The floristic assembly of some lineages of families like Annonaceae, Bromeliaceae, Rapataeceae and Myristicaceae for example is also supported by long distance dispersal (Givnish & Renner 2004; Renner 2005). Panbiogeography Croizat (1952) prompted the concept of panbiogeography, a development from the ideas of Buffon, Candolle, Hooker and Wulff (Morrone & Crisci 1995). It challenged the dispersal paradigm and used vicariance to explain disjunct distributions prior to acceptance of plate tectonic theory. Vicariance is the breaking up of a once larger distribution by a barrier of some kind. Distinctly related organisms having similar disjunct distributions at present were believed to share connecting land areas in the past. Distribution maps were drawn and areas with closely related taxa were connected to obtain biogeographic homology using track analysis method (Morrone & Crisci 1995; Humphries & Parenti1999). Even though the distributions of immediate ancestors are obtained, the actual pathway of biotic migrations is not explained by a panbiogeography approach (Nelson 1973). The movements of continents and oceans were ignored when explaining the disjunctions and the direction of migrations could not be addressed. Phylogeny and systematic knowledge was not considered and all the organisms were 8 treated equally, irrespective to their dispersal capabilities and their divergence times (Lomolino et al. 1998). Cladistic Biogeography The combination of Croizats ‘tracks’, which are actually a network or graph of hypothetical historical area connections, and Hennigs new cladistic methodology of reconstructing evolution lead to the development of cladistic biogeography (Rosen et al. 1988). This was an improvement upon previous methods, and meant phylogenetic relationships between species could be analytically incorporated, resulting in a more rigorous way to explain disjunct distributions (Morrone & Crisci 1995; Lomolino et al. 1998; Humphries & Parenti 1999). A cladistic biogeographic analysis has two steps, firstly the terminal taxa in taxon cladogram are replaced by area component to construct taxon area cladograms. Several taxon area cladograms are then combined to produce general area cladograms by using methods such as component analysis, brooks parsimony analysis, three area statements or reconciled trees (Morrone & Crisci 1995). These were assumed to represent biogeographic relationships among taxa and regions thus reveal the possible ancestral areas for the taxa. A drawback of cladistic biogeography is that incongruence between the area cladogram and the taxon cladogram is frequent due to dispersal, sympatric speciation and extinction. Even when multiple taxon cladograms have congruent topologies these may infact represent pseudo-congruence if splits occurred at different times (Donoghue & Moore 2003). All the taxa are treated equally irrespective to their age or dispersal ability (Humphries & Parenti 1999). 1.3.3 Approaches to analytical historical biogeography The area of phylogenetic studies has been developed with the advent of new sophisticated parametric methods and software packages such as BEAST (Drummond & Rambaut 2007) which are used to date phylogenies and infer biogeographic scenarios. Ancestral area reconstruction methods are used to infer the hypothetical ancestral areas at nodes of phylogenies under maximum parsimony, maximum likelihood or Bayesian criteria, and sometimes in conjunction with a model of geographic range evolution. Age estimates are determined using relaxed molecular 9 clocks and the distribution history is being inferred in terms of vicariance, long distance dispersal, speciation and extinction events within each clade (Crisp et al. 2011). Parsimony methods Perhaps the most basic ancestral area reconstruction method is the generalized parsimony approach, by accepting the fewest number of character state changes across the tree, where the transformation of character states are weighted equally (Lamm & Redelings 2009). Geography is coded as presence or absence for each taxon and which minimizes the instances of range evolution across phylogeny is considered as most parsimonious reconstruction. In complicated analyses where frequent dispersal and extinction occur with complex distributions, parsimony reconstructions can result in many equally parsimonious reconstructions making interpretation difficult (Lamm & Redelings 2009). Drawbacks of parsimony based approaches such as lack of incorporation of branch length information and the lack of a geographic model, lead to the search for novel approaches. Weighted ancestral area analysis (WAAA) In weighted ancestral area analysis a range of taxa and their ancestral areas are calculated and more weight is given to the distribution of basal or ancestral lineages (Lamm & Redelings 2009; Hausdorf 1998). Areas are optimized on to a tree and the number of times that area is gained and lost in the tree is calculated separately. Weighted gain steps and weighted loss step values are computed and a probability index is calculated from the ratio of above two values. The area with a high probability index is considered more likely to be a part of an ancestral area than the other. Branch length information is not considered and the results are entirely dependent on choosing which sister clade is deemed ancestral, approach now seen as spurious (Lamm & Redelings 2009). Dispersal-Vicariance analysis (DIVA) Parsimony, event based, quantitative methods such as DIVA (Ronquist 1997) allowed scientists to investigate ancestral area relationships based on dispersal and extinction events (Ronquist 1997; Buerki et al. 2011). It is considered as one of the first model based historical biogeographic methods. Species distribution areas are identified and 10 cost of range evolution is given depending on dispersal or extinction, with vicariance given a zero cost. Branch length is not taken in to account (Clark et al. 2008) and optimal reconstruction is produced by a parsimony criterion that minimizes dispersalextinction events (Ronquist 1997) and this can result in wider ancestral ranges (Kodandaramaiah 2010). Adding additional outgroups or restricting the number of possible ancestral ranges (Lamm & Redelings 2009) is used to overcome the problem, but this can result in discontinuous ancestral range reconstructions. Even in the absence of divergence times, general area relationships can be determined and prior knowledge of geological history of distributions are not needed. The rapid generation of results and user friendly software resulted in it being one of the most widely used methods in inferring ancestral area relationships. Bayes DIVA A weakness of DIVA was that a single tree topology was specified, with no account taken of other equally likely tree topologies. Bayes DIVA addresses phylogenetic uncertainty and improves upon DIVA by utilizing the posterior distribution of clades across multiple trees following Bayesian phylogenetic analysis. Ancestral area reconstructions are displayed as marginal distributions and the best reconstruction among all possible solutions is selected by integrating posterior distribution across all sampled trees (Nylander et al. 2008). This approach is implemented by the software package S-DIVA (Yu et al. 2010). Dispersal Extinction Cladogensis The parametric counterpart of DIVA, i.e. DEC model (Ree & Smith 2008) infers ancestral area reconstructions under a likelihood approach. It is a continuous time model, which treats local extinction and dispersal as stochastic processes and allows only single dispersal or extinction at any instant in time. This result in either range contraction or expansion (Ree & Sanmartin 2009; Lamm & Redelings 2009; Buerki et al. 2011) and for a given phylogeny, the likelihood of range inheritance of terminal taxa is calculated by integrating the rate of dispersal and extinction for each internal node. It proceeds through a continuous time Markov process giving the probabilities of each ancestral range as the result (Lamm & Redelings 2009; Ree et al. 2005). Unlike DIVA, 11 the DEC model explains the ancestral range inheritance by daughter lineages with vicariance and dispersal both treated as equally important (Kodandaramaiah 2010). However, a mechanism of identical range inheritance when the ancestral range comprises multiple unit areas is not addressed (Lamm & Redelings 2009). In contrast to DIVA, in DEC branch length information is incorporated and not weighting vicariance as the main factor in causing wide species ranges results in more realistic ancestral ranges. The ability for users to incorporate biological and physical parameters such as connections between areas at a given time and dispersal rates improves the flexibility and complexity (Clark et al. 2008; Kodandaramaiah 2010). DEC is implemented in two software packages, Lagrange (Ree & Smith 2008) and RASP (Yu et al. 2011). A geography matrix coded as binary presence-absence values and a phylogenetic tree with branch lengths proportional to expected changes are taken as input. Stochastic dispersal and local extinction along branches are optimised resulting in possible range inheritance scenarios at nodes (Ree & Sanmartín 2009; Webb & Ree 2012). Bayesian Binary MCMC (BBM) BBM implemented in RASP utilizes a full hierarchical Bayesian approach for inferring ancestral states of a phylogeny. The code from the Bayesian phylogenetic reconstruction programme MrBayes 3.1.2 (Ronquist & Huelsenbeck 2003) has been modified to perform geographic range reconstruction as a form of character evolution under one of four models (JC, JC+G, F81, F81+G) utilizing character state frequencies. Topological uncertainty and branch length are taken into account and the tree file and distribution file are taken as the input (Yu et al. 2011). 12 Figure 1.1 The processors assumed by different biogeographic analysis methods. Excerpted from Matzke 2013. Simulated Historical Island Biogeography (SHIBA) Simulated Historical Island Biogeography Analysis is a discrete temporal and spatial model based approach where a phylogeny is divided into time slicesand each time slice is integrated in a dynamic land area model (Webb & Ree 2012). Dispersal causes range expansion while extinction causes range contraction, with branching occurring within a time slice as a result of a speciation event. The ages of speciation events are incorporated into the land area model and probabilities of ancestral ranges are calculated (Webb & Ree 2012). SHIBA utilizes inputs as a distance matrix, a physical area matrix, a list of taxa, a chronogram and a presence absence matrix for taxa and has been used in few studies containing small number of taxa. Bay Area In the analysis where the number of area is large, a novel method, bay area is used where probablilities of biogeographic histories across the tree is calculated according to continuous time Markov Chain Monte Carlo process under Bayesian inference. 13 Resolution of geographic history is increased by increasing the number of areas which would reveals a better picture of events in earth’s history (Landis et al. 2013) Furthermore, use of grid system when defining geographic area would help to infer the biogeographic history of an area which lacks well defined boundaries. Recent advancement in biogeographic analysis allows likelihoods to be assigned to models of range evolution for a particular dataset. The package Biogeobears (Matzke 2014) allows choosing the best-fit model under a maximum likelihood framework by modeling transitions between discrete states (biogeographical ranges) along phylogenetic branches as a function of time. The DEC model calculates maximum likelihood ratios of ancestral states at speciation events, in a method similar to LAGRANGE (Ree et al. 2005). In addition to two free parameters; d (dispersal), e (extinction) included in the DEC model (Batalha-Filho et al. 2014; Landis et al. 2013) the new additional parameter J is added to model. The J parameter tests the contribution of founder event speciation to range expansion and controls the probability of two events during cladogenesis; founder event speciation versus sympatric and vicariant speciation (Matzke 2014). Different ancestral area reconstruction methods and characteristics are summarized in the table 1.1. Table 1.1 Characters of different ancestral area reconstruction methods. Name Optimality criterion Method Incorporates time no Software Cost based model Multiple trees no DIVA Parsimony approach Bayes DIVA Parsimony approach Cost based model yes no S-DIVA (Yu et al. 2010) DEC model SHIBA Likelihood approach No Yes no yes Bay area Bayesian probabilistic approach Continuous-time model Spatial-Temporal model Spatial-Temporal model no yes Lagrange (Ree & Smith 2008) SHIBA (Webb et al. 2012) Bay area (Landis et al. 2013) Probability approach 14 DIVA 1.1 (Ronquist 1997) 1.4 The current state of the field Ancestral area reconstruction methods are a rapidly evolving field, addressing the most important questions in biogeography such as when, where and how species evolve. There has been a recent proliferation of methods with the development of various model-based approaches. There are advantages as well as disadvantages of each method over the others but incorporation of geological and biological parameters to biogeographic models is beginning to reveal a better understanding about the geographic history of biodiversity. Due to the lack of testable hypotheses concerning long distance dispersal, and with the emergence of plate tectonic theory, vicariance was accepted as the major reason for disjunct plant distributions and long distance dispersal was considered as random noise thus taking less attention (Cowie & Holland 2006). However recent studies suggested that dispersal is a crucial factor in determining plant distributions due to the young ages of lineages which cannot be accepted as vicariance. For example, the highly endemic biodiversity in the Canary Islands provides no chances to integrate with a vicariance scenario since they never had continental connections, thus should be a result of long distance dispersal (Cowie & Holland 2006). Furthermore, there are wind current patterns, bird migratory routes, ocean currents in oceanic systems which support recurrent long distance dispersal patterns (Cowie & Holland 2006; Nathan 2006). 1.5 Summary Biogeography has been often considered as a narrative addition to phylogenetic studies. However with the advent of new sophisticated models and accumulation of data from various other fields such as ecology, geology, molecular dating has provided opportunities to test biogeographic hypothesis and then to infer biogeographic scenarios. The selection of data in hypothesis testing is fundamental, and should be independent such as fossil data rather than geological data which could be subjective and biased (Crisp et al. 2011). More often vicariance hypotheses are testable due to the evidence provided, however more careful examination of independent evidences provide space to test dispersal hypotheses. 15 Selection of an appropriate model for hypothesis testing, accurate data and adequate taxon sampling are crucial factors in any analysis in order to obtain a more reliable picture of historical biogeographic scenarios (Webb & Ree 2012). 16 CHAPTER 2: Introduction to Sri Lankan biogeography 2.1 Current Geology and climate of Sri Lanka The island of Sri Lanka is located south-east of the southernmost tip of peninsular India, with a total area of 65,610km2 and known as Ceylon until 1972. It has a close proximity to India, separated by the shallow Palk Strait which is 32 km wide. It is positioned in the Indian Ocean, to the southwest of the Bay of Bengal, between latitude 5° and 10°N, and longitude 79° and 82°E. Geologically Sri Lanka is overlain with younger rocks in places of metamorphic rocks of Precambrian origin. These rocks form three major lithotectonic units named the Vijayan Complex, the Highland Complex and the Vanni Complex. The Highland Complex is the largest, forming the backbone and covering 90% of the island. It is composed of supra crustal rocks, charnockites, enderbites, metabasits, metagabbro, metadiorites and some orthogneisses and is bounded by the Vanni Complex to the west and by the Vijayan Complex to the east. The origins of the formation of the different geological units dates back to the Precambrian. Jurassic formations include shales, carbonaceous shales and arkosic sandstone that are 180 million years old and are present in a small area of the Western coast. Miocene limestones underlie the northwestern part of the country and extend south in a relatively narrow belt along the west coast. The island also contains laterites, gravels, red earths and limestone of Pleistocene origin, whereas the Holocene formations include alluvial deposits, coastal sandstone, beach mineral sands, peat and gem gravels (Herath 1984). The narrow limestone shoal is found beneath the Palk Strait, also locally known as Rama’s or Adam’s Bridge based on the legend that it was built by Rama to rescue his consort Sitha, as mentioned in the Ramayana. The presence of sand belts beneath the coral have been used to controversially support that belief since corals usually form on rocky substrates. This reflects the complex geological history of the reefs, and there is some evidence to suggest they result from local uplift and this suggests that these land masses were in full connection until the last sea level rise 6000 years ago (Ashton& Gunatilleke 1987; McLoughlin 2001). 17 A recent study by Dissanayake et al. (2000) suggested closer geological affinities of Sri Lanka with Madagascar than with the Southern and Eastern parts of India, due to the presence of high grade basement rocks in Sri Lanka that are more similar to those of South-eastern Madagascar. Furthermore, evidence of a distinct mineral belt running from Antarctica through the highland series of Sri Lanka into Madagascar, Tanzania, and further north suggests close similarities between these lands dating back to Gondwanan times. Corundum, spinel, topaz, zircon, aquamarine, amethys are some common gems found in Sri Lanka and Madagascar and the presence of gems in eastern Antarctica confirms past links that are further supported by graphite formations which are common to all these land masses (Dissanayake & Chandrajith 1999). Figure 2.1 Geographic regions in Sri Lanka The current climate of Sri Lanka can be described as tropical and warm. The average annual temperature for the country as a whole ranges from 28 to 30 °C. Diurnal 18 temperatures may vary by 4 to 7 °C. January is the coolest month, with the temperature dropping to 5°C (in the highlands) with May being the hottest month. The island receives rain from a South-western monsoon from May to September and a North-eastern monsoon from December to February. South-western parts and windward slopes of the central highlands receive high rains during south western monsoons of around 2500mm and are classified as the wet zone. The North-eastern side receives high rains during December to February but the amount is not as much as in the Southwestern monsoons. Those areas that receive 1250mm-1750mm rainfall are known as the dry zone; this includes most of the North and North East of the island. However, North Western and South Eastern parts receive the lowest annual rain fall of less than 1250mm. An intermonsoonal period occurs from March until mid-May, with light, variable winds and evening thunder showers. 2.2 Palaeoclimate, geological changes According to plate tectonics, the huge land mass of Gondwana began to breakup during the late Jurassic 180 million years ago (Figure 2.2). The Indian (or Deccan) Plate, composed of both India and Sri Lanka, rifted from East Africa 158-160 mya (Briggs 2003; Conti et al. 2002; Ashton & Gunatilleke 1987) along with the Mascarene plate comprised of Madagascar and the Seychelles Plateau. This block moved northward in close proximity to Africa allowing exchange of floristic elements between landmasses (Gunatilleke &Gunatilleke 1984; Ashton & Gunatilleke 1987).The split between the Madagascar and Indian plus Seychelles plates occurred about 95-85 million years ago (Ashton & Gunatilleke 1987; Conti et al. 2002). About 65 million years ago the Seychelles Plateau separated from the Deccan Plate and remained fixed relative to Africa. Since then, the Deccan Plate drifted northward close to the African plate throughout the Cretaceous, with a 35° anti-clockwise rotation in the early Tertiary. The Deccan Plate moved north at about 20 centimetres (7.9 in) per year approaching equatorial latitudes during the Eocene(45 million years ago) and began colliding with Eurasia 55-65 Ma followed by a hard collision with the southern shore of Eurasia approximately 42-55 ma during the early Tertiary (Briggs 2003; Conti et al. 2002; Gunatillke & Gunatilleke 1984). It is believed that an aseasonal humid tropical 19 climate persisted on the Deccan Plate during its migration; this is supported by the presence of moribund laterites in the South-western area of the country, which can be formed only under wet seasonal climates (Ashton & Gunatilleke 1987). However, there is also evidence that the plate would have experienced aridification (Rutschmann & Eriksson 2004) as it moved north through 30-35 horse latitudess where patterns of atmospheric circulation result in very little precipitation and the formation of the desert belt (Lomolinoet al. 1998). Figure 2.2 Break-up of Gondwana showing vicariance and collision times between continental fragments. Excerpted from McLoughlin 2001. 2.3 Flora of Sri Lanka Sri Lanka has high floristic richness and species diversity, comprises 3900 species of land plants with the endemic elements heavily concentrated in the wet south-western quarter of the island. There are 2900 species of angiosperms on the island of which 830 are known to be endemic (Ashton & Gunatilleke 1987). Despite the fact that the flora of Sri Lanka has been extensively studied since the eighteenth century by various botanical experts, the biogeographic affinities of the flora remain poorly investigated (Ashton & Gunatilleke 1987). 20 Within the country eleven genera are known to be endemic, all of them showing restricted distributions in the wet zone including Stemonoporus (Dipterocarpaceae) 26 spp.and Sonerila (Melastomataceae) 22spp. An example of a more widespread genus that has high species endemism in Sri Lanka is Memecylon (Memecylaceae) which contains about 14 species distributed in lowland wet zone,eight species in the wet montane zone and eight species in the dry intermediate zone. Thirty of its species occur in Sri Lanka, of which 28 are considered to be endemics (Ashton & Gunatilleke 1987). The forests of Sri Lanka can be classified into nine forest types: 1. Montane forest 2. Submontane forest 3. Lowland rain forest 4. Moist monsoon forest 5. Dry monsoon forest 6. Riverine dry forest 7. Mangroves 8. Sparse forest 9. Forest plantations (forest survey) Among the forests distributed within the island, the lowland rain forest in the south western part of the country is considered very important, harbouring 90% of the endemics and high species diversity. The south western hills of Sri Lanka include most of the lowland and lower montane forests with biological significance as Sinharaja, Kanneliya, Nakiyadeniya, Dediyagala, Hiniduma etc. and the area is considered as the only aseasonal wet region in South Asia (Ashton & Gunatilleke 1987; Gunatilleke et al. 2005). 21 These forests are similar in floristic composition to the moist rain forests of the Western Ghats of Peninsular India suggesting past close affinities between the two land masses. These two may have acted as refugia for relicts of the Indian plate flora during changing climatic conditions. Due to their high endemicity and species diversity they have been declared as one of 25 biodiversity hotspots (Gunatilleke et al. 2005; Conti et al. 2002). Several species are endemic to both the Agastyamalai-Nilgiri Hills of The Western Ghats and the Sri Lankan highlands, including Abarema subcoriacea, Biophytum nudum, Chrysoglossum maculatum, Eugenia rotundata, Fahrenheitia zeylanica, Filicium decipens, Pavetta zeylanica, and Rubus micropetalus.This high endemicity might be a result of the isolated migration of the Indian plate in the time between the Paleocene and Miocene epochs. This time period could have been sufficient for the evolution of new genera (Ashton & Gunatilleke 1987). Another interesting feature is the absence of some Gondwanan plant and animal species in the Indian Peninsula despite their presence in Sri Lanka and Southeast Asia. Rapid latitudinal changes during the Late Cretaceous to Early Tertiary times caused climatic changes, resulting in the massive extinction of Gondwanan forms from India, but as an isolated island Sri Lanka was not affected as much (Rutschmann et al. 2004). Extensive volcanic eruptions 65 million years ago further impoverished the Indian biota and Sri Lanka may have acted as refuge harbouring Gondwanan Biota mainly in the Southwestern parts of the island (Gunatilleke et al. 2005; Conti et al. 2002). One striking example of this is the presence of Gondwanan Crypteroniaceae such as Axinandra zeylanica endemic to Sri Lanka with a large number of con-generic species in Southeast Asia but none in India. This could be explained through extinction of the genus from India during the Tertiary (Conti et al. 2002; Rutschmann et al. 2004) according to phylogenetic evidence and molecular dating. Thus, Sri Lanka may have retained many of the Gondwanan elements that India lost (Karanth 2006). There are not many angiosperm fossils in Sri Lanka, but fossils from India provide strong indications of past climatic history of these two landmasses. Late Cretaceous to early Tertiary fossils from India includes araucariaceous and podocarpaceous conifers and monocot families like Strelitziaceae, Zingiberaceae and Cyclanthaceae. Lakhanpal 22 (1970) identified 29 dicotyledon Cretaceous-early Tertiary fossils belonging to families including Clusiaceae, Tiliaceae, Combretaceae, Sonneratiaceae, and Anacardiaceae. Eocene deposits include macrofossils of Clusiaceae and Bombacaceae. This fossil content is similar to that of the Oligocene indicating the existence of a similar flora during those two time periods. Interestingly, lower Tertiary fossils include taxa from Juglandaceae and Myricaceae which might be a result of immigration from Laurasia (Lakhanpal 1970). There are a few Oligocene deposits and fossils of Ebenaceae, Dipterocarpaceae, Ericaceae, Malvaceae etc. of the Miocene from Rajhastan supporting the presence of humid tropical conditions at that time in these land masses (Gunatilleke & Ashton 1987; Lakhanpal 1970). Based on the fossil record Morley (2003) indicated that a substantial number of Indian sub-continental elements invaded Southeast Asia during the Oligocene when the Laurasian part of Southeast Asia and the Indian subcontinent were at similar latitudes and had similar climates. The distinct distribution of the genus Cotylelobium of Dipterocarpaceae in Sri Lanka and West Malesia argues for immigration from mainland Asia to Sri Lanka, or the reverse; the Indian subcontinent is therefore thought to be important in contributing to the biogeographic history of Southeast Asia (Ashton & Gunatilleke 1987). 2.4 Possible origins of Sri Lankan Flora The Sri Lankan flora comprises elements that could have reached the island in several contrasting ways. Firstly, as Sri Lanka and India originated as fragments of the southern supercontinent Gondwana, together with South America, Africa, Antarctica, Madagscar, New Guinea, Australia, New Zealand and New Caledonia. The breakup of the supercontinent Gondwana initiated during early Jurassic 180 million years ago (Mya), followed by subsequent continental drifts and sea floor spreading. The Deccan plate comprised of India and Sri Lanka was a part of Gondwana that initiated separation from Antarctica from 132 Mya followed by northward migration until collision with Asia c. 35 Mya. Being a part of Gondwana, part of Sri Lankan flora might be Gondwanan relicts. 23 However to be of Gondwana origin the disjunction should be reflected in phylogenetic splits occurring approximately 132 Mya-94 Mya concurrent with continental break-up. One good example is family Monimiaceae, which is represented by two species of Hortonia from Sri Lanka with one genus from West Africa and abundance in South America and Australia (Renner et al. 2010). The chronogram in Figure 2.3 depicts times of lineage splits which would be concordant with Gondwanan vicariance. Secondly, species could have arrived via the immigration of Laurasian lineages through Asia and India after the collision of the Deccan plate with Laurasia from c. 35 Mya. Thus, the species could have arrived via the immigration of Laurasian lineages through Asia and India which resulted in the mixing of the Deccan Gondwanan flora with the tropical flora occupying southern Laurasia. The collision of Deccan plate with southern Laurasia is accepted as the most marked possibility for interplate dispersal during the Tertiary (Morley 2000). However for this hypothesis to be accepted, we would expect that divergence times for lineages should not be older than 35 Mya and that Indo-Sri Lankan lineages would be nested within Laurasian ones. Thirdly, long distance dispersal could have played a major role in the assemblage of the Sri Lanka flora from other land masses such as Africa and/or South East Asia. This could have occurred at any point in time, it is the only viable scenario for tropical disjunctions younger than ~33 Mya and Indo-Sri Lankan lineages would be nested within African and/or South East Asian ones. Long distance dispersal has contributed to the sharing of floristic elements between tropical Asia and Africa, e.g. the appearance of Begonia afromigrata in Indochina which arrived from Africa during the Pliocene (de Wilde et al. 2011). There are other striking examples such as the occurrence of Gleditsia in South America with its closest relatives living in China. Also, long distance dispersal has been used to explain the colonization of some lineages within pantropical families such as Annonaceae, Myristicacae and Boraginaceae to their current distributions (Renner 2005). 24 Figure 2.3 Geological area cladogram adapted from Sanmartin & Ronquist 2004 representing relationships between Gondwanan fragments and time of phylogenetic splits based on paleogeographic data. Vicariance is assumed to be at primary fragmentation. Asteriks (*) mark nodes, which are dated at 70-60 Ma in alternative reconstructions. Another potentially important contributor to the Sri Lankan flora will have been diversification of lineages within the island (in situ speciation), resulting in the evolution of new species. Species that are endemic to Sri Lanka, and also have sister species in Sri Lanka, are likely to have arisen by in situ speciation, whereas endemics whose closest relatives occur elsewhere might be paleoendemics that once had wider distributions. 2.5 Study groups Sri Lanka is a key location for understanding patterns of migration and biome assembly among tropical plants. To investigate its biogeographic affinities, study groups should be pantropical in distribution, well-sampled and thoroughly investigated in other regions. Sapotaceae, Begoniaceae and Zingiberaceae are ideal in this respect since all have dated phylogenies produced by previous studies (Bartish et al. 2011, Thomas 2012, Poulsen et al. unpublished) that are somewhat lacking in representatives from Sri Lanka. 2.5.1 FamilyBegoniaceae Begoniaceae are a large pantropical family widely distributed in the tropics with greatest species diversity in Southeast Asia and South America. The largest genus 25 Begonia dominates the family with ca. 1800 species including 900 spp. in Asia with the bulk occurring in Southeast Asia (Hughes & Hollingsworth 2008, Rajbhandary et al. 2011). The genus Begonia has ten species present in Sri Lanka, of which two are known to be endemic (Dassanayake 1983), four are native and four are introduced. Within Sri Lanka, the two endemics are restricted to humid lowlands while the other native species are distributed in the hilly wet and intermediate zones. Begonia has migrated from an origin in Africa during the Oligocene or late Eocene to China and Southeast Asia across India or dispersed via Asia (Thomas et al. 2012; Rajbhandary et al. 2011). The Sri Lankan species Begonia malabarica is sister to two species from the Socotran archipelago and with them forms the western limit of the Asian Begonia clade (Rajbhandary et al. 2011). It is suggested that continental Asia is the ancestral area for the Socotran Begonia, which resulted from long distance dispersal and diversifications of South Indian-Sri Lankan lineages to the Arabian-Socotran region during the Late Miocene. Thus, it would be interesting to determine the migration history of Sri Lankan Begonia species and their role in the diversification of the genus to the current hotspots with complete taxon sampling with the DNA regions used by Thomas et al. (2012). 2.5.2 Family Sapotaceae Sapotaceae are a pantropical family of 53 genera and about 1100 species, mostly distributed in the tropical forests, but some extending to semi-arid and arid regions. The highest species richness is recorded in the tropical and subtropical regions of Asia and South America (Swenson & Anderberg 2005). In Sri Lanka, seven genera are native and all the nine species of Palaquium, three of Isonandra and four Madhuca are endemic (Dassanayake 1995). According to dated phylogenies, long distance transoceanic dispersal events from South America or Australasia across Atlantic or Indian oceans must have played a role in the diversification of Sapotaceae, rather than vicariance events such as the break-up of Gondwana (Bartish et al. 2011; Amstrong et al. 2014). Sri Lanka might have played an important role as a stepping stone in the movement of Sapotaceae between continents. 26 2.5.3FamilyZingiberaceae Zingiberaceae, the largest family of Zingiberales, are pantropically distributed with a diverse set of genera in Africa, Southeast Asia and Pacific with only one neotropical genus Renealmia (Kress et al. 2005; Kress & Specht 2006). There are twelve genera found in Sri Lanka. Out of 36 species, 12 are known to be endemic although the status of some of the species is doubtful (Dassanayake 1983). The family is widely distributed in wet lowland and mid-montane primary forests but some of the endemic species are rare and have not been re-collected since the type gathering. In addition to understand the affinity of the Sri Lankan Zingiberaceae with those of other regions, molecular data could be used to investigate species limits within the family, and hence examine whether in situ speciation has occurred within Sri Lankan Zingiberaceae. All species present in Sri Lanka belonging to each of the above families, their ecological status and distribution are summarized in the table 2.1. Table 2.1 The species present in Sri Lanka for each study group, ecological status and their distribution Ecological Family Species Begoniaceae Begonia tenera Endemic Sri Lanka Begonia thwaitesii Endemic Sri Lanka Begonia cordifolia Native Sri Lanka, India Begonia dipetala Native Sri Lanka, India Begonia malabarica Native Sri Lanka, India Palaquium pauciflorum Endemic Sri Lanka Palaquium thwaitesii Endemic Sri Lanka Palaquium grande Endemic Sri Lanka Palaquium hinmolpedda Endemic Sri Lanka Palaquium zeylanicum Endemic Sri Lanka Palaquium rubiginosum Endemic Sri Lanka Palaquium petiolare Endemic Sri Lanka Palaquium laevifolium Endemic Sri Lanka Palaquium canaliculatum Endemic Sri Lanka Isonandra lanceolata Native Sri Lanka, India, Brunei, Borneo Isonandra compta Endemic Sri Lanka Sapotaceae status 27 Distribution Zingiberaceae Isonandra montana Endemic Sri Lanka Isonandra zeylanica Endemic Sri Lanka Madhuca fulva Endemic Sri Lanka Madhuca moonii Endemic Sri Lanka Madhuca microphylla Endemic Sri Lanka Madhuca clavata Endemic Sri Lanka Madhuca neriifolia Endemic Sri Lanka Madhuca longifolia Native Sri Lanka, India Madhuca indica Native Sri Lanka, India Manilkara hexandra Native Sri Lanka, Thailand, Himalaya Mimusops elengi Native Sri Lanka, Malaysia. India, Thailand Xantolis tomentosa Native Sri Lanka, India, Burma Zingiber cylindricum Endemic Sri Lanka Zingiber wightianum Native Sri Lanka, India Amomum nemorale Endemic Sri Lanka Amomum trichostachyum Endemic Sri Lanka Amomum graminifolium Endemic Sri Lanka Amomum echinocarpum Endemic Sri Lanka Amomum hypoleucum Endemic Sri Lanka Amomum benthamianum Endemic Sri Lanka Amomum acuminatum Endemic Sri Lanka Amomum fulviceps Native Sri Lanka, India Amomum masticatorium Native Sri Lanka, India Amomum pterocarpum Native Sri Lanka, India Curcuma albiflora Endemic Sri Lanka Curcuma aromatica Native Sri Lanka, India Curcuma zedoaria Native Sri Lanka, India Curcuma oligantha Native Sri Lanka, India Alpinia fax Endemic Sri Lanka Alpinia rufecense Endemic Sri Lanka Alpinia abundiflora Native Sri Lanka, India Alpinia nigra Native Sri Lanka, Indo-Malaysia Alpinia malaccensis Native Sri Lanka, Malaysia Elettaria cardamomum Native Sri Lanka, India Hedichyum coronarium Native Sri Lanka, India, Malaysia 28 2.6 Distribution of Begoniaceae, Sapotaceae and Zingiberaceae in Sri Lanka Figure 2.4 Species distribution of Begonia in Sri Lanka Figure 2.5 Distribution of Palaquium species in Sri Lanka 29 Figure 2.6 Distribution of Isonandra species in Sri Lanka Figure 2.7 Distribution of Madhuca species in Sri Lanka 30 Figure 2.8 Distribution of Mimusops species in Sri Lanka Figure 2.9 Distribution of Manilkara hexandra in Sri Lanka 31 Figure 2.10 Distribution of Xantolis tomentosa in Sri Lanka Figure 2.11 Distribution of Alpnia fax, Alpinia abundiflora and Cyphostigma pulchellum species in Sri Lanka 32 Figure 2.12 Distribution of Amomum species in Sri Lanka Figure 2.13 Distribution of Curcuma and Zingiber species in Sri Lanka 33 2.7 Field collections in Sri Lanka Field collections were made at pre identified locations (list of locations is given below) defined based on the distribution of the species according to the flora of Ceylon and existing herbarium specimens in different herbaria. Transportation was made by a vehicle to the forests and inside the forests on foot. Field assistants were accompanied on all the excursions. For the DNA extractions leaf samples of each of the species were collected in silica. To prepare herbarium specimens a branch from each species was removed from the plant using secateurs and wrapped in newspaper. Then alcohol was added to prevent degradation of the specimens. All the information on the site and plant were recorded and the locations were recorded using a GPS. Plant habit, reproductive structures and any other important features were photographed. When collecting branches from trees tree climbers or poles were used. All the local connections were made to get the collection permits and specimens were taken through proper channels to the Royal Botanic Gardens of Edinburgh, where I carried out my lab work. Silica dried specimens were used for DNA extractions and herbarium specimens were prepared and deposited at the Royal Botanic Gardens Edinburgh. Figure 2.14 Localities of field collections in Sri Lanka 34 Table 2.2 Species collected in each family during the field work in Sri Lanka. Endemic species are indicated in bold. Family Genera Species present Species collected % sampling Begoniaceae Begonia Begonia tenera Begonia thwaitesii Begonia cordifolia Begonia dipetala Begonia malabarica Endemic-100% Native- 75% Sapotaceae Palaquium Begonia tenera Begonia thwaitesii Begonia cordifolia Begonia dipetala Begonia malabarica Begonia subpeltata Palaquium pauciflorum Palaquium thwaitesii Palaquium grande Palaquium hinmolpedda Palaquium zeylanicum Palaquium rubiginosum Palaquium petiolare Palaquium laevifolium Palaquium canaliculatum Isonandra lanceolata Isonandra compta Isonandra montana Isonandra zeylanica Madhuca fulva Madhuca moonii Madhuca microphylla Madhuca clavata Madhuca neriifolia Madhuca longifolia Manilkara hexandra Mimusops elengi Zingiber cylindricum Zingiber wightianum Curcuma albiflora Curcuma aromatica Curcuma zedoaria Curcuma oligantha Amomum nemorale Amomum graminifolium Amomum acuminatum Amomum echinocarpum Amomum fulviceps Amomum masticatorium Amomum pterocarpum Alpinia fax Alpinia abundiflora Alpinia rufescence Alpinia nigra Alpinia malaccensis Elettaria cardamomum Hedichyum coronarium Hedichyum flavecens Cyphostigma pulchellum Palaquium pauciflorum Palaquium thwaitesii Palaquium grande Palaquium hinmolpedda Palaquium zeylanicum Palaquium rubiginosum Palaquium petiolare Palaquium laevifolium Palaquium canaliculatum Isonandra lanceolata Isonandra compta Isonandra montana Isonandra zeylanica Madhuca fulva Madhuca clavata Madhuca neriifolia Madhuca longifolia Endemic- 100% Manilkara hexandra Mimusops elengi Zingiber cylindricum Zingiber wightianum Curcuma albiflora Curcuma zedoaria Curcuma oligantha Native- 100% Native – 100% Endemic- 100% Native - 100% Endemic- 100% Native – 66% Amomum nemorale Amomum graminifolium Amomum acuminatum Amomum fulviceps Amomum masticatorium Amomum pterocarpum Endemic- 75% Native - 100% Alpinia fax Alpinia abundiflora Alpinia nigra Alpinia malaccensis Endemic - 66% Native - 100% Elettaria cardamomum Hedichyum coronarium Native 100% Native 50% Cyphostigma pulchellum Endemic- 100% Isonandra Madhuca Zingiberacea e Manilkara Mimusops Zingiber Curcuma Ammomum Alpinia Elettaria Hedichyum Cyphostigma 35 Endemic- 100% Native- 100% Endemic- 60% Native- 100% 2.8 Species portrait: Rediscovery of Amomum nemorale (endemic) Many of the species collected for this project were found after not being seen for many decades. Some, such as Begona thwaitesii, were photographed for the first time. In particular, I highlight the rediscovery of Amomum nemorale, which was previously only known from type collection. Amomum nemorale (Thw.) Trimen, Cat. 92. 1885; Trimen, Handb. Fl. Ceylon 4: 251. 1898; Baker in Hook. F., Fl. Br. Ind. 6: 233. 1892. Type: C.P. 3703 (PDA, K, BM) Elettaria nemoralisThw., Enum. Pl. Zeyl. 319. 1861. Figure 2.15 Habit (left) and inflorescence (right) of Amomum nemorale. “Leafy stem 60-120 cm; lamina shortly petiolate (c. 1 cm), 14-30 (?)x 3-5 cm, narrowly lanceolate, acuminate, glabrous, the margins callose-denticulate towards the tip; ligule short, rounded, entire, ciliate. Peduncle 5-7 cm, clothed with broad, loose scales. Inflorescence 2.5-3.5 cm long, subglobose to oblong. Bracts 0.7-1.5 cm long, orbicular or obovate, membranous, lightly pubescent towards the tip. Bracteoles lanceolate. Calyx 1.7 cm long, glabrous. Corolla greenish-white, tube 3.5 cm long; lobes up to 1.2 cm long, oblong, obtuse, subequal. Labellum veined with pink, reniform, 3- lobed, lateral lobes falcately recurved, modlobe smaller and rounded, 2-3 fid, Lateral staminodes? Anther c. 4 m, thecae glabrous, connective not prolonged into a crest. Ovary 36 pubescent.Capsule smooth, subglobose, c. 1.75 cm in diameter; seeds few, large, enclosed in a white spongy pulp” (Flora of Ceylon volume 4) 2.9 Aims of the doctoral research and structure of the thesis Dated phylogenies are an important tool in historical biogeography because they provide a method for testing amongst alternative biogeographic hypotheses. The objectives of this study are, 1. To incorporate Sri Lankan taxa into worldwide phylogenies for three families: Sapotaceae, Begoniaceae and Zingiberaceae. 2. To estimate the relative contributions to the Sri Lankan flora of Gondwanan relicts, immigration from nearby landmasses (Asia), long distance dispersal from Southeast Asia, and long distance dispersal from Africa. 3. To examine the contribution of in-situ speciation to biodiversity in Sri Lanka. 4. To investigate the contribution of Sri Lanka to the assembly of biotas elsewhere, by examining whether examples exist of: a. clades of Asian taxa whose most early diverging members occur in Sri Lanka, indicating a group whose ancestors rafted northwards with the subcontinent and then radiated outwards into Asia. b. long distance dispersal events out of Sri Lanka, as might be indicated for example by a clade of Southeast Asian species whose closest relative was Sri Lankan and whose next closest relatives were also Sri Lankan or occurred further west. The thesis contains four empirical research chapters; The biogeography of Sri Lankan Begoniaceae (Chapter 3), The biogeography of Sri Lankan Sapotaceae (Chapter 4), The biogeography of Sri Lankan Zingiberaceae (Chapter 5) and the final chapter (Chapter 6) discussing main findings and conclusions. Each chapter is aimed at a scientific publication, thus there could be some content repetition. 37 References Armstrong, K.E. et al., 2014. Patterns of diversification amongst tropical regions compared: a case study in Sapotaceae. Frontiers in Genetics, 5(December), pp.1– 13. Ashton, P.S. & Gunatilleke, C.V.S., 1987. New light on the plant geography of Ceylon I Historical plant geography. Journal of Biogeography, 14(3), pp.249–285. Baldwin, B.G., 1992. Phylogenetic Utility of the Internal Transcribed Spacers of Nuclear Ribosomal DNA in Plants: An Example from the Compositae. Molecular Phylogenetics and Evolution, 1(1), pp. 3-16. Batalha-Filho, H. et al., 2014. Phylogeny and historical biogeography of gnateaters (Passeriformes, Conopophagidae) in the South America forests. Molecular Phylogenetics and Evolution, 79, pp.422–432. Bartish, I. V. et al., 2011. Vicariance or long-distance dispersal: Historical biogeography of the pantropical subfamily Chrysophylloideae (Sapotaceae). Journal of Biogeography, 38(1), pp.177–190. Borsch, T. & Quandt, D., 2009. Mutational dynamics and phylogenetic utility of noncoding chloroplast DNA. Plant Systematics and Evolution, 282, pp.169–199. Briggs, J.C., Collections, Z. & Museum, G., 2003. The biogeographic and tectonic history of India. Journal of Biogeography, 30, pp.381–388. Buffon, G. L. L. & Compte De, 1761. Histoire naturelle generale. Paris: Imprimerie Royal. Buffon, G. L. L. & Compte De, 1776. Servant de suite a l’histoire des animaux Quadrupedes (Supplement III) In: Histoire naturelle generale et particuliere. Paris: Imprimerie Royal. 38 Buerki S., et al,. 2011. An evaluation of new parsimony-based versus parametric inference methods in biogeography: a case study using the globally distributed plant family Sapindaceae. J Biogeog, 38, pp.531–550. Clark, J.R. et al., 2008. A comparative study in ancestral range reconstruction methods: retracing the uncertain histories of insular lineages. Systematic biology, 57(5), pp.693–707. Cowie, R.H. & Holland, B.S., 2006. Dispersal is fundamental to biogeography and the evolution of biodiversity on oceanic islands. Journal of Biogeography, 33, pp.193– 198. Conti, E. et al., 2002. Early Tertiary out-of-India dispersal of Crypteroniaceae: evidence from phylogeny and molecular dating. Evolution; international journal of organic evolution, 56(10), pp.1931–1942. Croizat, L., 1952. Manual of Phytogeography. The Hague: Dr. W. Junk. Darwin, C., 1859. On the Origin of Species by Means of Natural Selection, or the Preservation of Favoured Races in the Struggle for Life. London: John Murray. De Candolle, A. P., 1820. Essai elementaire de geographie botanique. De l’imprimerie de F.G. Levrault. De Candolle, A. P., 1855. Geographie Botanique Raisonee (2 volumes). Paris: Masson. De Wilde, J.J.F.E. et al., 2011. Pliocene intercontinental dispersal from africa to Southeast Asia highlighted by the new species begonia afromigrata (Begoniaceae). Taxon, 60(December), pp.1685–1692. Dissanayake, C.B. & Chandrajith, R., 1999. Sri Lanka–Madagascar Gondwana Linkage: Evidence for a Pan‐African Mineral Belt. The Journal of Geology, 107(2), pp.223–235. 39 Dissanayake, C.B., Chandrajith, R. & Tobschall, H.J., 2000. The geology, mineralogy and rare element geochemistry of the gem deposits of Sri Lanka. Bulletin of the Geological Society of Finland, 72(1-2), pp.5–20. Donoghue, M.J. & Moore, B.R., 2003. Toward an integrative historical biogeography. Integrative and comparative biology, 43, pp.261–270. Drummond, A.J. et al., 2006. Relaxed phylogenetics and dating with confidence. PLoS Biology, 4(5), pp.699–710. Drummond, A.J. et al., 2012. Bayesian phylogenetics with BEAUti and the BEAST 1.7. Molecular Biology and Evolution, 29, pp.1969–1973. Drummond, A.J. & Rambaut, A., 2007. BEAST: Bayesian evolutionary analysis by sampling trees. BMC evolutionary biology, 7, p.214. Forster, J.R., 1996. Observations Made During A Voyage Round the World. University of Hawai'i Press. Givnish, T.J. & Renner, S.S., 2004. Tropical Intercontinental Disjunctions: Gondwana Breakup, Immigration from the Boreotropics, and Transoceanic Dispersal. International Journal of Plant Sciences, 165, pp.S1–S6. Guindon, S. & Gascuel, O., 2003. A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Systematic biology, 52, pp.696–704. Gunatilleke, I. A. U.N. & Gunatilleke, C.V.S., 1984. Distribution of endemics in the tree flora of a lowland hill forest in Sri Lanka. Biological Conservation, 28(3), pp.275–285. Gunatilleke, I., Gunatilleke, C. & Dilhan, M., 2005. Plant biogeography and conservation of the south-western hill forests of Sri Lanka. Raffles Bulletin of Zoology Suppl, 12(12), pp.9–22. Hausdorf, B., 1998.Weighted ancestral area analysis and a solution of the redundant distribution problem. Systematic Biology, 47, pp. 445–456. 40 Hennig, W., 1966. Phylogenetic Systematics. 3rd edition. Translated by D.D. Davis and R. Zanderl. Urbana: University of Illinois Press. Herath, W., 1984. Geology and Occurrence of Gems in Sri Lanka *. J. Natn. Sci. Coun. Sri Lanka, 12, pp.257–271. Hughes, M. & Hollingsworth, P.M., 2008. Population genetic divergence corresponds with species-level biodiversity patterns in the large genus Begonia. Molecular Ecology, 17(11), pp.2643–2651. Humphries, C.J. & Parenti, L., 1999. Cladistic Biogeography: Interpreting Patterns of Plant and Animal Distribution. New York: Oxford Univ. Press. Jayasuriya, A.H.M., 1983. Begoniaceae. In: Dassanayake, M.D., Fosberg, F.R. (Eds.), Flora of Ceylon. Amerind Publishing, New Delhi, pp. 137–152. Kim, S.C. et al., 1999. The use of a non-coding region of chloroplast DNA in phylogenetic studies of the subtribeSonchinae (Asteraceae:Lactuceae). Plant Systematics and Evolution, 215, pp.85–99. Kodandaramaiah, U., 2010. Use of dispersal-vicariance analysis in biogeography - A critique. Journal of Biogeography, 37, pp.3–11. Kress, J.W. & Specht, C.D., 2006. The evolutionary and biogeographic origin and diversification of the tropical monocot Oder Zingiberales. Aliso, 22, pp.621–632. Kress, W.J. et al., 2005. The molecular phylogeny of Alpinia (Zingiberaceae): A complex and polyphyletic genus of gingers. American Journal of Botany, 92(1), pp.167–178. Lakhanpal, R. N., 1970. Tertiary floras of India and their bearing on the historical geology of the region. Taxon, 19(5), pp.675-694. Lamm, K.S. & Redelings, B.D., 2009. Reconstructing ancestral ranges in historical biogeography: Properties and prospects. Journal of Systematics and Evolution, 47(June), pp.369–382. 41 Landis, M.J. et al., 2013. Bayesian analysis of biogeography when the number of areas is large. Systematic Biology, 62(0), pp.789–804. Lomolino, M.V., Riddle, B.R. & Brown, J.H., 1998. Biogeography, 3rd edn. Sinauer Associates, Sunderland, MA. Lyell, C., 1830. Principles of Geology, Being an attempt to explain the former changes of the earth’s surface, by reference to causes now in operation. London: John Murray. Lyte, C., 1981. Sir Joseph Banks. 18th Cemtury Explorer, Botanist & Entrepreneur. Med Hist, 25(3), pp.329-330. Matzke, N.J., 2013. Probabilistic historical biogeography: new models for founder‐ event speciation, imperfect detection, and fossils allow im‐proved accuracy and model‐testing. Frontiers of biogeography, 5(4), pp.242-248. Matzke, N.J., 2014. Model Selection in Historical Biogeography Reveals that FounderEvent Speciation Is a Crucial Process in Island Clades. Systematic Biology, 63(6), pp.951–970. McLoughlin, S., 2001. The breakup history of Gondwana and its impact on preCenozoic floristic provincialism. Australian Journal of Botany, 49(3), pp.271–300. Morley, R. J., 2000. Origin and evolution of tropical rain forests. Chichester: John Wiley & Sons, Ltd. Morley, R.J., 2003. Interplate dispersal paths for megathermal angiosperms. Perspectives in Plant Ecology, Evolution and Systematics, 6(12), pp.5–20. Morrone, J.J. & Crisci, J.V., 1995. Historical biogeography: introduction to methods. Annual Review of Ecology and Systematics, 26, pp.373-401. Nathan, R., 2006. Long-distance dispersal of plants. Science (New York, N.Y.), 313(August), pp.786–788. 42 Nelson, G. J. 1973. Comments on Hennig's "phylogenetic systematics" and its influence on ichthyology. Syst. Zool, 21, pp.364-374. Nylander, J. a a et al., 2008. Accounting for phylogenetic uncertainty in biogeography: a Bayesian approach to dispersal-vicariance analysis of the thrushes (Aves: Turdus). Systematic biology, 57(2), pp.257–268. Praveen Karanth, K., 2006. Out-of-India Gondwanan origin of some tropical Asian biota. Current Science, 90(6), pp.789–792. Rajbhandary, S. et al., 2011. Asian Begonia: out of Africa via the Himalayas? Gardens’ Bulletin Singapore, 63, pp.277–286. Ree, R.H. et al., 2005. A likelihood framework for inferring the evolution of geographic range on phylogenetic trees. Evolution; international journal of organic evolution, 59(11), pp.2299–2311. Ree, R.H. & Sanmartín, I., 2009. Prospects and challenges for parametric models in historical biogeographical inference. Journal of Biogeography, 36, pp.1211–1220. Ree, R.H. & Smith, S. A, 2008. Maximum likelihood inference of geographic range evolution by dispersal, local extinction, and cladogenesis. Systematic biology, 57(1), pp.4–14. Renner, S.S. et al., 2010. Biogeography of the Monimiaceae (Laurales): A role for East Gondwana and long-distance dispersal, but not West Gondwana. Journal of Biogeography, 37(7), pp.1227–1238. Renner, S.S., 2005. Relaxed molecular clocks for dating historical plant dispersal events. Trends in Plant Science, 10(11), pp.550–558. Ronquist, F., 1997. Dispersal-vicariance analysis: A new approach to the quantification of historical biogeography. Systematic Biology, 46(l), pp.195–203. Ronquist, F. & Huelsenbeck, J.P., 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics, 19(12), pp.1572–1574. 43 Rosen, B.R., 1988. Biogeographic patterns: a perceptual overview. In Analytical Biogeography (Myers, A.A. and Giller, P.S., eds), pp. 23–55. Rutschmann, F. & Eriksson, T., 2004. Did Crypteroniaceae really disperse out of India? Molecular dating evidence from rbcL, ndhF, and rpl16 intron sequences. International Journal of plant Science, 165(S4), pp.S69–S89. Rutschmann, F., 2006. Molecular dating of phylogenetic trees: A brief review of current methods that estimate divergence times. Diversity and Distributions, 12, pp.35–48. Sanmartín, I. & Ronquist, F., 2004. Southern hemisphere biogeography inferred by event-based models: plant versus animal patterns. Systematic biology, 53(2), pp.216–243. Stamatakis, A., 2006. RAxML-VI-HPC: Maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics, 22(21), pp.2688–2690. Swenson, U. & Anderberg, A.A., 2005. Phylogeny, character evolution, and classification of Sapotaceae (Ericales). Cladistics, 21, pp.101–130. Swofford, D.L., 1993. PAUP*: Phylogenetic Analysis Using Parsimony. Thomas, D.C. et al., 2012. West to east dispersal and subsequent rapid diversification of the mega-diverse genus Begonia (Begoniaceae) in the Malesian archipelago. Journal of Biogeography, 39(1), pp.98–113. Von Humboldt, A. 1805. Essai sur la geographie des plantes accompagne d’un tableau physique des regions equinoxiales, fonde sur les mesures executees, depuis le dixieme degre de latitude boreale j’usqu’au dixieme degre de latitude australe, pendent les annees 1799, 1800, 1801, 1802 et 1803. Paris: Levrault Schoell. Von Humboldt, A., 1808. Ansichten der natur mit wissenschaftlichen erlauterungen. Tubingen, Germany: J.C. Cotta. 44 Wallace, A. R., 1860. On the Zoological Geography of the Malay Archipelago. Zoological Journal of the Linnaean Society, London 4: 172-184. Wallace, A. R., 1863. On the physical geography of the Malay archipelago. Journal of the Royal Geographical Societyof London 33: 217-234. Wallace, A. R., 1869. The Malay Archipelago. London: Macmillan. Wallace, A. R., 1876. The Geographical Distribution of Animals. London: Macmillan. Willdenow, K.L., 1811. The principles of Botany and of Vegetable Physiology. University of Edinburgh Press. Wegener, A., 1912. Die Entstehung der Kontinente. Geologische Rundschau 3(4): 276292. Wegener, A., 1966. The Origin of Continents and Oceans. New York: Dover Yang, Z. & Rannala, B., 2012. Molecular phylogenetics: principles and practice. Nature Reviews Genetics, 13(5), pp.303–314. Yu, Y., Harris, a. J. & He, X., 2010. S-DIVA (Statistical Dispersal-Vicariance Analysis): A tool for inferring biogeographic histories. Molecular Phylogenetics and Evolution, 56(2), pp.848–850. Zwickl, D.J., 2006. GARLI manual (version 0.95). Software manual, p.21. 45 CHAPTER 3: The early evolution and mode of range expansion in Asian Begonia Introduction Tropical areas contain a disproportionate amount of the world’s biodiversity. This diversity is not spread evenly throughout the tropics, but concentrated in certain regions. Many of these regions with exceptionally high endemism and diversity are under threat, and are called Biodiversity Hotspots, representing 1.4% of earth’s surface but containing 44% of flowering plant diversity (Myers et al. 2000). The origin of tropical hyperdiversity is only just beginning to be understood. The prevailing view of the importance of vicariance as a major influence on the evolution of the tropical flora has recently been challenged, as the arrival of molecular dating techniques and advancement of biogeographic models have revealed a much more recent origin for most Angiosperm families, and therefore more recent dispersal events provide much more plausible explanations for extant plant distribution patterns (Pennington & Dick 2004). Biodiversity hotspots provide an ideal opportunity to examine the origins and mode of range expansion of hyperdiverse tropical floras. Within continental Asia, three regions, Indo-Burma, Western Ghats and Sri Lanka, and South and central China have been declared as biodiversity hotspots with Indo-Burma and Western Ghats/Sri Lanka being placed amongst the hottest hotspots due to their rich endemic plant and vertebrate diversity (Myers et al. 2000). The biogeographic history of Malesia has been well documented and extensively studied (Crayn et al. 2014), however the broad scale biogeographic patterns of Continental Asia remain comparatively unknown despite the high floristic richness of the region. The plant family Begoniaceae makes an excellent candidate to test biogeographic patterns within continental Asia due to its high levels of species richness and endemism in the above biodiversity hotspots. Begoniaceae are a large pantropical family widely distributed in the tropics with greatest species diversity in Southeast Asia and South America. The largest genus Begonia dominates the family with ca. 1800 species including ca. 700 species in the neotropics, 160 species in Africa and 900 spp. in Asia with the bulk occurring in Southeast Asia (Goodall-Copestake et al. 2009; 46 Rajbhandary et al. 2011; Thomas et al. 2012). The monotypic Hillebrandia sandwicensisis a Hawaiian endemic (Clement et al. 2004). This morphologically diverse Begonia is characterised by asymmetrical leaves, unisexual flowers and monoecy, twisted, papillose stigmas and dry three winged capsules (Doorenbos et al. 1998, Judd et al. 2008). The Hawaiin endemic Hillebrandia deviates from Begonia by having more differentiated segments of the perianth semi inferior ovary and fruit dehiscence between styles (Clement et al. 2004, Forrest et al. 2005). Numerous phylogenetic, cytological and biogeographic studies have been carried out during past decades in order to resolve the relationships within this large, morphologically diverse family. The genus was circumscribed in to 63 sections by Doorenbos et al. (1998) and currently is classified into 69 sections (Hughes et al. 2015). All other sections except section Tetraphila shows restricted distribution in single continent; either in Asia, Africa or neotropics however single lineage of Tetraphila has been recently discovered in continental South East Asia (De wilde et al. 2011). Polyphyly of sections Sphenanthera, Platycentrum and Leprosae is evidenced (Forrest et al. 2005, Tebbit et al. 2006) based on nuclear ribosomal DNA. Several small sections were found to be nested within some large sections; section Baryandra within Philippine Diploclinum, sections Alicida and Putzeysia within Section Diploclinum and section Monopteron within Platycentrum (Rajbhandary et al. 2010; Thomas, 2010). Several chloroplast and nuclear markers have being used in resolving sectional and species level relationships as well as biogeographic analysis in Begonia (Dewitte et al. 2008). The phylogenetic relationships of African Begonia are well resolved by Plana (2003) by using 81 Begonia species covering 17 African sections with the use of chloroplast trnL intron. African lineages were found to be non-monophyletic and the study reveals the possibility of using trnL intron region in identifying major clades within the genera. Goodall-Copestake,(2005) suggested an African ancestry for both Asian and American Begonia based on ca. 13kb of sequence data from eleven regions of nuclear, 47 mitochondrial and chloroplast DNA of 31 Begonia species. Forrest et al.(2005) also confirmed African paraphyly based on nuclear large subunit and internal transcribed spacer sequences along with morphological data. Neotropical, Asian, and Socotran Begonia were found nested within the African Begonia. This pattern was later confirmed by Thomas et al. (2012). Within Asian Begonia, two well supported major clades were identified; one dominated by continental Asian taxa comprising species from sections Parvibegonia, Diploclinum, Platycentrum and Sphenanthera, and the second major clade comprise of sections Rideleyella, Bracteibegonia, Petermannia and polyphyletic Malesian section Diploclinium with Chinese section Coelocentrum being the sister. Himalayan Begonia have been found nested within Asian Begonia grade which fall in to two groups; one with a grade of tuberous, deciduous species and the other with evergreen rhizomatous species (Rajhbandry et al. 2011). Two major sections of Himalaya Begonia are identified; one old of Miocene age (sect Diploclinium) and the other comprising of sections Platycentrumand Sphenanthera which originated in the Pliocene. The nuclear ITS region resulted in a poorly resolved back bone for the phyogeny, however the results hint at China or Indochina being the potential areas of origins for Asian Begonia. The utility of chloroplast markers in resolving phylogenetic relationships were investigated by Thomas (2011), on Southeast Asian Begonia to test the monophyly of Asian sections and to build up the Asian Begonia phylogeny. Among those regions, three fast evolving non-coding cpDNA regions: ndhA intron, ndhF-rpl32 spacer and the rpl32-trnL regions contained higher percentage of variable and parsimony informative sites than the other chloroplast regions tested: matK, petD gene and intron, psbB gene, trnL intron. Further, large global Begonia datasets comprising Neotropical, Southeastern Asian and Philippine Begonia were based on these three chloroplast regions (Moonlight et al. 2015). Thus, in order to complement the lacking South Asian and Himalayan element and to obtain a highly resolved phylogeny 20 sequences were newly generated for the current study and added to existing Begonia dataset comprising 159 species. 48 I aim to shed light on the biogeographic history of the Continental Asian hotspots, focussing on three main hypotheses. Asian Begonia are known to be monophyletic (Thomas et al. 2012), however the point of entry to Asia is still uncertain. Thus, first I aim to test the point of entry to Asia prior to initial diversification from an African ancestor. Secondly, I investigate the routes of migration across continental Asia which explains extant distributions, and the biogeographic connections of the hotspots. Finally I examine the mode of range expansion and speciation within continental Asia, and the balance between vicariance and dispersal. Route of entry The Begonia lineage first originated in Africa, based on plastid and mitochondrial DNA phylogenies sampling thirty Begonia species with representatives from Africa, America and Asia (Goodall-Copestake et al. 2010). Asian and Socotran Begonia were shown to be monophyletic, confirmed in a more densely sampled study of 112 species by Thomas et al. (2012) who also demonstrated the Asian species began to diversify during the early to mid Miocene and subsequently dispersed into Malesia. Both Asian Begonia monophyly and an African origin for the genus were confirmed by Moonlight et al. (2015) using 268 samples from across the tropics. A second African lineage was discovered in Asia by de Wilde et al. (2011). This was a single dispersal event in Pleistocene, resulting in the generation of a single new species which has not played a further role in the broader picture of Asian Begonia diversity. There are several hypotheses on the route of entry of Begonia into Asia. One possibility is a migration from Africa to Asia through an Arabian corridor (Goodall-Copestake et al. 2010) during favourable warmer and wetter climatic periods in Paleogene to Neogene (Zachos et al. 2001). Extensive aridification and global cooling might resulted in extinctions, leaving only B. socotrana on Socotra (Goodall-Copestake et al. 2010). Long distance dispersal (LDD) is another possibility, although given the phylogenetic and population genetic patterns observed in Begonia LDD does not appear to be common (Hughes & Hollingsworth 2008). However the occurrence of B. afromigrata, an African element, in Laos and Thailand shows the feasibility of successful colonisation following LDD. The lack of evidence for a continuous wet forest corridor from Africa to Thailand during the Pleistocene favours long distance dispersal rather 49 than overland migration for this species (de Wilde et al. 2011).The route of entry for the monophyletic large radiation of Asian Begonia from Africa via LDD is perhaps more likely to be found in western or southern Asia given their proximity to the African area of origin for the genus. I aim to test the following hypotheses of routes of entry into Asia from Africa for the large genus Begonia (Figure 3.1). 1. Arabian corridor. During favourable climatic conditions Begonia could have been dispersed from Africa to Socotra and then migrated to Asia via a mesic Arabian corridor. If Socotra was a point on the route of entry to Asia, Socotran Begonia would form a sister clade to the rest of the Asian species. 2. Dispersal to Sri Lanka. If Sri Lanka was the route of entry, Sri Lankan Begonia would be the sister group to the rest of the Asian species. Long distance dispersal events could have taken place out of Africa to Sri Lanka during the time when the Deccan Plate containing Sri Lanka moved in a close proximity to Africa during its northward migration. Initial colonization in Sri Lanka could have been followed by dispersal to the Indian mainland followed by overland migration from West to East. Further, given the distribution of Begonia sect. Reichenheimea, found in Sri Lanka and with most diversity in Southeast Asia, there could have been some long dispersal from Sri Lanka to Southeast Asia. 3. Dispersal to India. If India acted as the route of entry via LDD, an Indian clade would take the sister position to the rest of the Asian species. Long distance dispersal events could have taken place out of Africa to the coastal mountains of western India during the northward migration of the Deccan Plate. 4. Dispersal to the Himalayas. Dispersal to the Himalayan range is another potential route of entry to Asia, and we would expect Himalayan Begonia to form a sister group to other Asian species, followed by migration to the rest of Asia, including to Sri Lanka via the Western Ghats. Both Socotra and Sri Lanka are isolated and we might expect dispersal to and from these relatively small areas to be infrequent, leading to a simple sister relationship in the phylogeny if these areas harbour species from the initial immigration. Concerning the Western Ghats of India or the Himalaya, we might expect a more complicated scenario 50 in the hypothetical phylogeny, since the large area provides more potential for multiple lineages and dispersal events, resulting in species from the Western Ghats or the Himalaya being paraphyletic with other Asian species nested within. Continental biogeography Even though molecular phylogenetic studies on Begoniaceae have greatly advanced our knowledge of the evolution of the family, the relationships and biogeography of continental early diverging Asian Begonia lineages remain poorly understood (GoodallCopestake et al. 2010; Thomas et al. 2012) and key samples from west and south Asia have not been included to date. Rajbhandary et al. (2011) increased the sampling of Himalayan species compared to earlier studies but failed to get an unequivocal biogeographic hypothesis as the phylogeny, based on ribosomal ITS sequences, was poorly supported.In this paper I have expanded the chloroplast sequence dataset from Thomas et al. (2012), which provides better phylogenetic support than ITS, by adding more samples from South Asia and the Himalayas. Thomas et al. (2012) coded continental Asia as a single geographic unit; my increased sampling allows me to expand this into five areas (India, Sri Lanka, Himalaya, China, Indo-China) to permit insight into development of the hotspot floras. 51 (1) (2) (3) (4) Figure 3.1 Phylogenetic representations showing the history of four hypothetical routes of entry of Begonia to Asia from Africa, and dispersal within Asia. The unrooted network at the bottom of the figure depicts hypothetical relationships among the major geographic areas with arrows indicating the placement of the root for each of the four colonization routes. 52 Mode of range expansion In addition to testing these broad biogeographic hypotheses, I investigate the mode of range expansion which has led to the monophyletic radiation generating ca. 900 species and covering all of tropical Asia. In order to obtain a complete picture of Asian Begonia biogeography, we need to understand the contributions of vicariance, dispersal and founder event speciation to the radiation. The parsimony, event based, quantitative method, Dispersal-vicariance analysis (DIVA) (Ronquist 1997) is a widely used method in inferring ancestral area relationships based on dispersal and extinction events (Ronquist 1997; Buerki et al. 2011) with vicariance given a zero cost. Phylogenetic branch length is not taken in to account (Clark et al. 2008) and optimal reconstruction is produced by a parsimony criterion that minimizes dispersal-extinction events (Ronquist 1997). DIVA can result in wide ancestral ranges (Kodandaramaiah 2010), and may not provide a good model for range evolution where it is not driven primarily by vicariance. The parametric counterpart of DIVA, Dispersal-Extinction-Cladogenesis (DEC) (Ree & Smith 2008) infers ancestral area reconstructions under a likelihood approach. Even though local extinction and dispersal are treated as free parameters, DEC assigns a weight of 1 to each allowed cladogenesis event, and the probability of each event is obtained through dividing by the sum of the weights of all the allowed events (Matzke 2014). Thus the cladogenesis model is fixed and range inheritance by daughter lineages by different mechanisms are considered to have an equal probability. DEC also could result in numerous equally likely ancestral reconstructions at the base (Moonlight et al. 2015) and the mechanism of identical range inheritance when the ancestral range comprises multiple unit areas is not addressed (Lamm & Redelings 2009). During cladogenesis when the ancestor is restricted to a single area, both DIVA and DEC models allow speciation within that area. However, when an ancestor occurs in more than one area, DEC assumes that one daughter lineage will always be restricted to one area and permits separation of the single area from the remaining ancestral range. However DIVA, on the other hand, permits classical vicariance where each daughter can occupy more than one area. When the ancestral area is wide, DIVA does not permit speciation within one area whereas DEC does (Ronquist & Sanmartín 2011). 53 Recent advancement in biogeographic analysis allows likelihoods to be assigned to models of range evolution for a particular dataset. The package Biogeobears (Matzke 2014) allows choosing the best-fit model under a maximum likelihood framework by modeling transitions between discrete states (biogeographical ranges) along phylogenetic branches as a function of time. The DEC model calculates maximum likelihood ratios of ancestral states at speciation events, in a method similar to LAGRANGE (Ree et al. 2005). In addition to two free parameters; d (dispersal), e (extinction) included in the DEC model (Batalha-Filho et al. 2014, Landis et al. 2013) the new additional parameter J is added to the model. The J parameter tests the contribution of founder event speciation to range expansion and controls the probability of two events during cladogenesis; founder event speciation versus sympatric and vicariant speciation (Matzke 2014). Founder event speciation is potentially important in lineage splitting especially in island systems, and here I use the functionality of Biogeobears to find the best fit model for the range expansion of Asian Begonia. My study is the first to incorporate endemic Sri Lankan Begonia in a global Begoniaceae phylogeny, potentially permitting new insights into the early evolution and biogeography of Asian Begonia. Materials and Methods Taxon sampling In order to represent potentially early divergent lineages in Asian Begonia the following samples (20 species; five Sri Lankan species, one Indian species, six Southeast Asian species and eight Himalayan species) were added to a previously published Asian Begonia DNA sequence alignment (Thomas et al. 2012) comprising 106 species: Begonia tenera and B. thwaitesii (both Sri Lankan endemics); B. cordifiolia, B. malabarica and B. dipetala (distributed in both in Sri Lanka and Western Ghats in India); B. albo-coccinia (Western Ghats). To shed more light on continental Asian Begonia biogeography, eight Himalayan species were added: B. diocia, B. flagellaris, B. picta, B. hatacoa, B. tribenensis, B. bryophila, B. panchtharensis and B. rubella. Voucher information for sequences used in the analysis is listed in table3.1. In addition, 53 samples of Begonia sect. Baryandra from the Philippines were added to complement the sampling in the Malesian region (Hughes et al. 2015). Asian Begonia have been 54 proved as monophyletic by previous studies (Goodall-Copestake et al. 2010; Rajbhandary et al. 2011; Thomas et al. 2012). Thus, I chose four African and five South American species as an out group for my study. The data set comprised 179 taxa representing 14 sections of Begonia. Voucher information for all the species used in the analysis are listed in appendix 1. Table 3.1: Voucher information for newly generated sequences for the study Species name Voucher/living accession EDNA number B. cordifolia Thwaites Kumarage 14 (E) EDNA13-0032556 B. thwaitesii Hook Kumarage 23 (E) EDNA13-0032557 B. dipetala Graham Kumarage 25 (E) EDNA13-0032558 B. malabarica Lam. Kumarage 28 (Photo voucher available) EDNA13-0032559 B. tenera Dryand Kumarage 68 (E) EDNA13-0032560 B. albo-coccinea Hook Photo voucher available B. puspitae Ardi 20111539 EDNA12-0029751 B. sublobata Jack 20101649 EDNA12-0025038 B. spec. 20100763 EDNA12-0025039 B. sect. Reichenheimea 20111543 EDNA12-0029747 B. sect. Reichenheimea 20111545 EDNA12-0029749 B. sect. Reichenheimea 20112191 EDNA12-0025037 B. picta Sm. 5993 EDNA14-0035336 B. tribenensis C.R.Rao 6043 EDNA14-0035338 B. bryophila Ined. 6100 EDNA14-0035340 B. rubella Ham. Ex D.Don 6000 EDNA14-0035341 B. hatacoa Buch.Ham. 5971 EDNA14-0035344 B. panchtharensis S.Rajbh. 5968 EDNA14-0035345 B. flagellaris Hara 6010 EDNA14-0035346 B. diocia Ham. Ex D.Don 13651 EDNA14-0035352 DNA Extraction, Sequencing and Alignment Total genomic data was extracted from silica gel dried material using the DNeasy Plant Mini Kit (Qiagen, UK) according to the manufacturer’s protocols. 25 μl PCR reactions were setup for amplification of both ITS and chloroplast regions. 55 Three non coding regions of plastid DNA (ndhA intron, ndhF-rpl32 spacer, rpl32-trnL spacer) were used in the current study to complement the previously available data. The primer sequences used in the study are listed in Table 3.2. For the amplification of plastid DNA regions, each 25μl of PCR mixture contained 15.25μl of ddH2O, 2.5μl of 10x reaction buffer, 1.25μl of 25mM MgCl2, 2.5 μl of 2mM dNTPs, 0.75μl of 10μM forward primer, 0.75 μl of 10μM reverse primer, 0.8μl0.4% BSA, 0.2 μl of Biotaq DNA polymerase (Bioline, UK) and 1μl of DNA template. The temperature profile included an initial template denaturation step of 80oC for 5 minutes, followed by 30 cycles of denaturation at 95oC for 1 min, primer annealing at 50oC for 1 min, primer extension at 65oC for 4 min and a final extension step at 65oC for 5 min. Amplified products were run on a 1% Agarose gel with Syber Safe as the staining agent and visualized in UV transiluminator. The PCR purification was done using EXOSAPIT as 7 μl reaction mixtures. 5μl of PCR product was mixed with 2μl of EXOSAP and incubated for 37oC for 15 minutes followed by 80o C for 15 minutes. Table 3.2: Primers used in the study for amplification of plastid DNA DNA region Primer Primer sequence References ndhA intron ndhAx1 ndhAx2 GCY CAA TCW ATT AGT TAT GAA ATA TTC GGT TGA CGC CAM ARA TTC CA Shaw et al. 2007 Shaw et al. 2007 ndhF-rpl32 rpl32-R ndhF Beg1F Beg2R CCA ATA TCC CTT YYT TTT CCA A GAA AGG TAT KAT CCA YGM ATA TT TGG ATG TGA AAG ACA TAT TTT GCT TTT GAA AAG GGT CAG TTA ATA ACA A Shaw et al. 2007 Shaw et al. 2007 Thomas et al. 2014 Thomas et al.2014 Rpl32-trnL rpl32-F trnL CAG TTC CAA AAA AAC GTA CTT A CTG CTT CCT AAG AGC AGC GT Shaw et al. 2007 Shaw et al. 2007 Sequencing PCR was done using purified PCR products as 10μl mixtures using 5.68μl of ddH2O, 2μl of sequencing buffer, 0.32μl of primer, 1μl of Big dye and 1μl of template. The sequencing PCR protocol was, denaturation at 95oC for 30 sec, followed by 24 cycles of primer annealing at 50oC for 20sec, extension at 60oC for 4 min. Separate forward and reverse sequencing PCR s were carried out and products were sent to the Genepool facility at the University of Edinburgh (Genepool, UK) for BigDye Terminator Cycle sequencing. 56 Sequences were manually aligned in Mesquite (3.03; Maddison & Maddison 2015). The following bases were excluded due to the ambiguities in the alignment or missing data at the region ends: 55-103 462-476 1092-1111 1230-1249 1263-1280 2046-2064 21362150 2258-2284 2368-2379 3370-3410 3571-3604 3971-4012 4485-4507. Phylogenetic analysis and divergence time estimation The plastid data set was analyzed separately under Bayesian inference (BI) using MrBayes 3.2.1 (Ronquist et al. 2012) in the CIPRES science gateway V. 3.3 (Miller et al. 2010) and treated as single partitions. MCMC runs were carried out for 10,000,000 generations and sampled every 1000 generations. A 25% burn-in was set to discard the first set of trees and the remaining trees were summarised as a 50% majority rule consensus tree, visualized in FigTree (Figure 3.2) (Rambaut 2009). Models of sequence evolution were determined using jmodel test 2.1.3 (Posada et al. 2012). Twenty four models were tested and the optimal model was chosen using the Akaike Information Criterion (AICc) and the Bayesian Information Criterion (BIC). The GTR+I+G model was the most probable model under both criteria for the 3-region dataset. Bayesian divergence time estimation was performed using BEAST v.1.8.0 (Drummond & Rambaut 2007). The data set was treated as a single partition and run under a lognormal relaxed molecular clock. The only Begonia macrofossils known (Sults & Axsmith 2011) is too young to provide a calibration point and lacks the synapomorphies for the placement of the family. Thus, the dates obtained from primary calibration of broadly sampled phylogeny of Cucurbitales-Fagales dataset by Thomas et al. 2012 were used as a secondary calibration. A similar calibration has been utilized in more recent studies on the divergence time estimates of Begoniaceae (Moonlight et al. 2015; Hughes et al. 2015; Rajbhandary et al. 2011). Two calibration points were used for the divergence time estimates for the most recent common ancestor for Begonia. The crown group was given a mean age of 24 Ma and all Asian Begonia a mean age of 18.2 Ma. To account for the uncertainty of the age 57 estimates of primary analysis,the crown group was modelled with a normal distribution with a standard deviation of 3.57 Ma (Thomas et al. 2014). Four separate Markov Chain Monte Carlo (MCMC) runs were carried out for 10000000 generations sampling every 1000 generations under Birth-Death model of speciation. Plots of the logged parameters for each run were visualised using Tracer v.1.5 (Drummond & Rambaut 2007) to confirm convergence between runs. Trees were combined in LOGCOMBINER (Drummond & Rambaut 2007) and burn-in was set for 25% for initial sample for each run and a single maximum clade credibility tree was obtained from Tree Annotator v.1.7.5 (Drummond & Rambaut 2007) and visualized in fig tree v.1.4.0 (Rambaut 2009). Biogeographic analysis Twenty-two geographic areas were coded based on extant distributions and areas of endemism, to permit a higher resolution of the range expansion in continental Asia and the mode of range expansion across the entire study area: (1) Americas; (2) Africa; (3) Socotra; (4) India; (5) Sri Lanka; (6) Himalaya; (7) China; (8) Indo-China; (9) Peninsular Malaysia; (10) Sumatra; (11) Borneo; (12) Sulawesi; (13) Papua New Guinea; (14) Negros; (15) Java-Lesser Sunda; (16) Luzon; (17) Lanyu&Batan; (18) Panay; (19) Palawan; (20)Mindanao; (21)Sibuyan; (22) Biliran. The data matrix was prepared by coding presence/ absence in each of the areas (Appendix 2) based on data in Hughes et al. (2015). The majority of Begonia species are narrow endemics (Hughes & Hollingsworth 2008), hence the maximum number of areas permitted in the ancestral range reconstructions was constrained to two in order to reflect biological reality and to constrain analysis time. Only the widespread Begonia longifolia and Begonia palmata are found in more than two geographic areas as coded in this study. They were coded as present in China and Indo-China by considering their current distribution hotspots and likely area of origin in Rajbhandary et al. (2011). Ancestral areas at internal nodes were constructed using Biogeobears (Matzke 2013) under four models; DIVA-like (Ronquist 1997), DIVA-LIKE+J, DEC (Ree & Smith 2008) and DEC+J. Log likelihood values for each model were compared to identify the best model for inferring ancestral ranges at nodes. 58 Phylogenetic analyses - three-region matrix Descriptive statistics for 179 taxa plastid dataset and nucleotide partitions including amplicon length, alignment length, number of included characters and number of variable characters are given in the table 3.3. Trees derived from the non coding plastid DNA sequences of 179 taxa are presented in figures 3.2 & 3.3. Neotropical and Asian Begonia are nested within an African Begonia grade, with Asian+Socotran Begonia resolving as a strongly supported clade with PP=1.0 (Clade I). The Asian+Socotran Begonia clade has two major subclades, clades II and III (Figure 3.3). Table 3.3 Dataset descriptive statistics DNA region Alignment length Number of variable characters Number of informative characters Combined data set 4271 1537 (35.98%) 838 (19.62%) ndhA 1419 467 (32.91%) 240 (16.91%) ndhF-rpl32 1036 412 (39.76%) 240 (23.16%) rpl32-trnL 1816 658 (36.23%) 358 (19.71%) Clade II is a geographic mosaic and consists of species mainly from Socotra, Sri Lanka, India, Himalaya, and Eastern Asia; however it is not supported (PP=0.28). It contains the western limit of Asian Begonia and is the sister to clade III comprising largely eastern Asian Begonia. Clade III also does not show any support (PP=0.13) with the early diverging lineage of the Himalayan Begonia dioica being the sister to all other taxa. Hence clades II, III and B. dioica form three effectively unresolved lineages at the base of the Asian+Socotra clade. Within clade III three species from China and Indochina (B. morsei, B. ningmingensis and B. masoniana) form a strongly supported clade (PP=1.0) and B. kingiana from Peninsular Malaysia is sister to a largely Malesian clade comprising Philippine Begonia (Clade XIII) and other South East Asian Begonia (Clade XIV). Two Socotran species, B. socotrana and B. samahensis, together with Indo-Sri Lankan species form clade IV with PP=0.92 with Socotran Begonia being the sister group to the Indo-Sri Lankan elements (clade V). Clade V is strongly supported (PP=1.0) comprising 59 a grade of Indian and Sri Lankan Begonia (B. malabarica, B. dipetala, B. cordifolia) with two Sri Lankan endemics, B. tenera and B. thwaitesii, nested within. The remaining Indian taxa form clade VIII (B. floccifera, B. albo-coccinea, B. malabarica) and are sister to everything else in the poorly supported Clade VII (PP= 0.59), making Indian Begonia paraphyletic. The Indo-Sri Lankan species, B. malabarica appears separately in clade V and Clade VIII. Sri Lankan B. malabarica grouped with B. dipetala and it is highly supported as monophyletic (PP=1.0) while Indian B. malabarica grouped with B. albo-coccinea with high support value (PP=1.0). Clade IX (PP=0.99) is dominated by continental Asian (China+Indochina) Begonia with Himalayan species highly nested within. Clade XI (PP=1.0) consists of a paraphyletic group of Himalayan species intermixed with species from China, Indochina, Peninsular Malaysia, Sumatra and Sulawesi. The largely Philippine Begonia sect.Baryandra (Clade XIV) resolves as monophyletic (PP=1.0) while other Southeast Asian Begonia form Clade XV, PP= 0.58. Divergence time estimates The mean divergence time estimate for Asian Begonia is 15.5 Mya. The age of the Socotran/Indo-Sri Lankan clade (Clade IV) is 13.5 (8.0-19.2) Mya which is similar to age estimates by Thomas et al. (2012) of 13.6 (7.3–19.5) Mya. The Indo-Sri Lankan element (Clade V) shows a divergence time of 7.7 (3.9-12.6) Mya while the Sri Lankan endemics (Clade VI) show a more recent Pleistocene diversification of 0.8 (0.21.8) Mya. The rest of the Indian taxa (Clade VIII) have a stem age of 15.1 (9.6-21) Mya and a crown age of 5.9 (3-9.4) Mya. Clade IX, dominated by species from Indo-China and Himalaya begins to diversify 14.4 (7.7-17.2) Mya. Himalayan Begonia (Clade XI) have an origin 7.9 (4.8-11.6) Mya and comprises two main sub clades. One is endemic to the Himalaya, and the other also contains species from Sumatra, Sulawesi, China and Indochina; clade XII which has a crown age of 5.2 (3.1-7.9) Mya. 60 Begonia diocia from Himalaya shows an early branching position in the phylogeny at 15.8 Mya and is weakly supported at the base of clade III. South East Asian Begonia (clades XIV and XV) has a crown age of (12.5 Mya). Biogeographic analysis and ancestral area reconstructions Amongst the models used in BIOGEOBEARS for ancestral area reconstructions, the models with the J parameter included gave results with higher likelihood values (Table 3.4), with DEC+J giving the highest likelihood score of -337.91, and was chosen as the best fitting model for the data. Table 3.4 d (dispersal), e (extinction), j (j value, founder-event speciation) LnL (log likelihood) for each of the geographic range evolution models compared in Biogeobears. Model d e j LnL DEC 0.0045 0.0213 0 -417.47 DEC+J 7.00E-04 0.0013 0.0091 -337.91 DIVA LIKE 0.0043 0.009 0 -402.38 DIVA LIKE + J 8.00E-04 0 0.0089 -341.79 61 Figure 3.2 Bayesian majority rule consensus tree resulted from non coding region of plastid DNA. Bayesian posterior probability (PP) support values are indicated next to the nodes. 62 Figure 3.3 Maximum-clade-credibility chronogram of a Beast analysis of the three-region Begonia data set. Node heights indicate mean ages. Numbers at nodes represent clades in Table 3.2. Branches coloured according to their optimal range reconstructions under the DEC+J model in the package Biogeobears. Pie charts show the relative probability of ancestral state reconstructions at selected nodes. Geological epochs are indicated by background colour: light grey, Miocene (5.3–2.3 Mya); mid-grey, Pliocene (0.26–5.3 Mya); dark grey, Holocene and Pleistocene (0–0.26 Mya). Dotted lines indicate posterior clade probabilities less than 0.95. 63 Figure 3.4 Maximum clade credibility chronogram of the three-region Begonia dataset. Node heights indicate mean ages. Node bars indicate 95% highest posterior density date ranges. Numbers inside boxes at each node represent node numbers and values next to nodes are the posterior probability values for each node. 64 The most probable ancestral areas for Asian Begonia clade II are India (D)=0.66, Sri Lanka (E)=0.14,Indochina (H)=0.1. My results in terms of tree topology best match my hypothesis of India being the point of entry to Asia. India Begonia did not resolve as a monophyletic sister group to rest of the Asian Begonia, but are paraphyletic at the base, thus potentially are the origin for all other Asian Begonia. However, these results depend on the rather poorly supported topology at the base of the clade, and should hence be interpreted as a hypothesis which needs further testing. This lack of resolution is probably be due to very rapid range expansion of Asian Begonia after initial colonization. Clade III potentially has an Himalayan origin with probability values of Himalaya (F)=0.45, China (G)=0.2, Indochina (H)=0.11, however this is rather uncertain due to poor support at the base of the clade. India is constructed as most probable ancestral area for clade IV containing Socotran and Indo Sri Lankan Begonia with probabilities of India (D)=0.62, Sri Lanka (E)=0.18, Socotra (C)=0.11. A dispersal to Socotra from India during the mid-Miocene is suggested, with the two Socotran endemics B.socotrana and B. samahensis arising from more recent Pleistocene speciation in situ. The Indo Sri Lankan Clade V has an Indo Sri Lankan origin (India+Sri Lanka, DE=0.5) and two Sri Lankan endemics B. tenera and B. thwaitesii are the result of Pleistocene immigration from India followed by in situ speciation thereafter. The rest of the Indian species in Clade VIII have an Indian origin with area probabilities of India (D)=0.78, India+Sri Lanka(DE)=0.22. The ancestral area for Continental Asian Begonia inClade IX is constructed as Indochina (H=0.97), which is also the ancestral area for Himalayan Begonia Clade X (Indochina, H=0.95). Within clade X, multiple dispersal events between China, Indochina, Peninsular Malaysia, Sumatra, Sulawesi and Himalaya during MiocenePliocene are inferred. The geographic area of origin for Clade XIII is constructed as China during MidMiocene; China (G) = 0.38, Peninsular Malaysia (I) = 0.22, China+Sumatra (G+J) = 65 0.14 and the biogeographic history within this region is fully resolved by Thomas et al. (2012). Table 3.5 Posterior probabilities, divergence ages and ancestral area probabilities; C= Socotra, D= India, E= Sri Lanka, F= Himalaya, G= China, H=Indochina, I= Peninsular Malaysia, J= Sumatra, S= Palawan, P= Luzon, R= Panay. Clade No PP Divergence age Ancestral Area Probability Clade I 1 16.2 (10.34-22.51) D= 0.32 F=0.21 H= 0.1 Clade II 0.28 15.54 D= 0.66 E=0.14 H=0.1 Clade III 0.13 15.85 F= 0.45 G= 0.2 H= 0.11 Clade IV 0.92 13.48 (8.08-19.19) D=0.62 E=0.18 C=0.11 Clade V 1 7.68 (3.91-12.62) DE= 0.5 D= 0.3 E=0.2 Clade VI 1 2.82 (1.16-5.01) DE= 0.67 E=0.31 Clade VII 0.59 15.13 (9.57-20.95) D= 0.69 H=0.15 Clade VIII 1 5.92 (2.98-9.42) D= 0.78 DE= 0.22 Clade IX 1 14.43 (9.24-20.34) H= 0.97 G= 0.01 Clade X 1 9.83 (5.98-13.89) H= 0.95 FH= 0.04 Clade XI 1 8.54 (5-12.21) F= 0.86 H= 0.05 FH= 0.07 Clade XII 1 5.19 (3.06-7.87) F= 0.83 FH= 0.08 GH= 0.04 Clade XIII 1 13.55 (8.49-19.21) G= 0.38 I= 0.22 GJ= 0.14 Clade XIV 1 8.72 (5.5-12.6) S=0.36 P=0.34 R=0.19 Clade XV 0.58 10.82 (6.73-15.49) J=0.99 E=0.12 GH= 0.01 Discussion Route of entry to Asia The Afro-Arabian plate collided with Eurasia around 25 Mya during the late Oligocene (Yuan et al. 2005). The appearance of the Gulf of Aden rift caused the Arabian peninsula to become separated from continental Africa and Socotra ca. 10 Mya (Ghebreab 1998). My first hypothesis of Socotra being the point of route of entry to Asia, possibly facilitated by a mesic Arabian corridor, is rejected, as SocotranBegonia are not a sister group to all other Asian species, but nested within and likely to be of Indian origin. Further, evaporate and calcrete deposits suggest dry and warm conditions for areas of western Asia and the Arabian Peninsula during mid Miocene (Morley 2000; 66 Scotese 2003). Thus, overland migration of Begonia from Africa to Asia via dry corridor seems unlikely. Since the separation of the Indian and African plates 132 mya (Ali & Aitchison 2008) Sri Lanka has remained static relative to India, together forming the Deccan plate. Sri Lanka is separated from India by the shallow, narrow Palk Strait. Although the Deccan plate could potentially have provided a refuge for African elements during its northward journey, in the case of Begonia these would have arrived by LDD given the age of the Asian clade. The hypothesis of Sri Lanka as the point of entry to Asia for Begonia is rejected, from the ancestral area reconstruction and also since Sri Lankan Begonia are not sister to other Asian species; indeed the Sri Lanka endemics are resolved as Pleistocene immigrants from India. There is no evidence for Himalaya to be the point of entry to Asia due to the young age of the Himalayan clade and the nesting of most Himalayan species within eastern Asian Begonia. The point of entry of Begonia to Asia is obscure, since the topology at the base of the clade is poorly supported with PP=0.28. This lack of resolution is probably due to very rapid range expansion during the Miocene. India harbours 61 native Begonia species (Rashid & Rahman 2012), however only 4 species (B. albo coccinea, B. floccifera, B. malabarica, B. dipetala) have been incorporated in the biogeograhic analysis due to practical difficulties in getting material from India. Indian Begoniabelong to early diverging lineages, and lack of sampling may have contributed to the low back bone support in the phylogeny. If more Indian material could be incorporated in future analysis that would be beneficial in order to contribute towards a better resolved phylogeny due to breaking up some of the long early divergent branches in the tree. Assuming the presented topology is correct, India is the most probable area for the route of entry to Asia and the initial colonizer could have undergone a very rapid radiation during favourable warm and moist conditions during Miocene in the Indian subcontinent (Yuan et al. 2005; Zachos et al. 2001). Subsequent dispersals then occurred to the islands Lanka and Socotra and to Southeast Asia via Continental Asia. 67 There is some evidence for a plant dispersal phase which occurred between Africa and India via Madagascar when the Deccan plate moved in close proximity to Africa during Cenomanian to Turonian times until Madagascar was separated from India and Seychelles about 90 Mya (Briggs et al. 2003; Conti et al. 2002 ; McLoughlin 2001). It is considered as one of the major inter plate dispersal paths for megathermal angiosperms and evidenced by plant families such as Sapindaceae, Palmae and Myrtaceae (Morley 2003; Morley 2000). This dispersal path was believed to cease when Madagascar was separated from India and Seychelles. Madagascar harbours 44 Begonia species, of which 42 are considered as endemic. Only one species from Madagascar, B. goudotii is incorporated in the analysis here. However, previous studies (Plana et al. 2003) have shown other Madagascan species to belong to a single clade. The Madagascan Begoniasampled here does not show a close affinity to Indo-Sri Lankan Begonia and the dates of entry are too young to support a dispersal pathway between Madagascar and Indo-Sri Lankan region for Begoniaceae. The timing of entry of 15 Ma for Asian Begonia is highly congruent with the midMiocene climatic optimum peaking during 17-15 Mya when the global ice volume was low and bottom seawater temperatures were slightly higher (Zachos et al. 2001). Further, the moist and warm climates led to the expansion of megathermal vegetation in Asia up to Southern Japan in the north east and up to the northwest of the Indian subcontinent (Zachos et al. 2001; Morley 2000). The occurrence of rain forest taxa in the mid-Miocene fossil floras from Rajasthan in the northwest of the Indian peninsula confirms the existence of rain forests during that time (Lakhanpal 1970; Raven & Axelrod 1974). However, migration of Begonia across dry Arabian corridor is unlikely, thus long distance dispersals from Africa are more likely to have facilitated the middle Miocene arrival of Begonia into India and subsequent radiation thereafter. Biogeography of Asian Begonia Although the route of entry of Begonia into Asia remains somewhat speculative, a clearer picture of biogeographic range evolution of Begonia amongst the Asian hotspots has been resolved. 68 Early evolution Based on the topology in Figure 3.3, the initial diversification of Asian Begonia occurred in India during the mid-Miocene followed by multiple dispersal events; two independent dispersals to Eastern continental Asia, a single dispersal to Sri Lanka and a single dispersal to Socotra. The warm phase during the mid-Miocene climatic optimum resulted in the expansion of mega thermal vegetation in Asia with spreading of rain forests throughout the continental Asia facilitate the east to west migration of Begonia to current hotspots under favourable climatic conditions. Further, dispersal from continental Asia to the Sunda Shelf region can potentially be explained by overland dispersal during the mid-Miocene when the moist and warm climates were predominant in South east Asia (Thomas et al. 2012). Himalaya The geological history of the Himalayas and the Tibetan Highlands is still is poorly understood and the timing of uplift and the evolution of regional climate remains enigmatic (Kutzbach et al. 1989, Molnar et al. 2010). Two separate origins can be identified for Himalaya Begonia; firstly, clade XI where Himalayan species are highly nested within an eastern Asian Begonia grade with an area of origin optimised as Indo-China (F) (P=0.97) and the second consisting of the phylogenetically isolated B. diocia which is sister to Chinese and South East Asian Begonia, albeit with very low support. The diversification of Himalayan Begonia sect Diploclinium, B. picta, B. tribenensis, B. rubella, B. flagellaris and B. bryophila (part of the ‘Diploclinium Grade’ of Rajbhandary et al. (2011) began 7.9 (4.8-11.4) Mya during late Miocene which is highly congruent with the onset of South Asian monsoons at 7.4 Mya (Copeland 1997). This is concurrent with the uplift of the Himalaya and the Tibetan plateau ca. 10 ma which reached sufficient altitudes to alter rain patterns in the Himalayan region (Guo et al. 2004; Molnar et al. 2010; Zheng et al. 2004). The Diploclinium grade is characterised by tuberous, seasonally adapted Begonia in which blooming is restricted to three months after the onset of monsoons followed by fruit dispersal during the following dry period (Rajbhandary et al. 2011). 69 The second major Himalayan grade composing Begonia sect Platycentrum and Sphenanthera intermixed with Begonia from China, Indochina, Peninsular Malaysia, Sumatra, Sulawesi, which diversified during 5.2 (3.1-7.9) Mya, indicating back and forth dispersal between the Himalaya and other regions. Continuing rapid uplift of the Tibetan plateau resulted in further intensification of East Asian monsoons. The species in this grade are evergreen, and have fruits adapted to rain-splash seed dispersal thus occupied in habitats where water is more constantly available (Rajbhandary et al. 2011). More recent speciation during the Pleistocene may have occurred in response to Pliocene-Pleistocene climatic cycles and changes in the monsoon intensity (Guo et al. 2004; Zheng et al. 2004). The position of Himalayan B. diocia in the phylogeny is enigmatic and its placement at the base of the phylogeny is effectively unresolved (Figure 3.3). Further, the age of 15.8 Mya for the species predates the onset of South Asian monsoons during late Miocene, and whether the Himalaya had achieved sufficient height during that time to receive monsoon rain is uncertain. This relatively old lineage in Asian Begonia is further evidence for an area of origin in the region of the Indian Subcontinent. Thus, most Himalayan Begonia are a result of back dispersal from Indo-china during late Miocene followed by further back and forth dispersal between the Himalaya and Indo-China during the Pliocene-Pleistocene which can be linked with climatic cycles during that time (Janssens et al. 2009). A similar scenario is observed in the genus Impatiens (Balsaminaceae), one of the most species rich genera of flowering plants with hotspots occurring in tropical Africa, the Himalayan region, Madagascar, South India and Sri Lanka and Southeast Asia. Himalayan species appear to have a Chinese origin with multiple separate dispersal events occurring during the late Miocene (Janssens et al. 2009). Sri Lankan biogeography The Deccan plate, comprising of India and Sri Lanka underwent remarkable climatic and vegetational changes during its northward journey to Eurasia (Conti 2002; McLoughlin 2001; Morley 2003). The plate would have experienced aridification as it moved north through the horse latitudes at 30-35 degrees where patterns of atmospheric circulation result in very little precipitation and the formation of the desert belt 70 (Lomolino et al. 1998), potentially causing a substantial amount of taxa to become extinct (Rutschmann & Eriksson 2004). The massive volcanism at the CretaceousTertiary boundary during 65 Ma and extensive aridification during early Tertiary due to the uplift of Himalaya resulted in further impoverishment of allochthonous African elements from the Indian flora (Rutschmann & Eriksson 2004; Morley 2003; Conti 2002; McLoughlin 2001) leaving the Western Ghats and Sri Lanka as refugial areas for those flora. The base of the Indo Sri Lankan clade V comprises of B. dipetala and B. malabarica and confirms a likely Indian origin for Sri Lankan Begonia with a more recent entry to Sri Lanka during the late Miocene ca. 7.7 Mya (3.9-12.6) which coincides with the development of South Asian monsoons during late Miocene 7.4 Mya (Copeland 1997). The intervening continental shelf between Sri Lanka and India provided an intermittent land connection until the last sea level rise 6000 years ago during the Holocene (McLoughlin 2001; Ashton & Gunatilleke 1987) and has potential to act as a migratory path for floristic exchange between the two land masses. However as the endemic Sri Lanka Begoniaare restricted to altitudes of 1000-1200m, the overland migration via a lowland land bridge seems unlikely. The Western Ghats of India and the sub montane forests in Sri Lanka could have act as archipelago-like systems facilitating plant dispersals among them. The Sri Lankan endemics, B. tenera and B. thwaitesii, are the result of speciation in situ. Pleistocene deposits from the wet zone of Sri Lanka provides evidence for intermittent periods of seasonal tropical climate (Ashton & Gunatilleke 1987) which would favor in the recent radiation of tropical taxa in the island. The monophyletic grouping of the Indian origin sample of B. malabarica with B. albococcinea is odd morphologically, since they have completely different life forms; B. malabarica is a woody cane-like species much more similar to B. dipetala, whilst B. albo-coccinea is a rhizomatous herb. This placement is best explained by hybridization resulting in chloroplast capture by the Indian B. malabarica lineage. The grouping of Sri Lankan B. malabarica with Sri Lankan B. dipetala is sensible and the preferred phylogenetic hypothesis for this species. Incongruence between plastid and nuclear data sets can be mainly due to hybridization, however some factors such as recombination, gene paralogy and pseudogene formation 71 can result in incongruences between nuclear and plastid datasets (Feliner & Rosello 2007; Linder & Rieseberg 2004; Small et al. 2001). Further analysis with additional DNA markers and samples need to be carried out in order to resolve this uncertainty in the placement of B. malabarica from India and Sri Lanka. The use of the nuclear ITS region might be a solution to examine this incongruence further, as it is bi- parentally inherited and can potentially track the paternal lineage, as shown in Hughes et al. (2015). Both nuclear and chloroplast genotypes can become introgressed in to other species resulting in interspecies hybridization and it is highly evidence in the species of Begonia sect Baryandra. This would result in a “plastid pool” where there is combination of different genotypes of neighbouring species would ultimately result in species with new genotypes (Hughes et al. 2015). Socotra The ancestral area reconstruction for Socotran Begonia and the lack of a sister relationship to other Asian Begonia favours an Indian origin (India (D)=0.62), thus long distance dispersal from India during the mid-Miocene is inferred. The conspicuous long branches suggest an extinction of taxa or long isolation of Socotran lineage during the Neogene drought towards the end of the Miocene (Yuan et al. 2005; Morley 2000) in the Arabian and Socotran region. These endemics are also the result of more recent Pliocene-Pleistocene speciation. Exacum (Gentianaceae) which is distributed in Africa, Madagascar, Socotra, the Arabian peninsula, Sri Lanka, India, the Himalayas, mainland Southeast Asia and northern Australia, exhibits a similar biogeographic scenario to Begonia where long distance dispersal followed by extensive range expansion is found to be the most plausible reason for its extant distribution (Yuan et al. 2005). Sri Lanka and southern India were initially colonized by a Madagascan ancestor via long-distance dispersal and subsequently dispersed to Socotra-Arabia, northern India, and mainland Southeast Asia around the northern Indian Ocean Basin when the climate was warm and humid. The species were survived in isolation in refugia such as Socotra-Arabia, southern India–Sri Lanka, and perhaps mainland Southeast Asia during the Neogene drought and secondary diversification resulted in current distribution patterns (Yuan et al. 2005). 72 Pollination mechanisms, seed and fruit dispersal Begonia species exhibits a low intraspecific gene flow, due to poor seed dispersal. Most Begonia are zoophilous, which are pollinated by generalized pollinators such as stingless bees (Trigona species), honey bees (Apis cerana), and bumble bees (Bombus ephippiatus) and flies (Hughes & Hollingsworth 2008; Kiew 2005; Dewitte et al. 2011). Begonia female flowers do not offer rewards to pollinators, however they mimic the same colouration of yellow or orange colour of androecia of male flowers. Male flowers offer pollen for pollinators, and pollinators mistakenly visit female flowers and pollination can takes place (Renner 2006; Schemske et al. 1996). However some species in the neotropics such as Begonia boliviensis and Begonia ferrugenia are bird pollinationg species which possesses synpetalous, tubular perigones which is an adaptation for bird pollinating species (Hughes 2002). Zoochory is common in African Begonia and though has been directly observed in Asian Begonia (Tebbitt et al. 2006). However, some Asian sections like Sphenanthera and Leprosapossess indehiscent, thick, fleshy pericarps with bright colours which which is an indication for dispersal animals such as birds, bats and other vertebrates (Tebbitt et al. 2006). However, majority of Asian species are anemochorous which produce large amount of minute seeds in large quantities. The capsules are dry with a membranous pericarp and the winged seeds are released through the slits when the capsules are shaken by the wind. The seeds produced are very tiny, size ranging from 300-600μm in length, which can be carried out by the wind effectively (de Lange & Bouman 1992; Kiew 2005). However, wind dispersal responsible only for short distance dispersals which is proven by limited gene flow and geographic isolation by most Begonia populations (Hughes &Hollingsworth 2008; Matolweni et al. 2000). Further, some alternative dispersal mechanisms are seen in some Asian Begonia such as section Platycentrum. Their coriaceous pericarp, rain ballist capsule and unequal wings are characteriscs of a rain ballist syndrome. At the maturity pedicels get curved and the pericarp and smaller wings facing upwards while larger wing facing downwards. The rain drops cause the splash cups to move up and down thus allowing the seeds to get released from dehisced septa. The species in section Parvibegonia show bot different 73 mechanism, the capsules get dried and hand downward at maturity allowing the seeds to get dispersed by wind or rain (Tebbitt et al. 2006; Kiew 2005; Rajbhandry et al. 2011). Thus most of the species are geographically isolated with narrow geographical ranges. Thus it is considered, long term isolation has led to more speciation than genetic drift between populations (Hughes & Hollingsworth 2008; Dewitte et al. 2011). The genome of Begonia is highly diversed with large variation in chromosome length, width, total chromosome number and large secondary constrictions can be observed in those genotypes (Dewitte 2009a). Most of species possess chromosomes with 0.53.6μm in length and 0.3-1.5μm in width and the chromosomes of South American Begonia are smaller than that of Asian, African and Middle American Begonia (Dewitte 2009a). A wide range of chromosome numbers has been reported, ranging from 2n=16 (B. rex) to 2n=156 (B acutifolia) (Doorenbos et al. 1998; Legro & Doorenbos 1969). It is suggested that X=13 (2n=26) is the basic chromosome number for Begonia across the world and others are a result of polyplodisation which should have been played a wide role in the diversification of lineages. Ployploides are either produced by means of somatic mutations in meristematic cells or unreduced gametes (Bretagnolle & Thompson 1995; Otto & Whitton 2000; Dewitte et al. 2011). Among 70 investigated genotypes, 10 produced unreduced gametes and this can be observed in both species and hybrids (Dewitte et al. 2009b). A high level of heterozygosity is transferred thorough 2n gametes and in some hybrids, viability only constrained within the 2n gametes. Thus, the polyploidy can be considered as a major factor contributing to diversification of lineages however it is considered that the frequency of occurrence is very low in the wild (Dewitte et al. 2011). Mode of range expansion My results confirm the importance of founder event speciation in Begonia, in which a daughter lineage jumps to an area completely outside the ancestral range immediately followed by speciation. The J parameter controls the probability of two events during cladogenesis; founder event speciation versus sympatric and vicariant speciation (Matzke 2014). 74 My present study confirms a recent origin (15.5 Mya) for Asian Begonia and a history dominated by a small number of long-distance dispersal events followed by extensive radiations in current hotspots. The youth of the Sri Lankan elements challenges the relict vicariant history for the evolution of the flora, and adds to a growing body of evidence supporting youthful tropics at the species level (Pennington et al. 2015). The early divergence of the Himalayan B. dioica is surprising and a hint at relictual elements in a relatively young montane landscape which begs further investigation. References Ashton, P.S. & Gunatilleke, C.V.S., 1987. New light on the plant geography of Ceylon I . Historical plant geography. Journal of Biogeography, 14(3), pp.249–285. Ali, J.R. & Aitchison, J.C., 2008. Gondwana to Asia: Plate tectonics, paleogeography and the biological connectivity of the Indian sub-continent from the Middle Jurassic through latest Eocene (166-35??Ma). Earth-Science Reviews, 88(3-4), pp.145–166. Batalha-Filho, H. et al., 2014. Phylogeny and historical biogeography of gnateaters (Passeriformes, Conopophagidae) in the South America forests. Molecular Phylogenetics and Evolution, 79, pp.422–432. Bretagnolle, F. & Thompson, J.D. (1995). Gametes with the somatic chromosome number: mechanisms of their formation and role in the evolution of autoploid plants. New Phytologist, Vol.129, No.1, (Januari 2005), pp. 1-22. Briggs, J.C., Collections, Z. & Museum, G., 2003. The biogeographic and tectonic history of India.Journal of Biogeography, 30, pp.381–388. Buerki S., et al,. 2011. An evaluation of new parsimony-based versus parametric inference methods in biogeography: a case study using the globally distributed plant family Sapindaceae. J Biogeog, 38, pp.531–550. Burton, N., 2012. Phylogenetics of Begonia sect. Gireoudia. University of Edinburgh. 75 Copeland, P., 1997. The when and where of the growth of the Himalaya and the Tibetan Plateau. In: Ruddiman, W.F. (ed) Tectonic Uplift and Climate Change. New York: Plenum Press. Clark, J.R. et al., 2008. A comparative study in ancestral range reconstruction methods: retracing the uncertain histories of insular lineages. Systematic biology, 57(5), pp.693–707. Clement, W.L. et al., 2004. Phylogenetic position and biogeography of Hillebrandia sandwicensis (Begoniaceae): A rare Hawaiian relict. American Journal of Botany, 91(6), pp.905–917. Conti, E. et al., 2002. Early Tertiary out-of-India dispersal of Crypteroniaceae: evidence from phylogeny and molecular dating. Evolution; international journal of organic evolution, 56(10), pp.1931–1942. Crayn, D.M., Costion, C. & Harrington, M.G., 2014. The Sahul – Sunda floristic exchange : dated molecular phylogenies document Cenozoic intercontinental dispersal dynamics. Journal of Biogeography, 42, pp.11–24. De Lange, A. & Bouman, F. 1992. Studies on BegoniaceaeIII. Seed micromorphology of the genus Begonia in Africa: Taxonomic and ecological implications. Wageningen AgriculturalUniversity papers: 1-82. De Wilde, J.J.F.E. et al., 2011. Pliocene intercontinental dispersal from africa to Southeast Asia highlighted by the new species Begoniaafromigrata (Begoniaceae). Taxon, 60, pp.1685–1692. Dewitte, A. et al., 2009a. Genome size variation in Begonia. Genome, 52 (10), pp. 829838. Dewitte, A. et al., 2009b. Occurrence of viable unreduced pollen in a Begonia collection. Euphytica, 168 (1), pp. 81-94. 76 Dewitte, A., Twyford, A.D., Thomas, D.C., Kidner, C.A. & Van Huylenbrock, J., 2011. Begonia: Genome Dynamism, Population Processes and Phylogenetic Patterns. The Dynamical Processes of Biodiversity-Case studies of Evolution and Spatial Distribution. pp 27-52. Doorenbos, J., M. S. M. Sosef, and J. J. F. E. de Wilde . 1998 . The sections of Begonia including descriptions, keys and species lists. Studies in Begoniaceae VI. Wageningen Agricultural University Papers 98-2. Drummond, A.J. & Rambaut, A., 2007. BEAST: Bayesian evolutionary analysis by sampling trees. BMC evolutionary biology, 7, p.214. Feliner, G.N., & Rosello, J.A. 2007. Better the devil you know? Guidelines for insightful utilization of nrDNA ITS in species-level evolutionary studies in plants. Molecular Phylogenetics and Evolution 44:911-919. Forrest, L.L., Hughes, M., and Hollingsworth, P. 2005. A phylo- geny of Begonia using nuclear ribosomal sequence data and morphological characters. Syst. Bot. 30(3): 671–682. Ghebreab, W., 1998. Tectonics of the Red Sea region reassessed. Earth Science Reviews, 45, pp.1–44. Goodall-Copestake, W.P. et al., 2010. The early evolution of the mega-diverse genus Begonia (Begoniaceae) inferred from organelle DNA phylogenies. Biological Journal of the Linnean Society, 101, pp.243–250. Goodall-Copestake, W.P., Harris, D.J. & Hollingsworth, P.M., 2009. The origin of a mega-diverse genus: Dating Begonia (Begoniaceae) using alternative datasets, calibrations and relaxed clock methods. Botanical Journal of the Linnean Society, 159, pp.363–380. Guo, Z. et al., 2004. Late Miocene-Pliocene development of Asian aridification as recorded in the Red-Earth Formation in northern China. Global and Planetary Change, 41(3-4), pp.135–145. 77 Hall, T., 1999. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symposium Series, 41, pp.95–98. Hughes, M. 2002. Population structure and speciation in Begonia L. PhD thesis. University of Glasgow, Glasgow. Hughes, M. et al., 2015. Begonia Resource Centre. Hughes, M. et al., 2015. The Miocene to Pleistocene colonization of the Philippine archipelago by Begonia sect. Baryandra (Begoniaceae). American Journal of Botany, 102, pp.695–706. Hughes, M. & Hollingsworth, P.M., 2008. Population genetic divergence corresponds with species-level biodiversity patterns in the large genus Begonia. Molecular Ecology, 17(11), pp.2643–2651. Janssens, S.B. et al., 2009. Rapid radiation of Impatiens (Balsaminaceae) during Pliocene and Pleistocene: Result of a global climate change. Molecular Phylogenetics and Evolution, 52(3), pp.806–824. Judd, et al., 2008. Plant Systematics: A phylogenetic approach, 3rd ed. Sinauer Associates, Sunderland. Kearse, M. et al., 2012. Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics, 28(12), pp.1647-1649. Kiew, R. 2005. Begonias of Peninsular Malaysia. Natural History Publications (Borneo), Kota Kinabalu. Kodandaramaiah, U., 2010. Use of dispersal-vicariance analysis in biogeography - A critique. Journal of Biogeography, 37, pp.3–11. 78 Kutzbach, J.E., et al., 1989. Sensitivity of climate to Late Cenozoic uplift in South-ern Asia and the American West: numerical experimentsJ. Geophys. Res, 94, pp.18 393–18 407. Lakhanpal, R. N., 1970. Tertiary floras of India and their bearing on the historical geology of the region. Taxon, 19(5), pp.675-694. Lamm, K.S. & Redelings, B.D., 2009. Reconstructing ancestral ranges in historical biogeography: Properties and prospects. Journal of Systematics and Evolution, 47(June), pp.369–382. Landis, M.J. et al., 2013. Bayesian analysis of biogeography when the number of areas is large. Systematic Biology, 62(0), pp.789–804. Legro, RAH. & Doorenbos, J. (1969). Chromosome numbers in Begonia 1. Netherlands Journal of Agricultural Sciences, Vol.17, pp. Linder, C.R., & Rieseberg, L.H. 2004. Reconstructing patterns of reticulate evolution in plants. American Journal of Botany 91: 1700-1708. Lomolino, M.V., Riddle, B.R. & Brown, J.H., 1998. Biogeography, 3rd edn. Sinauer Associates, Sunderland, MA. Maddison, W. P. & Maddison, D.R., 2015. Mesquite: a modular system for evolutionary analysis. Version 3.03 http://mesquiteproject.org. Matolweni, L.O., Balkwill, K., & McLellam, T. 2000. Genetic diversity and gene flow in the morphology variable, rare endemics Begonia dregei and Begonia homonyma (Begoniaceae). American Journal of Botany, 87,pp.431-439. Matzke, N.J., 2014. Model Selection in Historical Biogeography Reveals that FounderEvent Speciation Is a Crucial Process in Island Clades. Systematic Biology, 63(6), pp.951–970. 79 McLoughlin, S., 2001. The breakup history of Gondwana and its impact on preCenozoic floristic provincialism. Australian Journal of Botany, 49(3), pp.271–300. Miller, M.A. et al., 2015. A RESTful API for Access to Phylogenetic Tools via the CIPRES Science Gateway. Evolutionary Bioinformatics, 11, pp,43-48. Molnar, P., Boos, W.R. & Battisti, D.S., 2010. Orographic Controls on Climate and Paleoclimate of Asia: Thermal and Mechanical Roles for the Tibetan Plateau. Annual Review of Earth and Planetary Sciences, 38(1), pp.77–102. Moonlight, P.W. et al., 2015. Continental-scale diversification patterns in a megadiverse genus: the biogeography of Neotropical Begonia. Journal of Biogeography, 42(6), pp.1137–1149. Morley, R. J., 2000. Origin and evolution of tropical rain forests. Chichester: John Wiley & Sons, Ltd. Morley, R.J., 2003. Interplate dispersal paths for megathermal angiosperms. Perspectives in Plant Ecology, Evolution and Systematics, 6(12), pp.5–20. Myers, N. et al., 2000. Biodiversity hotspots for conservation priorities. Nature, 403(6772), pp.853–8. Otto, SP. & Whitton, J. (2000). Polyploid incidence and evolution. Annual Review of Genetics, 34, pp. 401-437. Pennington, R.T. & Dick, C.W., 2004. The role of immigrants in the assembly of the South American rainforest tree flora. Philosophical transactions of the Royal Society of London. Series B, Biological sciences, 359(1450), pp.1611–1622. Pennington, R.T., Hughes, M. & Moonlight, P.W., 2015. The Origins of Tropical Rainforest Hyperdiversity. Trends in Plant Science, 20(11), pp.693–695. Plana, V. (2003) Phylogenetic relationships of the Afro-Mala- gasy members of the large genus Begonia inferred from trnL intron sequences. Systematic Botany, 28, pp.693–704. 80 Posada, D. et al., 2012. jModelTest 2: more models, new heuristics and parallel computing. Nature Methods, 9, pp.772–772. Rajbhandary, S., Hughes, M. & Shrestha, K.K., 2010. Distribution patterns of Begonia in the Nepal Himalaya. Journal of Plant Science, 7(0), pp. 73-78. Rajbhandary, S. et al., 2011. Asian Begonia: out of Africa via the Himalayas? Gardens’ Bulletin Singapore, 63, pp.277–286. Rashid, M.E. & Rahman, M.A., 2012. Updated Nomenclature and taxonomic status of the plants of Bangladesh included in Hook. F., THE FLORA OF BRITISH INDIA: VOLUME-II. Bangladesh J. Plant Taxon, 19(2), pp.173-190. Raven, P. H., & Axelrod, D.I., 1974. Angiosperm biogeography and past continental movements. Ann. Missouri Bot. Gard, 61,pp.539– 673. Rambaut, A., 2009. FigTree, a graphical viewer of phylogenetic trees. Institute of Evolutionary Biology University of Edinburgh. Ree, R.H. et al., 2005. A likelihood framework for inferring the evolution of geographic range on phylogenetic trees. Evolution; international journal of organic evolution, 59(11), pp.2299–2311. Ree, R.H. & Smith, S. A, 2008. Maximum likelihood inference of geographic range evolution by dispersal, local extinction, and cladogenesis. Systematic biology, 57(1), pp.4–14. Renner, S.S. 2006. Rewardless flowers in the angiosperms and the role of insect cognition in their evolution, Plant pollinator interactions: from specialization to generalization. University of Chicago Press, Chicago. Ronquist, F., 1997. Dispersal-vicariance analysis: A new approach to the quantification of historical biogeography. Systematic Biology, 46(l), pp.195–203. Ronquist, F. et al., 2012. MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space. Systematic Biology, 61, pp.539–42. 81 Ronquist, F. & Sanmartín, I., 2011. Phylogenetic Methods in Biogeography. Annual Review of Ecology, Evolution, and Systematics, 42(1), pp.441–464. Rutschmann, F. & Eriksson, T., 2004. Did Crypteroniaceae really disperse out of India? Molecular dating evidence from rbcL, ndhF, and rpl16 intron sequences. International Journal of plant Science, 165(S4), pp.S69–S89. Schemske, D.W., et al. 1996. Decceit pollination in the monoecious, Neotropical herb Begonia oaxacana (Begoniaceae). Floral biology: Studies on floral evolution in animal-pollinated plants: 292-318. Scotese, C.R. 2003. PALAEOMAP Project. http://www.scotese.com/. Small, R.L., Cronn, R.C., & Wendel, J.F. 2004. Use of nuclear genes for phylogeny reconstructions in plants. Australian Systematic Botany 17, 145-170. Stults , D. Z. , &Axsmith, B.J., 2011. First macrofossil record of Begonia (Begoniaceae). American Journal of Botany 98 : 150 – 153 . Stamatakis, A., 2006. RAxML-VI-HPC: Maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics, 22(21), pp.2688–2690. Tebbitt, MC.; Lowe-Forrest, LL.; Santoriello, A.; Clement, WL. & Swensen, SM. 2006. Phylogenetic relationships of Asian Begonia, with an emphasis on the evolution of rain-ballist and animal dispersal mechanisms in sections Platycentrum, Sphenanthera and Leprosae. Systematic botany, 31(2), pp. 327-336. Thomas, D.C., 2010. Phylogenetics and historical biogeography of Southeast Asian Begonia L. (Begoniaceae). PhD thesis. University of Glasgow, Glasgow. Thomas,D.C.,Hughes,M.,Phutthai, T.,Rajbhandary,S.,Rubite, R., Ardi, W.H. & Richardson, J.E., 2011a. A non-coding plastid DNA phylogeny of Asian Begonia (Begoniaceae): evidence for morphological homoplasy and sectional polyphyly. Molecular Phylogenetics and Evolution, 60, pp. 428–444. 82 Thomas, D.C. et al., 2012. West to east dispersal and subsequent rapid diversification of the mega-diverse genus Begonia (Begoniaceae) in the Malesian archipelago. Journal of Biogeography, 39(1), pp.98–113. Yuan, Y.-M. et al., 2005. Phylogeny and biogeography of exacum (gentianaceae): a disjunctive distribution in the Indian ocean basin resulted from long distance dispersal and extensive radiation. Systematic biology, 54(1), pp.21–34. Zachos, J. et al., 2001. Trends, rhythms, and aberrations in global climate 65 Ma to present. Science (New York, N.Y.), 292(April), pp.686–693. Zheng, H. et al., 2004. Late Miocene and mid-Pliocene enhancement of the East Asian monsoon as viewed from the land and sea. Global and Planetary Change, 41(3-4), pp.147–155. 83 CHAPTER 4: Biogeographic history of Sri Lankan Sapotaceae Introduction Sapotaceae are a pantropical tree family, mostly distributed in tropical rain forests, but some extending to semi-arid and arid regions. The highest species richness is recorded in the tropical and subtropical regions of Asia and South America (Swenson et al. 2014). The family is composed of 53 genera and about 1100 species (Pennington 1991; Swenson et al. 2014; Anderberg 2005; Bartish et al. 2005; Bartish et al. 2011) It has a wide economic utility such as timber (Manilakara bidentata, M. huberi, M. obovata and M. kauki ), chewing gum (Manilkara zapota), edible fruit (sapodilla plum, Manilkara zapota; star apple, Chrysophyllum cainito) and in the cosmetics industry; Vitellaria paradoxa (Pennington 1991; Govaerts et al.2001; Mathews 2009). The most recent classification by Pennington (1991) divides the family into three subfamilies Chrysophylloideae, Sapotoideae, and Sarcospermatoideae and five tribes: Chrysophylleae, Isonandreae, Omphalocarpeae, Mimusopeae, and Sideroxyleae. Sarcospermatoideae is a small subfamily with a narrow distribution in South and Southeast Asia. It comprises of Sarcosperma and possibly Eberhardtia (Smedmark et al. 2006), two small genera which are sisters to the large subfamilies Chrysophylloideae and Sapotoideae. Chrysophylloideae is composed of 600 species of shrubs and medium understory to giant canopy trees of rainforests in Africa, Australasia, and South America, although the numbers are likely to increase as a result of ongoing phylogenetic and taxonomic research (Swenson et al. 2008). Sapotoideae is a pantropical subfamily comprising 543 species with the highest species richness recorded in Indo Pacific region and the rest occurring in Africa and America (Smedmark et al. 2006). Higher taxa within Sapotaceae have overlapping morphological variation making the generic delimitation within the problematic. The Tribe Mimusopeae is distinguished by having tripartite corolla lobes, petaloid staminodes and a basal-basiventral seed scar, which is further divided in to three subtribes; Mimusopinae, Manilkarinae and Glueminae (Pennington 1991). 84 Several chloroplast and nuclear markers have been widely utilised in the phylogeny and classification of Sapotaceae. The chloroplast region ndhF was utilized in the construction of the first Sapotaceae phylogeny; however the monophyly of the tribe Mimusopeae was not supported due to lack of variability with the use of a single chloroplast marker (Anderberg & Swenson 2003). However, later with the incorporation of morphological data a more resolved phylogeny was obtained and the family was divided into three subfamilies; Sarcospermatoideae, Sapotoideae and Chrysophylloideae. Sapotaceae were confirmed to have evolved in two separate lineages; Isonandrae-Mimusopeae-Syderoxyleae and Chrysophylleae-Omphalocarpeae. Among these tribes, Sideroxyleae are monophyletic, Isonandreae are polyphyletic while Mimusopeae are paraphyletic (Swenson & Anderberg 2005) Later, additional chloroplast non-coding intergenic spacers; trnH-psbA, trnC-TrnD, trnC-psbM, psbM-trnD were utilized by Smedmark et al. (2006) in an analysis of subfamilial level relationships in Sapotoideae. Eberhardtia aurata is the sister for two major sub clades; Sideroxyleae and Sapoteae and within Sapotoideae, Sideroxyleae and Sapoteae, each were strongly supported as monophyletic. Polyphyly of the Isonandrae was confirmed; however it was poorly supported due to lack of sampling in the phylogeny (Smedmark et al. 2006). A phylogeny of New Caledonian Sapotaceae was constructed by Bartish et al. (2005) using nuclear ribosomal DNA and the results conflicted with existing phylogenies. Some genera like Pouteria, Niemeyerawere found to be non monophyletic while unrelated genera confined to Australia, New Caledonia and neighbouring islands were clustered in single clades (Bartish et al. 2005). This non monophyly was further confirmed by Swenson et al. 2007, with the addition morphological data and increased taxon sampling. Nuclear DNA provided useful information on species level relationships, however fail to obtain relationships at deeper nodes (Swenson et al. 2007). Later, phylogenetic relationships within theNiemeyera complex and generic delimitations were investigated by Swenson et al. (2008) by utilizing both nuclear and chloroplast markers. Nuclear internal transcribed data (ITS) were combined with ETS 85 (External Transcribed Spacers) which resulted in better resolution (Swenson et al. 2008). The intergenic spacers; trnDGUC–trnTGGU, trnSGCU–trnGUUC, rpoB–trnCGCA and trnSUGA–trnfMCAU and trnH-psbA were utilized in the same study, however they resulted in a poorly resolved phylogeny. Among the chloroplast markers tested, trnHpsbA and trnS-trnG proven to be useful in resolving phylogenetic relationships at low levels whereas rpoB-trnC and trnD-trnT showed less variability, thus less informative. All the recent work based on morphological and different molecular markers (both nuclear and chloroplast data) reveal the need of a new classifications for Sapotaceae (Swenson et al. 2008) Throughout the past decades, biogeographers have struggled to explain tropical intercontinental disjunctions and vicariance was accepted as the major driving force for many years with examples cited in families like Dipterocarpaceae (Ashton & Gunatilleke 1987) and Monimiaceae (Renner et al. 2010). However, with the development of dated phylogenies and ancestral area reconstruction methods, many tropical families such as Begoniaceae (Thomas et al. 2012) have been shown to have a more recent origin demonstrating the importance of long distance dispersal in determining distributions of plant genera. This has also been demonstrated in many other groups, such as Araucaria and Nothofagus (Setoguchi et al. 1998; Swenson et al. 2001; Cook & Crisp 2005). The family Sapotaceae is an ideal candidate to study the biogeographic history of tropical families, with its pantropical distribution and the fact that it occupies various habitats. Several biogeographic analysis have been carried out for different groups within the family (Chrysophylloideae; Bartish et al. 2011, Isonandreae; Richardson et al. 2014, Manilkara; Armstrong et al. 2015), however a wider biogeographic analysis across the family has not been carried out so far. The dated phylogeny based on previously published data by Swenson et al.(2008) demonstrated the an early diversification sub-family Chrysophylloideae in Africa during the late Cretaceous approximately 73-83 Ma (Bartish et al. 2011). A single vicariance event between South America and Australia is postulated. Multiple long distance 86 dispersals from Africa to the Neotropics, between Australia and New Caledonia and between Africa and Madagascar and a single dispersal back to Africa from the Neotropics are inferred, thus long dispersal events are considered as the major driving force for current disjunctions of Chrysophylloideae across the tropics (Bartish et al. 2011). Smedmark & Anderberg (2007) suggest a northern hemisphere origin for Sideroxyleae in the early tertiary followed by extensive extinctions due to increasingly unfavourable climatic conditions, specifically decreases in temperature. The distributions of extant species in the new world may have been facilitated via the North Atlantic Land Bridge during the early Eocene and this is supported by the occurrence of Sapotaceae pollen in North America and Europe (Smedmark & Anderberg 2007; Bartish et al. 2011; Morley 2000). The pantropical distribution of Manilkara (Sapotoideae) has also been shown, through the generation of a dated phylogeny, to be the result of long distance dispersal events rather than vicariance caused by continental break-up (Armstrong et al. 2014). Nuclear ITS, plastid rpl32-trnL, rpl16-trnK and trnS-trnF were utilized in the study, however plastid data were not incorporated in the BEAST analysis due to less variability and less sampling (Armstrong et al 2014) Sri Lanka is a key locality for understanding patterns in tropical plant biogeography. It is part of the Indian continental plate that was part of Gondwana. In Sri Lanka, Sapotaceae contains 27 species from all tribes and 25 species occur in the low or upland wet zone. The highest number of species is recorded in the tribe Isonandreae with species from Palaquium, Madhuca, Isonandra, Payena and Diploknema. All the nine species of Palaquium, five out of seven species of Madhuca and three out of four Isonandra present in the country are endemic thus making it an important family to look for phylogenetic and biogeographic relationships of the Sri Lankan flora. The Isonandreae are also an ideal group to study diversification patterns in continental Asia and South East Asia due to the diversity in India, Indochina, Sunda, Australia and Pacific islands (Richardson et al. 2014). The purpose of this study is to determine the age and biogeographic affinities of Sri Lankan representatives of Sapotaceae by placing them in a global phylogenetic context. 87 I aimed to test the following hypothesis on origin and the diversification of Sri Lankan flora. The breakup of the supercontinent Gondwana initiated during early Jurassic 180 million years ago (Mya), followed by subsequent continental drifts and sea floor spreading, was considered one of the major causes for tropical plant disjunctions we observe today. The Deccan plate comprised of India and Sri Lanka was a part of Gondwana that began to separate from Antarctica from 132 Mya followed by northward migration until collision with Asia c. 35 Mya. Being a part of Gondwana, the Sri Lankan flora might be composed of Gondwanan relicts (Ashton & Gunatilleke 1987) . However to be of Gondwana origin the disjunction should be reflected in phylogenetic splits occurring approximately 132 Mya-94 Mya, concurrent with continental break-up. The collision of Deccan plate with southern Laurasia from c. 35 Mya is accepted as the most marked possibility for interplate dispersal during the Tertiary (Morley 2000). Thus, the Sri Lankan species could also have arrived via the immigration of Laurasian lineages through Asia and India, resulting in the mixing of the Deccan Gondwanan flora with Laurasian lineages. For this hypothesis to be accepted, we would expect that divergence times for lineages should not be older than 35 Mya and that Indo-Sri Lankan lineages would be nested within Laurasian ones. Thirdly, long distance dispersal could have played a major role in the assemblage of the Sri Lanka flora from other land masses such as Africa and/or South East Asia. Longdistance dispersal could have occurred at any point in time, it is the only viable scenario for tropical disjunctions younger than ~33 Mya and Indo-Sri Lankan lineages would be nested within African and/or South East Asian ones. Finally, a potentially important contributor to the Sri Lankan flora will have been diversification of lineages within the island (in situ speciation), resulting in the evolution of new species. Species endemic to Sri Lanka with sister species from Sri Lanka are considered as representing in situ speciation. More recent PliocenePleistocene speciation may have resulted from changing climatic conditions during glacial/inter-glacial periods. 88 Thus, I aimed to estimate the relative contributions to the Sri Lankan flora of Gondwanan relicts, overland immigration from nearby landmasses (Asia), long distance trans-oceanic dispersal from Southeast Asia, and long distance dispersal from Africa and in situ speciation with the use of a representative plant family Sapotaceae. Spatial relationships within Isonandreae from Sri Lanka were examined with a fine scale geographic coding for the Asian Sapotaceae in order to obtain a better understanding of biogeography within the continental Asia which was not addressed in previous Sapotaceae studies. Materials and methods Plant material and Taxon sampling Both nuclear and chloroplast data were generated in my study. The ITS data set comprised 163 accessions and chloroplast data set comprised 136 accessions with species from all tribes of Sapotaceae. The sequences of a selection of taxa representing tribes Sideroxyleae and Mimusopeae were included in order to provide fossil calibration points within each tribe (Richardson et al. 2014; Armstrong et al. 2014). Sarcosperma was used as the outgroup, as it is the sister to other lineages of Sapotaceae in previous studies (Anderberg & Swenson 2003). In order to complement South Asian elements lacking in previous studies, 24 species from Sri Lanka (100% Sri Lankan representatives) were added representing five genera Palaquium, Isonandra, Madhuca, Manilkara, Mimusops. Twenty four sequences were generated for the ITS matrix. The chloroplast matrix comprised 136 sequences and 86 of these sequences were newly generated (23 Sri Lankan representatives+ 63DNA extractions from EDNA bank) for trnH-psbA, trnC-D and ndhF. Voucher specimens of the newly generated ITS sequences and chloroplast sequences are listed in table 4.1 and table 4.2 respectively. Voucher information of all the samples used in the present study are listed in appendix 3. 89 Table 4.1 Voucher specimens of the newly generated ITS sequences EDNA number Species Origin Voucher EDNA13-0032656 Palaquium_laevifolium_SL Engl. Sri Lanka Kumarage 60 (E) EDNA13-0032663 Isonandra_lanceolata_SL Wight Brunei, Indonesia, Borneo, Sri Lanka Kumarage 47 (E) EDNA13-0032654 Madhuca_sp_SL Sri Lanka Kumarage 58 (E) EDNA13-0032657 Madhuca_fulva_SL J.F.Macber Sri Lanka Kumarage 62 (E) EDNA13-0032658 Madhuca_neriifolia_SL H.J.Lam Sri Lanka Kumarage 63 (E) EDNA13-0032661 Madhuca_longifolia2_SL Sri Lanka Photo voucher J.F.Macber available EDNA13-0032648 Palaquium_thwaitesii_SL Trim. Sri Lanka Kumarage 43 (E) EDNA13-0032650 Palaquium_canaliculatum_SL Sri Lanka Kumarage 45 (E) Sri Lanka Kumarage 64 (E) Sri Lanka Kumarage 59 (E) Engl. EDNA13-0032653 Palaquium_hinmolpedda_SL P.Royen EDNA13-0032655 Palaquium_pauciflorum_SL Engl. EDNA13-0032660 Palaquium_rubiginosum_SLEngl. Sri Lanka Kumarage 65 (E) EDNA14-003582 Isonandra_montana_SL Gamble Sri Lanka Kumarage 76 (E) EDNA14-003579 Isonandra_sp1_SL Sri Lanka Kumarage 97 (E) EDNA14-003580 Palaquium_sp3_SL Sri Lanka Kumarage 57 (E) EDNA13-0032651 Palaquium_grande_SL Engl. Sri Lanka Kumarage 46 (E) EDNA13-0032649 Palaquium_petiolare_SL Engl. Sri Lanka Kumarage 44 (E) EDNA14-003581 Palaquium_sp5_SL Sri Lanka Kumarage 93 (E) EDNA14-003578 Isonandra_sp2_SL Sri Lanka Kumarage 74 (E) EDNA13-0032659 Isonandra_compta_SL Dubard Sri Lanka Emanuelsson 3039(S) EDNA14-003583 Isonandra_zeylanica_SL Sri Lanka Kumarage 72 (E) EDNA13-0032662 Manilkara_hexandra_SL Dubard Sri Lanka Photo voucher available 90 EDNA14-003584 Mimusops_elengi_SL L. Sri Lanka Kumarage 98 (E) EDNA14-003585 Madhuca_clavata_SL Jayas. Sri Lanka Kumarage 99 (E) EDNA14-003577 Palaquium_sp4_SL Sri Lanka Kumarage 75 (E) Table 4.2 Voucher information for newly generated chloroplast sequences (trnH-psbA, trnC-trnD, ndhF) EDNA number Species name Origin Voucher EDNA09-02183 Madhuca elmeri Merr. Ex. H.J.Lam Indonesia, Kalimantan Wilkie P1 347 (E) EDNA09-02184 Madhuca laurifolia H.J.Lam Peninsular Malaysia Wilkie 843 (E) EDNA09-02185 Madhuca motleyana Baehni Peninsular Malaysia Wilkie 837 (E) EDNA09-02303 Palaquium rostratum 2 Burck Indonesia, Kalimantan Slik CMF9452 (L) EDNA09-02304 Palaquium sericeum H.J.Lam Indonesia, Kalimantan Slik CMF9737 (L) EDNA09-02305 Payena maingayi 2 C.B.Clarke Peninsular Malaysia Wilkie 841 (E) EDNA09-02306 Payena lucida 2A.d.Candolle Borneo Ambri et al. AA1604 (L) EDNA09-02307 Madhuca sp. Peninsular Malaysia Wilkie 834 (E) EDNA09-02309 Payena lucida 3A.d.Candolle Peninsular Malaysia Wilkie 845 (E) EDNA09-02312 Palaquium microphyllum King & Gamble Malaysia Pennington, Kochummen & Wong (K) EDNA09-02313 Payena obscura Burck Peninsular Malaysia Wilkie 880 (E) EDNA09-02316 Madhuca kingiana 2 H.J.Lam Malaysia, Sarawak Wilkie 856 (E) EDNA09-02317 Payena leerii Kurz Malaysia Wilkie 811 (E) EDNA09-02318 Palaquium obovatum H.J.Lam Thailand Middleton 4387 (E) EDNA09-02319 Palaquium sp. Solomon Islands Poulsen et al. 2488 (E) EDNA09-02320 Palaquium sp.2 Indonesia, Sulawesi Armstrong 370 (E) EDNA09-02325 Madhuca sp. Peninsular Malaysia Wilkie 834 (E) EDNA09-01452 Palaquium sericeum H.J.Lam Indonesia, Kalimantan Slik CMF9737 (L) 91 EDNA09-00716 Palaquium clarkeanum King & Gamble Malaysia Wilkie 501 (E) EDNA09-00717 Palaquium gutta Baillon Malaysia Wilkie 504 (E) EDNA09-00718 Palaquium obovatum Engler Thailand Middleton 4387 (E) EDNA09-00719 Palaquium oxleyanum Pierre Malaysia Wilkie 527 (E) EDNA09-00721 Palaquium xanthochymum Pierre ex Burck Malaysia, Borneo Wilkie 544 (E) EDNA09-00722 Palaquium formosanum Hayata Taiwan,Philippines Chung & Anderberg 1421 (HAST) EDNA09-00723 Palaquium amboinense 3 Burck Indonesia (native to New Guinea) Wilkie 813 (E) EDNA09-00952 Palaquium beccarianum van Royen Malaysia,Papua, Borneo Wijesundara s.n. (K) EDNA09-00953 Payena leerii Kurz Malaysia Wilkie 811 (E) EDNA09-00954 Madhuca kingiana 1 H.J.Lam Malaysia, Sarawak Wilkie 856 (E) EDNA09-00955 Palaquium sumatranum Burck Indonesia, Java Wilkie 823 (E) EDNA09-00956 Palaquium rostratum 1 Burck Indonesia, Kalimantan Slik CMF9452 (L) EDNA09-00957 Palaquium maingayi King & Gamble Peninsular Malaysia Wilkie 841 (E) EDNA09-00958 Madhuca malaccensis H.J.Lam Peninsular Malaysia Wilkie 832 (E) EDNA09-00959 Payena obscura Burck Peninsular Malaysia Wilkie 880 (E) EDNA09-00987 Palaquium galactoxylum H.J.Lam Australia Bartish and Jessup 9 (S) EDNA09-00990 Pouteria maclayana Baehni Indonesia,Malaysia,PN G Armstrong 316 (E) EDNA09-00992 Palaquium calophyllum Pierre ex Burck Indonesia, Kalimantan Wilkie et al 1/477 (E) EDNA09-00993 Palaquium eriocalyx H.J.Lam Indonesia Wilkie et al. 8/147 (E) EDNA09-01129 Palaquium rigidum Pierre ex Dubard Malaysia Wilkie 878 (E) EDNA09-01130 Palaquium leiocarpum 1 Boeriage Malaysia, Sarawak Wilkie 870 (E) 92 EDNA09-01131 Palaquium pseudorostratum H.J.Lam Malaysia, Sarawak Wilkie 857 (E) EDNA09-01132 Palaquium ridleyi King & Gamble Malaysia, Sarawak Wilkie 858 (E) EDNA09-01133 Palaquium rufolanigerum P.Royen Malaysia, Sarawak Wilkie 859 (E) EDNA09-01134 Madhuca kuchingensis Yii & P.Chai Malaysia, Sarawak Wilkie 860 (E) EDNA09-01135 Palaquium walsurifolium Pierre ex Dubard Malaysia, Sarawak Wilkie 877 (E) EDNA09-01136 Madhuca sarawakensisH.J.Lam Sarawak, Malaysia Wilkie 863 (E) EDNA09-01137 Madhuca erythrophylla H.J.Lam Sarawak, Malaysia Wilkie 867 (E) EDNA09-01138 Madhuca kunstleri H.J.Lam Malaysia, Sarawak Wilkie 868 (E) EDNA09-01140 Palaquium calophyllum Pierre ex Burck Indonesia, Kalimantan Wilkie et al 1/477 (E) EDNA09-01141 Palaquium herveyi King & Gamble Malaysia, Sarawak Wilkie 871 (E) EDNA09-01142 Palaquium hexandrum Baillon Malaysia, Sarawak Wilkie 872 (E) EDNA09-01143 Madhuca barbata T.D.Penn. Sarawak, Malaysia Wilkie 873 (E) EDNA09-01144 Palaquium cryptocariifolium P.Royen Malaysia, Sarawak Wilkie 874 (E) EDNA09-01378 Madhuca sericea H.J.Lam Sarawak, Malaysia Wilkie 879 (E) EDNA09-01379 Madhuca oblongifolia Merrill Sarawak, Malaysia Wilkie 861 (E) EDNA09-01380 Madhuca korthalsii H.J.Lam Sarawak, Malaysia Wilkie 876 (E) EDNA09-01381 Madhuca proxila Yii & P.Chai Sarawak, Malaysia Wilkie 875 (E) EDNA09-01382 Madhuca lancifolia H.J.Lam Sarawak, Malaysia Wilkie 853 (E) EDNA09-01383 Madhuca sp. nov Peninsular Malaysia Wilkie 834 (E) EDNA09-01384 Palaquium amboinense 2 Burck Indonesia (native to New Guinea) Wilkie 813 (E) EDNA09-01386 Madhuca pachyphylla (K.Krause) ined. Indonesia, West Papua Armstrong 313 (E) EDNA09-01388 Burckella polymera P.Royen Indonesia, West Papua Armstrong 326 (E) 93 EDNA09-01389 Pouteria firma (Miq.) Baehni Indonesia, West Papua Armstrong 305 (E) EDNA09-01451 Palaquium quercifolium Burck Indonesia, Kalimantan Slik CMF6780 (L) EDNA130032656 Palaquium_laevifolium_SL Sri Lanka Kumarage 60 (E) EDNA130032663 Isonandra_lanceolata_SL Wight Brunei, Indonesia, borneo, Sri Lanka Kumarage 47 (E) EDNA130032654 Madhuca_sp_SL Sri Lanka Kumarage 58 (E) EDNA130032658 Madhuca_neriifolia_SL H.J.Lam Sri Lanka Kumarage 63 (E) EDNA130032661 Madhuca_longifolia2_SL Sri Lanka Photo voucher EDNA130032648 Palaquium_thwaitesii_SL Trim. Sri Lanka Kumarage 43 (E) EDNA130032650 Palaquium_canaliculatum_SL Sri Lanka Kumarage 45 (E) EDNA130032653 Palaquium_hinmolpedda_SL Sri Lanka Kumarage 64 (E) EDNA130032655 Palaquium_pauciflorum_SL Engl. Sri Lanka Kumarage 59 (E) EDNA130032660 Palaquium_rubiginosum_SL Sri Lanka Kumarage 65 (E) EDNA14-003582 Isonandra_montana_SL Gamble Sri Lanka Kumarage 76 (E) EDNA14-003579 Isonandra_sp1_SL Sri Lanka Kumarage 97 (E) EDNA14-003580 Palaquium_sp3_SL Sri Lanka Kumarage 57 (E) EDNA130032651 Palaquium_grande_SL Engl. Sri Lanka Kumarage 46 (E) EDNA130032649 Palaquium_petiolare_SL Engl. Sri Lanka Kumarage 44 (E) EDNA14-003581 Palaquium_sp5_SL Sri Lanka Kumarage 93 (E) EDNA14-003578 Isonandra_sp2_SL Sri Lanka Kumarage 74 (E) EDNA130032659 Isonandra_compta_SL Dubard Sri Lanka Emanuelsson 3039 EDNA14-003583 Isonandra_zeylanica_SL Jeuken J.F.macber available Engl. P.Royen Engl. (S) 94 Sri Lanka Kumarage 72 (E) EDNA130032662 Manilkara_hexandra_SL Dubard EDNA14-003584 Mimusops_elengi_SL L. Sri Lanka Kumarage 98 (E) EDNA14-003585 Madhuca_clavata_SL Jayas. Sri Lanka Kumarage 99 (E) EDNA14-003577 Palaquium_sp4_SL Sri Lanka Kumarage 75 (E) Sri Lanka Photo voucher available DNA extraction, amplification and sequencing Total genomic data was extracted from silica gel dried material using the DNeasy Plant Mini Kit (Qiagen, UK) according to the manufacturer’s protocols. 25 μl PCR reactions were setup for amplification of both ITS and chloroplast regions. Nuclear ITS rDNA was amplified using ITS 5p and ITS 8p. Each 25μl of PCR mixture contained 5.75μl of ddH2O, 2.5μl of 10x reaction buffer, 1.25μl of 25mM MgCl2, 2.5 μl of 2mM dNTPs, 0.75μl of 10μM forward primer (ITS5p), 0.75μl of 10μM reverse primer (ITS8p), 10μl Betain, 0.25 μl of 0.4% BSA, 0.25μl of Biotaq DNA polymerase (Bioline, UK) and 1μl of DNA template. The PCR temperature profile was, template denaturation at 95oC for 5 min followed by 34 cycles of denaturation at 95oC for 30 Sec, primer annealing at 50oC for 30 sec, primer extension at 72oC for 1.5 min followed by a final extension step at 72oC for 8 min. Chloroplast data were obtained for three regions. The trnH-psbA spacer was amplified using the primers described by Hamilton (1999). The trnC-trnD region (consisting of the trnC- petN spacer, the petN gene, the petN-psbM spacer, the psbM gene, and the psbM-trnD spacer) was amplified in two segments; the trnC-psbM region with the trnC (Demesure et al. 1995) and psbM2R (Lee & Wen 2004) primers, and the psbM-trnD spacer with the psbM1 (Lee & Wen 2004) and trnD (Demesure et al. 1995) primers. The 3´end of ndhF was amplified with primers ndhF5 and ndhF10 (Olmstead & Sweere 1994). The primer sequences are listed in Table 4.3. For the trnH-psbA region, 25μl of PCR mixture contained 15.25μl of ddH2O, 2.5μl of 10x reaction buffer, 1.25μl of 25mM MgCl2, 2.5 μl of 2mM dNTPs, 0.75μl of 10μM forward primer, 0.75μl of 10μM reverse primer, 0.8μl of 0.4% BSA, 0.2μl of Biotaq DNA polymerase (Bioline, UK) and 1μl of DNA template. The temperature profile 95 included an initial template denaturation step of 96oC for 5 minutes, followed by 34cycles of denaturation at 96oC for 45 s, primer annealing at 53oC for 1 min, primer extension at 72oC for 30 sec and a final extension step at 72oC for 5 min For the ndhF region, 25μl of PCR mixture contained 15.55μl of ddH2O, 2.5μl of 10x reaction buffer, 1.25μl of 25mM MgCl2, 1 μl of 2mM dNTPs, 1 μl of 10μM forward primer (ndhF5), 1μl of 10μM reverse primer (ndhF10), 1.5μl of 0.4% BSA, 0.2μl of Biotaq DNA polymerase (Bioline, UK) and 1μl of DNA template. The temperature profile included an initial template denaturation step of 95oC for 5 minutes, followed by 29 cycles of denaturation at 95oC for 45 s, primer annealing at 51oC for 45 sec, primer extension at 72oC for 3 min and a final extension step at 72oC for 10 min. For the trnC-D region (consisting of the trnC–petN spacer, the petN gene, the petN– psbM spacer, the psbM gene, and the psbM–trnD spacer), 25μl of PCR mixture contained 15.25μl of ddH2O, 2.5μl of 10x reaction buffer, 1.25μlof 25mM MgCl2, 2.5μl of 2mM dNTPs, 0.75μl of 10μM forward primer, 0.75μl of 10μM reverse primer, 0.8μl of 0.4% BSA, 0.2 μl of Biotaq DNA polymerase (Bioline, UK) and 1μl of DNA template. The temperature profile included an initial template denaturation step of 94oC for 5 minutes, followed by 34 cycles of denaturation at 94oC for 1min, primer annealing at 52oC for 2 min, primer extension at 72oC for 2 min and a final extension step at 72oC for 2 min. Table 4.3 Nuclear (ITS) and chloroplast primer sequences used in Sapotaceae DNA sequence generation. Region Primer name Primer sequences References ITS ITS5p GGAAGGAGAAGTCGTAACAAG Moeller & Cronk (1997) ITS8p CACGCTTCTCCAGACTACA Moeller & Cronk (1997) trnH- trnH ACTGCCTTGATCCACTTGGC Hamilton (1999) psbA psbA CGAAGCTCCATCTACAAATGG Hamilton (1999) trnC- trnC CCAGTTCAAATCTGGGTGTC Demesure et al. (1995) psbM psbM2R TTCTTGCATTTATTGCTACTGC Lee & Wen (2004) psbM- psbM1 GCGGTAGGAACTAGAATAAATAG Lee & Wen (2004) trnD trnD GGGATTGTAGTTCAATTGGT Demesure et al. (1995) ndhF ndhF5 GTCTCAATTGGGTTATATG Olmstead et al. (1994) ndhF10 CCCCCTA(CT)ATATTTGATACCTTCTC Olmstead et al. (1994) 96 Amplified products were run on a 1% Agarose TBE gel with Syber safe as the staining agent and visualized in UV transiluminator. The PCR purification was done using EXOSAP IT as 7 μl reaction mixtures. 5μl of PCR product was mixed with 2μl of EXOSAP and incubated at 37oC for 15 minutes followed by 80o C for 15 minutes. Sequencing PCR was done using purified PCR products as 10μlmixtures using 5.68μl of ddH2O, 2μl of sequencing buffer, 0.32μl of primer, 1μl of Big dye and 1μl of template. The sequencing PCR protocol was, denaturation at 95oC for 30 sec, followed by 24 cycles of primer annealing at 50oC for 20sec, extension at 60oC for 4 min. Separate forward and reverse sequencing PCR were carried out and products were sent to the Genepool facility at the University of Edinburgh (Genepool, UK) for BigDye Terminator Cycle sequencing. Sequence editing and alignment Newly generated DNA sequences were edited in GeneiousR7 (7.1.4) (Kearse et al. 2012) and aligned manually in BioEdit 7.1.3 (Hall 1999) and checked for indels. Some regions were excluded from the ITS dataset due to ambiguous alignments or missing data at the region ends; 105-115, 156-163, 229-233, 335-344, 795-796, 803-807, and 1025-1036. Ambiguous alignment regions 1330-1350, 1770-1796, 4012-4026 and 5445-5463 were excluded from the chloroplast dataset. Phylogenetic analysis Both chloroplast and nuclear datasets were analyzed separately under Bayesian inference (BI) using MrBayes 3.2.1 (Ronquist et al. 2012) in the CIPRES science gateway V. 3.3 (Miller et al. 2010) and treated as single partitions. MCMC runs were carried out for 10,000,000 generations and sampled every 1000 generations. A 25% burn-in was set to discard the first set of trees and the remaining trees were summarised as a 50% majority rule consensus tree, visualized in FigTree (Rambaut 2009)and checked for hard incongruence between ITS and chloroplast trees. Plastid data did not provide a fully resolved tree and could not be combined with the ITS data due to well supported incongruence. Further, the ITS data gives more resolution than the chloroplast data and combination with exsisting ITS data comprising a high species number allowed better taxon sampling. Lower sample size in the plastid data would 97 reduce the reliability of biogeographic reconstructions; thus only the ITS data were used in the divergence time estimation and ancestral area reconstructions. Models of sequence evolution were determined using jModeltest 2.1.3 (Posada et al. 2012). Maximum likelihood topologies were used to estimate the optimal evolutionary model and twenty four models were tested under the Akaike Information Criterion (AICc) and the Bayesian Information Criterion (BIC). Bayesian divergence time estimation Bayesian divergence time estimation was performed for nuclear ITSDNA alignment using BEAST v.1.8.0 (Drummond & Rambaut 2007). GTR was selected as the nucleotide substitution model with a gamma distribution of rates among sites and a proportion of invariant sites. The data set was treated as a single partition and uncorrelated relaxed lognormal clock model was selected in order to relax the assumption of a molecular clock and allow for rate heterogeneity between lineages. The tree prior was set to random birth death speciation process, with a randomly generated starting tree. Four separate Markov Chain Monte Carlo (MCMC) runs were carried out for 10,000,000 generations, sampling every 1000 generations, under a Birth-Death model of speciation. Plots of the logged parameters for each run were visualised using Tracer v.1.5 (Drummond & Rambaut 2007) to confirm convergence between runs. Time series plots of all parameters were analyzed in Tracer v.1.5 (Drummond & Rambaut 2007) to check for convergence and to confirm adequate effective sample sizes (ESSs). Trees were combined in LOGCOMBINER (Drummond & Rambaut 2007) with the burn-in set to 25% for initial sample for each run and a single maximum clade credibility (MCC) tree was obtained from Tree Annotator v.1.7.5 (Drummond & Rambaut 2007) and visualized in fig tree v.1.4.0 (Rambaut 2009). The tree was calibrated using fossils at three nodes. Sideroxyleae pollen from the early Eocene of England dated at 47.8-56 Mya (Gruas-Cavagnetto 1976) was used to constrain the minimum age of the Sideroxyleae crown node. One of the criticisms of fossil based calibrations is that they only provide minimum age estimates, as the fossil could have been formed after the age for which the clade it represents was formed. By 98 constraining the crown node instead of the stem node I bias in favour of older age estimates. A log normal prior was used to constrain the age, with an offset of 52.2 Mya and a mean of 0.001. A Mid-Eocene (37.2–48.6Mya) Tetracolporpollenites pollen grain from the Isle of Wight was used to constrain the age of the crown node of the tribe Mimusopeae. This pollen grain closely resembles Tieghemella heckelii, a monotypic genus in Mimusopeae, and was used to constrain the age of the tribe Mimusopeae with an offset of 42.9 and mean of 0.095 so that the 95% probability limits lie within the midpoint of 42.9 Mya and the upper boundary of the mid Eocene (48.6 Mya).The final calibration point is based on a series of Oligocene (23–33.9 Mya) fossil leaves from Ethiopia (Jacobs et al. 2005) which was placed at the Manilkara crown node with an offset of 28 Mya and mean of 0.1. Area delimitation and Ancestral Area Reconstructions Eleven areas were coded based on the extant distribution, areas of endemism and geological history. In order to obtain a clear picture of South Asian biogeography, India, Sri Lanka and Himalaya were coded as separate geographic areas. East Asian elements were composed of an area East of Himalaya combined with China and Indochina. The Middle East element was composed of Iran, Turkey and the Arabian Peninsula which belongs to the Irano-Turanian flora (Takhtajan 1986).In Southeast Asia, the Sahul and Sunda Shelves were coded as separate states within the Malesian floristic region, which stretches from the Isthmus of Kra on the Malay Peninsula to Fiji (Takhtajan 1986; Van Welzen et al. 2005). Madagascan species were coded separately while species from Reunion, Comoros, Canary Islands and Cape Verde islands were coded as Africa. Seychelles and America were also assigned two geographic regions and all areas were coded according to their placement on different tectonic plates and the existence of distinct floras. Thus, the coded areas are (1) America; (2) Africa; (3) Madagascar; (4) Seychelles; (5) Middle East; (6) India; (7) Sri Lanka; (8) Himalaya; (9) East Asia; (10) Sunda Shelf; (11) Sahul Shelf. The data matrix was prepared coding for presence/ absence in each of the areas depending on the collection details, voucher specimens and current distributions. The maximum number of areas in ancestral ranges was constrained to two to avoid wide ancestral ranges and excessive analysis time. Thus, all taxa were coded as present either 99 in a single area or two. Mimusops elengi which occurs in India, Sri Lanka and Burma and Isonandra lanceolata which occurs in India, Sri Lanka and Borneo were coded as present in India and Sri Lanka; the samples were collected from the latter and the region has the highest number of collections for the species. Ancestral areas within internal nodes were constructed using Biogeobears (Landis 2013) in the R package under four models; DIVA like (Ronquist 1997), DIVA LIKE+J, DEC (Ree & Smith 2008), DEC+J. In addition to two free parameters; d (dispersal), e (extinction) included in DEC model (Batalha-Filho et al. 2014, Landis 2013), the new additional parameter “J“ is added to the model to account for founder event speciation, which is not addressed by the models in DIVA and DEC. Log likelihood values for each model were compared and the model with the highest value was chosen as the best model for inferring ancestral ranges at nodes. Results Phylogenetic Analysis-ITS The consensus tree resulting from chloroplast data has a very low clade support and the placement of some major taxa in the phylogeny is taxonomically doubtful. Thus, only nuclear ITS phylogeny is carried towards the results and discussion. Table 4.4 Descriptive statistics of ITS and plastid data DNA region Amplicon length Alignment length Number of variable characters Number of informative characters ITS 1036 1024 562 (54.88%) 399 (38.96%) Chloroplast 5463 5444 1082 (19.87%) 423 (7.77%) The maximum clade credibility tree resulting from Bayesian analysis is shown in Figure 4.2. It comprises of a basal grade including the outgroup Sarcosperma, Eberhardtia, a grade of Sideroxylon, Xantolis, Englerophytum, Omphalocarpum, Pouteria, Neolemonniera, Lecomtedoxa, Northia, Capurodendron, Inhambanella within which is nested a large clade of all other species (clade III). 100 The Clade III is weakly supported as monophyletic (PP=0.23) and comprises tribes Mimusopeae and Isonandreae. An early diverging clade of Baillonella, Vitellaria and Vitellariopsis is resolved (clade IV; PP=1). Clade V, containing all Mimusopinae species examined, it is not supported as monophyletic (PP=0.47). The Manilkarinae clade (Clade VI) comprising the genera Manilkara, Labramia, Faucherea and Labourdonnaisis strongly supported as monophyletic (PP=1.0). Clade VII containing the Manilkara grade is strongly supported as monophyletic (P=0.95) and Letestua durissima is nested within. The Isonandreae (clade VIII) is highly supported as monophyletic (pp=1) and comprises of sub-clades Palaquium (Clade IX, PP=1), Isonandra (Clade XIII, PP=1) and clade XV (PP=1) containing Madhuca, Burckella, Diploknema and Payena species. Palaquium (clade IX) is not monophyletic, with Aulandra longifolia and Diploknema butyracea nested in it. Within Clade IX, clade X is strongly supported as monophyletic (PP=1) with P. impressionervium from west of Sunda Shelf and Palaquium petiolare endemic to Sri Lanka being the sister clade to Sri Lankan endemic Palaquium clade. All other Sri Lanka Palaquium species are strongly supported as monophyletic (PP=1). The clade XI (PP=1) comprises of all other Palaquium species, dominated by species from Sunda Shelf, with Palaquium species from east of the line and China, and Aulandra longifolia and Diploknema oligomera nested within. The genus Isonandra (Clade XIII) is not monophyletic and instead forms a grade within which Madhuca utilis and Madhuca crassipes are nested. The Sri Lankan endemic Isonandra montana is sister to the other species in the strongly supported clade XIII (PP=1) and with other Sri Lankan species nested within sub clade XIV with high support (PP=1). Clade XVI is moderately supported (PP=0.87) with Madhuca hainanensis from China and Indo Sri Lankan Madhuca longifolia being the sister clade to other taxa. The Sri Lankan endemic Madhuca clavata and Burckella species are nested within it. Within sub clade XVII, Diploknema is the sister to Payena clade (PP=0.52). All the Payena species (Payena clade) are strongly supported as monophyletic (PP=1). 101 The sub clade XVIII is strongly supported (PP=1) as monophyletic and is dominated by Madhuca species from Sunda Shelf with and contains clade XIX comprising inter alia Sri Lankan endemic Madhuca with a moderate support (PP=0.86). Divergence time estimates Mean ages with 95% HPD confidence intervals for key nodes and area probabilities are reported in Table 4.5. The MCC tree from the BEAST analysis (Figure 4.2) resolves the mean crown age of the tribe Mimusopeae as 43.8 (42.9-46.2) Mya in the Mid Eocene. The mean age of sub tribe Mimusopinae (Clade V) is estimated to be 35.9 Mya and Manilkarinae (Clade VI) is 32.9 (29.4-58) Mya, both having originated during the Oligocene. The origin for genus Mimusops is 10 (5.5-16) Mya while Mimusops elengi show a more recent origin during the late Miocene 6.8 (3.5-11.1) Mya. The genus Manilkara (Clade VII) is resolved as 28.8 (28.0-31) Mya, with Indo Sri Lankan Manilkara hexandra diverging from19.0 (11.3-25.5) Mya during the early Miocene. Table 4.5 Posterior probabilities, divergence ages and ancestral area probabilities; A= America, B= Africa, C= Madagascar, D= Seychelles, E= Middle East, F= India, G= Sri Lanka, H= Himalaya, I= East Asia, J= Sunda Shelf, K= Sahul Shelf. Clade No Clade Name PP Divergence time (Mya) Area probabilities I Sapotaceae 1 92.49 (70.56-119.64) BI=0.43, BJ=0.43, IJ=0.11 II Sideroxylon clade 1 70.33 (60.11-82.75) B=0.52, AB=0.45, A=0.03 1 43.5 (42.94-44.9) B=0.95 III IV Vitellaria clade 1 34.47(23.54-43.12) B=1.00 V Mimusopinae 0.47 35.86 B=1.00 VI Manilkarineae 1 32.9 (29.37-58) B=0.86, BC=0.13 VII Manilkara 28.81 (28.05-30.94) B=0.99 VIII Isonandreae 1 50.97 (43.54-59.76) J=0.82, G=0.18 IX Palaquium 1 42.86 (33.86-52.46) J=0.71, GJ=0.16, G=0.13 X 1 31.93 (22.88-41.43) G=0.57, J=0.28, GJ=0.15 XI 1 29.6 (20.82-40.88) J=1.00 1 46.31 (38.5-54.45) J=0.78, G=0.16, GJ=0.06 1 36.91 (27.76-47.25) J=0.42, G=0.38, GJ=0.20 0.36 28.49 G=0.61, J=0.34, GJ=0.03 1 38.79 (31.77-46.09) J=0.91, G=0.05, GJ=0.02 XII Madhuca/Isonandra /Payena XIII Isonandra XIV XV Madhuca 102 XVI 0.87 33.14 (25-40.02) J=0.46, G=0.38, GJ=0.06 XVII Diploknema/Payena 0.52 34.66 (26.76-42.25) J=0.83, IJ=0.06, HJ=0.06 XVIII Madhuca 1 27.47 (20.89-35.14) J=1.00 0.86 8.37 (4.16-13.3) J=0.97, GJ=0.01, JK=0.01 XIX The mean crown age for Isonandreae (Clade VIII) is 51 (43.5-59.8) Mya and the divergence time estimate for Palaquium clade (Clade IX) is 42.9 (33.7-52.5) Mya, during the middle Eocene. The Sri Lankan Palaquium shows an initial origin 31.9 (22.9-41.4) Mya and diversified into two lineages, endemic P. petiolare during 27.1 (17.9-38) Mya and the second lineage giving rise to endemic taxa 16.4 (10.7-23.1) Mya during the mid-Miocene. The clade XI shows an initial origin during the Oligocene 29.6 Mya and further a rapid diversification during the mid-Miocene. The basal taxon of clade XIII, Sri Lankan endemic Isonandra montana shows an early divergence 36.9(27.8-47.2) Mya while the second lineage (Clade XIV) diverged during 28.5 Mya followed by more recent speciation of endemics 5.3 (2.0-10.8) Mya. The crown node age for clade XVI is 33.1 (25-40.0)Mya with Indo Sri Lankan Madhuca longifolia shows an origin 27.4 (17.3-36) Mya and endemic Madhuca clavata 31.5 Mya. The clade XIX comprises of Sri Lankan endemic Madhuca species that began to diversify from 11.5 (4.2-13.3) Mya followed by speciation of endemics more recently during the Pleistocene. Ancestral area reconstruction and intercontinental dispersal events Among the models compared in BIOGEOBEARS; DEC, DEC+J, DIVA LIKE, DIVA LIKE+J, the DEC+J model resulted in higher likelihood value, thus chosen as the best fit model for my data and the results are given in Table 4.6. 103 Table 4.6 d (dispersal), e (extinction), j (j value, founder-event speciation) LnL (log likelihood) for each of the geographic range evolution models compared in Biogeobears. Model d e j LnL DEC 0.0018 0.0034 0 -284.17 DEC+J 8e-04 0 0.0104 -264.49 DIVA LIKE 0.0019 0.0014 0 -290.50 DIVA LIKE+J 8e-04 0 0.0098 -271.96 The tribe Mimusopeae, sub-tribe Manilkarinae, Mimusopineae and the genera Tieghemella, Autranella, Mimusops, Manilkara, Labramia and Faucherea/Labourdonnaisia are all inferred to have African ancestry. Mimusops shows an initial origin in Africa during the late Eocene and long distance dispersals have occurred to Madagascar and Indo Sri Lankan region during the late Miocene. Manilkara shows a similar pattern of diversification which began in Africa during the late Eocene, and Indo Sri Lankan Manilkara hexandra and Eastern Asian Manilkara kauki shows an African origin, during the early Miocene. Neotropical Manilkara also have an African ancestry, which dispersed to South America during early Miocene 21.61 (12.97-27.68) Mya. 104 Figure 4.1. Bayesian majority rule consensus tree based on nuclear ITS data. Bayesian posterior probability (PP) support values are indicated next to the nodes. 105 Figure 4.2 Maximum-clade-credibility chronogram of a Beast analysis of the ITS region Sapotaceaedata set. Node heights indicate mean ages. Numbers at nodes represent clades in Table 4.2. Branches coloured according to their optimal range reconstructions under the DEC+J model in the pacakge Biogeobears. Pie charts show the relative probability of ancestral state reconstructions at selected nodes. Dotted lines indicate posterior clade probabilities less than 0.95. 106 Figure 4.3 Maximum clade credibility chronogram of Sapotaceae ITS dataset. Node heights indicate mean ages. Node bars indicate 95% highest posterior density date ranges. Numbers inside boxes at each node represent node numbers and values next to nodes are the posterior probability values for each node. 107 Figure 4.4 Maximum clade credibility chronogram of a Beast analysis of the Sapotaceae chloroplast dataset. Values next to nodes are the posterior probability values>0.5. 108 Isonandreae have originated from an African ancestor, and dispersed to Sunda Shelf followed by rapid diversification within and in surrounding areas. The Palaquium clade IX has an origin in Sunda shelf during the Eocene 42.9 (33.7-52.5) Mya, with diversification occurring during the Oligocene-Miocene with subsequent dispersal to the Sahul Shelf twice 1.1 (0.04-3.5) Mya and 4.1 (1.8-6.9) Mya, and Eastern Asia 3.7 (1.13.8) Mya during the late Miocene. Sri Lanka was reached by a single dispersal event and further diversification occurred in two lineages; one giving rise to P. petiolare during the Oligocene and the other lineage giving rise to ten endemic taxa during the Miocene 16.4 (10.7-23.1) Mya. The genus Isonandra shows an initial origin on the Sunda Shelf 46.3 (38.5-54.4) Mya and subsequently spread to other areas; 2 separate dispersals to Sri Lanka and a single dispersal to the Sahul Shelf. The Sri Lankan endemic Isonandra montana has an origin during the late Eocene 36.9 (27.8-47.2) Mya and the rest of Sri Lankan endemics in the genus show an entrance to the island from Sunda shelf during Oligocene 28.5 Mya and further speciation of endemics thereafter. The ancestral area for Indo Sri Lankan Isonandra lanceolata is constructed as Sri Lanka, thus short distance dispersal from Sri Lanka to India has occurred more recently 0.76 Mya during the Pleistocene. Sri Lankan Madhuca have three origins. The origin for Indo-Sri Lankan Madhuca longifolia and Chinese Madhuca hainanensis is constructed as Sri Lanka, thus short distance dispersal has occurred to the Indian continent and further dispersal to Eastern Asia has occurred during the Oligocene 27.4 (17.3-35.9) Mya. Madhuca clavata shows an early Oligocene entry to Sri Lanka during 31.5 Mya and further a more recent entry occurred during the mid-Miocene 11.5 Mya followed by in situ speciation of endemics thereafter. The Burckella clade nested within clade XVI shows an origin in Sunda Shelf and there are two dispersals to the Sahul Shelf and a single long distance dispersal to India which occurred during the Mid Miocene 10.9 (5.2-17.1) Mya. Discussion A non-Gondwanan origin of Sri Lankan Sapotaceae The hypothesis of an ancient Gondwanan origin for the family in Sri Lanka is not supported by my data, as all lineages show origins that considerably post date the 109 breakup of Gondwana. All three larger genera, Palaquium, Isonandra and Madhuca, originated during the Eocene 31.9 (22.9-41.4) Mya, 36.9 (27.8-47.2) Mya and 33.1 (2540.0) Mya respectively, followed by further diversification thereafter. There is some evidence, based on other dated phylogenies, for Sri Lankan lineages being of Gondwanan origin such as in Crypteroniaceae. The family provides a classic example for overland migration referred to as the “out of India hypothesis”, with arrival of some ancient Gondwanan lineages to Asia by rafting on the Indian plate (Conti et al. 2002; Rutschmann & Eriksson 2004; Moyle 2004; Karanth 2006). It comprises three genera, Crypteronia that is widely distributed in South East Asia, Dactylocladus with one species (Dactylocladus stenostachys) endemic to Borneo and Axinandra zeylanica endemic to Sri Lanka with three other Axinandra species occurring in the Malay Peninsula and the northern part of Borneo (Conti et al. 2002; Rutschmann & Eriksson 2004; Moyle 2004; Karanth 2006; Morley 2003; Renner et al. 2010). However the absence of these taxa in Peninsular India today is attributed to massive volcanism during the Cretaceous-Tertiary boundary and associated extensive aridification in India during the late Tertiary which resulted in substantial extinction (Morley 2003; Karanth 2006). Being an island perhaps Sri Lanka was less affected by these climatic changes and thus could have acted as a refugial area facilitating the persistence of these ancient Gondwanan lineages such as Axinandra zeyalnica (Ashton & Gunatilleke 1987; Morley 2003). One striking example is the Sri Lankan endemic monotypic genus Hortonia in Monimiaceae, with other genera in the family occurring in America, Africa, Madagascar and The Mascarenes, New Caledonia, Australia, New Zealand and the Malesian region (Renner et al. 2010). The dated phylogeny of Monimiaceae suggests a Gondwanan origin for Sri Lankan Hortonia which dates back to 71 (57-84) Mya during the late Cretaceous. One possibility of this ancient lineage of Hortonia in Sri Lanka is explained as rafting on the Deccan plate to Asia (Ashton & Gunatilleke 1987). The other explanation is long distance dispersal of Hortonia from Antarctica to Sri Lanka, however the huge distance between two land masses ca 2100 Km (Ali & Aitchison 2008), raises the question of ability to cross such long distance over water according to Renner et al. (2010). 110 Other families have been speculated to be of Gondwanan origin but dated phylogenies need to be produced to confirm this. The family Dilleniaceae is one example of a classic Gondwanan distribution, with all Asian genera in the family found on the Indian Peninsula and Sri Lanka with their closest relatives in Madagascar and the Seychelles. Raven & Axelrod (1974) considered that Schumacheria in the family Dilleniaceae rafted on the Deccan plate to Asia based on Dickson’s observations (1967-1969) that the Sri Lankan endemics share share many characters with the rare Bornean genus Didesmandra. Dipoterocarpaceae, a family well represented in Sri Lanka and with a wider distribution in South America, Africa, Madagascar, India, Southeast Asia, and Malesia also exhibits a Gondwanan distribution, potentially facilitated by rafting on the Deccan plate (Ashton & Gunatilleke 1987; Givnish & Renner 2004; Ducousso et al. 2004; Renner et al. 2010). The endemic genus Vateriopsis in the Seychelles has its closest relative Vateria in India and Sri Lanka, and has large wingless fruits that lack dormancy. It has been argued that these features provide evidence for a vicariant origin as they are not adapted to any form of long distance dispersal (Ashton & Gunatilleke 1987). Laurasian lineages My second hypothesis was to test for any Laurasian lineages of Sapotaceae in Sri Lanka. The Deccan plate, composed of both India and Sri Lanka, collided with the Southern coast of Laurasia during the Eocene between 55-40 Mya (Briggs et al. 2003; Aitchison et al. 2008; Morley 2003), a profound tectonic event which led to the uplift of the Himalayas (White & Lister 2012; Ali & Aitchison 2008; Conti et al. 2002; Briggs et al. 2003). Since then, due to the connection with Asia, a vast immigration of Laurasian taxa and mixing of the floras from Deccan Gondwana and Laurasia was hypothesized (Ashton & Gunatilleke 1987, Morley 2000). However only a few migrants are evident; lower Tertiary fossils of Juglandaceae and Myricaceae in Assam provide evidence for the beginning of Laurasian immigrations (Ashton & Gunatilleke 1987). Families distributed in the Northern hemisphere with Laurasian origin such as Pinus, Hamamelidaceae, Juglandaceae, Myricaceae, Fagaceae and Clethraceae are completely lacking in South Asia (Ashton & Gunatilleke 1987). However, present study does not support any laurasian lineages in Sri Lanka for the family Sapotaceae and this may be 111 due to the migration difficulties caused by the Himalayan barrier and by the associated development of arid conditions after the montane uplift (Ashton & Gunatilleke 1987). Long distance dispersals from Africa A relatively close connection between Africa and the Deccan plate was facilitated via Madagascar, which may have allowed many plant taxa to disperse from Africa to the Deccan plate during its northward journey in the late Cretaceous (Morley 2003; Morley 1998; Aitchison et al. 2008). As the plate drifted along the west coast of Africa, small islands and land bridges may have facilitated short to medium distance over water dispersals during the early to late Cretaceous (Briggs et al. 2003). The distance between the African coast and the Deccan plate was ca. 420 km, which remained constant until c. 84 Mya and pollen records during that period provide strong evidence for dispersals (Morley 2003; Morley 1998; Rutschmann & Eriksson 2004 ). However, this dispersal path ceased when the Deccan plate separated from Madagascar Late Cretaceous (94 – 84 Mya); since then there have been few dispersal events between Africa and the Deccan plate during its northward journey to contact with Asia (Morley 2003; Morley 2000). There are two major dispersal phases from Africa to Asia evidenced based on macrofossil records, one occurring soon after the collision with Asia during the Eocene, 54-36 Mya and a further phase during the late Miocene 10-5 Mya (Morley 2000). The tribe Mimusopeae evolved 52 Mya, followed by diversification 43 Mya during the Eocene when warm and wet climatic conditions were prevailing in the Northern hemisphere (Armstrong et al. 2014). Asian representatives of the genera Manilkara and Mimusops are nested within a grade of predominantly African genera (Tieghemella, Autranella, Baillonella, Vitellaria, and Vitellariopsis) suggesting an African origin that is highly supported (>95%) in ancestral area reconstructions. One possibility for the pantropical distribution of Manilkara and Mimusops could be attributed to migration through the boreotropics which could have occurred 65-45 Mya via the North Atlantic Land Bridge. However, the timing of pantropical distribution of the genera is too young to accept this hypothesis. However timings congruent with a boreotropical migration are consistent with several other families such as Burseraceae (Weeks et al. 2005), Malpighiaceae (Davis et al. 2002) and Meliaceae (Muellner et al. 2006). The expansion of tropical forests during the mid-Miocene climatic optimum, 17-15 Mya (Zachos et al. 112 2001) may also have facilitated the expansion of tropical taxa between Africa and Asia (Zhou et al. 2012). This pathway is followed by several animals such as primitive catarrhines, hominoids and lorisoids for the entry to Eurasia from Africa in the Early to Middle Miocene and some plant groups such as Uvaria (Annonaceae) (Zhou et al. 2012). Another possible explanation is migration from Africa to Asia after the collision of the Afro Arabian plate with Asia during the late Oligocene 25 Mya (Samuei et al. 1997). This could explain the migration of Asian Manilkara from Africa, as this pathway was available during the time frame of the split of this clade from its African relatives. However, due to the young ages evident in the dated phylogeny of Sapotaceae, the most plausible explanation for the disjunct pantropical distribution of Manilkara and Mimusopsis long distance dispersal from Africa to Madagascar, Asia and the Neotropics (this study; Armstrong et al. 2014). The recent origin of Sri Lankan Manilkara hexandra provides evidence for dispersal from Africa to the Indo-Sri Lankan region during the Miocene 19.04 (11.3-25.52) Mya, as does Mimusops elengi 6.81 Mya (3.4711.12) which subsequently spread eastward into Malesia (Armstrong et al. 2014). Long distance dispersals from Southeast Asia My results confirm the Sundanian origin for Sri Lankan Isonandreae which is congruent with the results of Richardson et al. (2014). The Sri Lankan lineages are nested in Sundanian ones, indicating an origin in Sunda shelf followed by six independent longdistance dispersals; one during the Eocene 36.9 Mya, four during the Oligocene 33.1 Mya, 31.9 Mya, 31.5 Mya, 28.5 Mya and one during the mid-Miocene 11.5 Mya. The Indian plate moved extremely close to the Malay Peninsula during its northward migration (Hall 2001) and the collision with Laurasian coast brought India and South East Asia to similar latitudes and within the same climatic belt. Further, the collision involved a glancing contact with Sumatra and then Burma in the late Paleocene onward 57 Mya facilitating more opportunities for floristic exchange (Ali & Aitchison 2008). However, the Indian flora is suggested to have been much more aggressive in migration, evidenced by the sudden appearance of Indian taxa in the Paleocene and Early Eocene in the palaeo floras of South East Asia and which resulted in the enrichment of the South East Asian flora (Morley 2003; Morley 2000). 113 The genus Palaquium has an ancestral area reconstructed as Sunda Shelf during the mid Eocene (Clade IX, Figure 4.2). It migrated to Sri Lanka giving rise to the endemic Palaquium petiolare during Oligocene and to a clade of endemic taxa (CladeX, Figure 4.2) during the Miocene. Palaquium impressionervium is likely a single back dispersal to Sunda shelf from Sri Lanka during the late Oligocene, but this supported by tree topology only. Among Isonandra species, Isonandra montana was a much earlier arrival on Sri Lanka from Sunda Shelf during the Eocene (36.91 Mya, Clade XIII, Figure 4.2) followed by another dispersal phase during the Oligocene giving rise to other endemic Isonandra species (Isonandra zeylanica, Isonandra compta, Isonandra sp1, Isonandra sp 2) on the island (Clade XIV, Figure 4.2). The ancestral area reconstruction for Madhuca (Clade XVIII) shows a Sundanian origin during the early Oligocene. The earliest arrival to the Indo Sri Lankan region occurred during the early Oligocene (33.14 Mya), giving rise to Mudhuca longifolia. Sri Lanka is constructed as the ancestral area for Indo Sri Lankan Madhuca longifolia and Chinese Madhuca hainanensis, consistent with a dispersal from Sri Lanka to India and then to Eastern Asia during the early-Mid Oligocene (27.4 Mya, Clade XVI, Figure 4.2), possibly by overland migration. Sri Lanka is separated from India by a narrow sea barrier called Palk strait and there have been intermittent land connections between these two land masses until sea level rises 6000 years ago (Ashton & Gunatilleke 1987; McLoughlin 2001). During the period from the late Eocene (39-36 Mya) and Oligocene (36-25 Mya)warm pre humid conditions were prevailing across the Indian sub-continent (Morley 2000) and Pleistocene deposits from the wet zone of Sri Lanka suggest the existence of intermittent periods of seasonal tropical climate (Ashton & Gunatilleke 1987). Thus, relatively short distance dispersal over land or water would potentially facilitate this migration and further radiations during favourable climatic conditions. Madhuca clavata, a point endemic species in Sri Lanka, shows an origin in Sunda shelf (Clade XVI, Figure 4.2) during the Oligocene and another more recent dispersal event is evident within Sri Lankan Madhuca, during the Mid Miocene, which resulted in further evolution of endemics (Clade XIX, Figure 4.2). 114 The phylogeny presented here is consistent with relatively recent intercontinental dispersal being the major factor in disjunctions in the extant tropical flora. Sri Lankan Sapotaceae have a more recent origin, beginning in the Eocene ca. 36 Mya followed by speciation of endemics more recently during Pliocene-Pleistocene. The major direction of dispersal to Sri Lanka is from the east, predominantly from Sundania which has played a prominent role in enrichment of Sri Lankan flora. The warm temperate and perhumid climate in the south western part of Sri Lanka, with nearest analogues in Sumatra, perhaps may have facilitated these prevalent dispersals to Sri Lanka through providing favourable habitats (Ashton & Gunatilleke 1987). More intense sampling of Indian and Indochinese taxa is necessary to further test this hypothesis. The Sapotaceae fruits possess different sizes ranging approximately from 1-10 cm, all with a fleshy, sweet colourful pericarps provide a hint for animal dispersals such as mammels and birds. Mimosopeae fruits range from 1.5-10cm, with a fleshy, sweet pericarp which are dispersed by various mammals and birds. Asian and African Manilkara are dispersed by frugivorous birds such as doves and pigeons (Corlett 1998, snow 1981). It has been recorded that primates such as spider monkeys, howler monkeys, capuchins and tamarins (Chauvet et al. 2004; Chapman 1989; Oliviera & Ferrari 2000) and fruit bats (Uriarte et al. 2005) are responsible for fruit and seed dispersal of Manilkara in neotropics. Palaquium species possess yellowish green coloured fleshy fruits with average size ranging from 2-4 cm possibly a hint for animal dispersal such as monkeys and birds. Smooth, fleshy, red-orange coloured fruits of Isonandra species with average size (1.4 cm) possibly eaten by small birds and carry over long distances. Fruits ofMadhuca species are large in size, about 3-4 cm long and the large size indicates the possible vectors as tropical bats, which are too large to be eaten and carried by birds. Thus, all Sapotaceae species possess a large fruits and seeds which are too bulky to be carried out and dispersed by wind. Thus, they possibly carried out by mammals, such as rodents and monkeys for short distances and in the gut content of birds across long distances. Most of the trans-oceanic dispersals included in the study are dated back to Middle Miocene thermal maximum 23-12Mya) and warmer climates should have been 115 contributed to an increase in hurricanes and tornadoes which were capable of carrying large seeds and propagules over long distances (Graham 2006, Nathan et al. 2008). Phylogenetic studies have confirmed the capability of transoceanic dispersals of some other families with large fleshy fruits such as, Annonaceae (Su & Saunders 2009); Adansonia, Bombacaceae (Baum et al. 1998); Atelia, Leguminosae (Ireland et al. 2010); Andira, Leguminosae (Skema 2003); Commiphora, Burseraceae (Weeks et al. 2007); Macherium, Leguminosae (Lavin et al. 2000); Symphonia, Clusiaceae (Dick et al. 2003). Incongruence between plastid and nuclear data Other studies of Sapotaceae have relied on ITS datasets in ancestral area reconstructions due to the hard topological incongruence between chloroplast and nuclear datasets (Armstrong et al 2014; Richardson et al. 2015). One of the major reasons for utilizing ITS sequences is mainly due to their high variability since plastid data were not informative enough to test for alternative hypothesis. Chloroplast capture across long distances has been identified as the major cause for hard incongruences between plastid and nuclear data and it is demonstrated in several pant groups. According to Armstrong et al. 2014, M.hexandra (Sri Lanka) and M. littoralis (Myanmar) are placed with other two Asian species in the nuclear phylogeny. However, those two species form a strongly supported clade (PP=1) in a plastid phylogeny with two African species, M. mochisia (Zambia) and M. concolor (South Africa), which is an indication of a chloroplast capture event (Armstrong et al. 2014). A possible hypothesis is hybridization between African and Asian lineages, with the African lineage donating the plastid DNA to the Asian species giving rise to M. hexandra and M. littoralis in Asia. That hypothesis could be confirmed by the origin of Manilkara in Africa and dispersal events from Africa to Asia and diversification thereafter. Some other species in Sapotaceae such as Chrysophyllum cuneifolium and Nesoluma sps also show evidence of chloroplast capture, this time between South American and African lineages (Swenson et al. 2008; Smedmark & Anderberg 2007). Chloroplast capture has also been found in genera such as Nothofagus (Acosta & Premoli 2010), Thuja (Peng & Wang 2008), Gossypium (Wendel et al. 1995), Heuchera (Soltis et 116 al.1991; Soltis & Kuzoff 1995) and Boykinia (Soltis et al. 1996). However, these incongruences should be investigated with the use of additional markers and the exact mechanisms of underlying incongruences beg further explanations. In situ speciation My results confirm that in situ speciation is an important contributor to the richness of the Sri Lankan flora. Species that are endemic to Sri Lanka, and also have sister species in Sri Lanka, are likely to have arisen by in situ speciation, whereas endemics whose closest relatives occur elsewhere might be paleoendemics that once had wider distributions. Within Sri Lankan Sapotaceae, more rapid radiation of endemics occurred during the Pliocene-Pleistocene, with ten endemic species appearing during the Pleistocene, possibly as a result of the changing climatic conditions during inter glacial periods. During the period of glacial maxima, the global temperature was cooler and drier, around 3-4oC in the equatorial regions and lowland rainforests in the equatorial zones trap species until favourable conditions (Bonnefille et al. 1992; Sosef 1994). The glacial cycles would have led to shifting and fragmentation of the climatic zones within Sri Lanka, where species would survive in isolation until environmental conditions become favourable. These contractions could have resulted in allopatric speciation giving rise to new species in isolated forest patches within the island and re-expansion of their ranges in the warm and wetter climate conditions during the Pleistocene cycles (Plana et al. 2004; Richardson et al. 2014). References Acosta, M. C. & Premoli, A. C., 2010.Evidence of chloroplast capture in South American Nothofagus (subgenus Nothofagus, Nothofagaceae). Molecular Phylogenetics and Evolution 54(1), pp. 235-242. Aitchison, J.C., Ali, J.R. & Davis, A.M., 2008. When and where did India and Asia collide? Geological Bulletin of China, 27(9), pp.1351–1370. Ali, J.R. & Aitchison, J.C., 2008. Gondwana to Asia: Plate tectonics, paleogeography and the biological connectivity of the Indian sub-continent from the Middle 117 Jurassic through latest Eocene (166-35??Ma). Earth-Science Reviews, 88(3-4), pp.145–166. Anderberg, A.A. & Swenson, U., 2003. Evolutionary lineages in Sapotaceae (Ericales): a cladistic analysis based on ndhF sequence data. International Journal of Plant Sciences, 164, pp.763–767. Armstrong, K.E. et al., 2014. Patterns of diversification amongst tropical regions compared: a case study in Sapotaceae. Frontiers in Genetics, 5(December), pp.1– 13. Ashton, P.S. & Gunatilleke, C.V.S., 1987. New light on the plant geography of Ceylon I . Historical plant geography. Journal of Biogeography, 14(3), pp.249–285. Bartish, I. V. et al., 2005. Phylogenetic relationships among Sapotaceae (Ericales) in Australia and Oceania: Molecular evidence for generic polyphyly and repeated dispersal. Amer. J. Bot. 92(4), pp.667–673. Bartish, I. V. et al., 2011. Vicariance or long-distance dispersal: Historical biogeography of the pantropical subfamily Chrysophylloideae (Sapotaceae). Journal of Biogeography, 38(1), pp.177–190. Batalha-Filho, H. et al., 2014. Phylogeny and historical biogeography of gnateaters (Passeriformes, Conopophagidae) in the South America forests. Molecular Phylogenetics and Evolution, 79, pp.422–432. Baum, D. A. et al., (1998). Biogeography and floral evolution of baobobs (Adansonia, Malvaceae) as inferred from multiple data sets. Systematic Biology, 47(2),pp.181207. Bonnefille, R.F. et al., 1992. Quantitative estimates of full glacial temperatures in equatorial Africa from palynological data. Clim. Dyn, 6, pp.251–257. Briggs, J.C., Collections, Z. & Museum, G., 2003. The biogeographic and tectonic history of India. Jounal of Biogeography, 30, pp.381–388. 118 Chapman, C. A., 1989. Primate seed dispersal - the fate of dispersed seeds. Biotropica, 21(2), pp.148-154. Chauvet, S. et al., 2004. Seed fate of two Sapotaceae species in a Guianan rain forest in the context of escape and satiation hypotheses. Journal of Tropical Ecology, 20, pp. 1-9. Conti, E. et al., 2002. Early Tertiary out-of-India dispersal of Crypteroniaceae: evidence from phylogeny and molecular dating. Evolution; international journal of organic evolution, 56(10), pp.1931–1942. Cook, L.G. & Crisp, M.D., 2005. Not so ancient: the extant crown group of Nothofagus represents a post-Gondwanan radiation. Proceedings of the Royal Society B: Biological Sciences, 272, pp.2535–2544. Corlett, R. T. (1998). Frugivory and seed dispersal by vertebrates in the Oriental (Indomalayan) Region. Biological Reviews, 73(4), pp.413-448. Davis, C.C. et al., 2002. Laurasian migration explains Gondwanan disjunctions: evidence from Malpighiaceae. Proceedings of the National Academy of Sciences USA, 99, pp.6833–6837. Demesure, B., Sodzi, N. & Petit, R.J., 1995. A set of universal primers for amplification of polymorphic non-coding regions of mitochondrial and chloroplast DNA in plants. Molecular ecology, 4(1), pp.129–131. Dick, C. W. et al., 2003. Molecular sytematic analysis reveals cryptic tertiary diversification of a widespread tropical rain forest tree. The American Naturalist, 162(6),pp.691-703. Drummond, A.J. & Rambaut, A., 2007. BEAST: Bayesian evolutionary analysis by sampling trees. BMC evolutionary biology, 7, pp.214. 119 Ducousso, M. et al., 2004. The last common ancestor of Sarcolaenaceae and Asian dipterocarp trees was ectomycorrhizal before the India-Madagascar separation, about 88 million years ago. Molecular Ecology, 13(1), pp.231–236. Givnish, T.J. & Renner, S.S., 2004. Tropical Intercontinental Disjunctions: Gondwana Breakup, Immigration from the Boreotropics, and Transoceanic Dispersal. International Journal of Plant Sciences, 165, pp.S1–S6. Govaerts, R. et al., 2001. World checklist and bibliography of Sapotaceae London: Royal Botanic Gardens, Kew. Graham, A., 2006. Modern processes and historical factors in the origin of the African element in Latin America. Annals of the Missouri Botanical Garden, 93(2),pp.335339. Gruas-Cavagnetto, C., 1976. Etude palynologique du Paleogene du sud de l’Angleterre. Cah.Micropal, 1, pp.1–49. Hall, R., 2001. Cenozoic reconstructions of SE Asia and the SW Pacific: Changing patterns of land and sea. Faunal and Floral Migrations and Evolution in Se AsiaAustralasia, (Figure 1), pp.35–56. Hall, T., 1999. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symposium Series, 41, pp.95–98. Hamilton, M.B., 1999. Four primer pairs for the amplification of chloroplast intergenic regions with intraspecific variation. Mol. Ecol. 8, pp.521–523. Ireland, H. E. et al., 2010. Biogeographical, ecological and morphological structure in a phylogenetic analysis of Atelia (Swartzieae, Fabaceae) derived from combined molecular, morphological and chemical data. Botanical Journal of the Linnean Society, 162, pp.39-53. 120 Jacobs, B. F. et al., 2005. Oligocene terrestrial strata of northwestern Ethiopia: a preliminary report on paleoenvironments and paleontology. Palaeontol. Electron, 8, pp.1–19. Kearse, M. et al., 2012. Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics, 28(12), pp.1647-1649. Landis, M.J. et al., 2013. Bayesian analysis of biogeography when the number of areas is large. Systematic Biology, 62(0), pp.789–804. Lavin, M. et al.,2000. Africa, the odd man out: Molecular biogeography of dalbergioid legumes (Fabaceae) suggests otherwise. Systematic Botany, 25(3),pp.449-467. Lee, C. & Wen, J., 2004. Phylogeny of Panax using chloroplast trnC-trnD intergenic region and the utility of trnC-trnD in interspecific studies of plants. Molecular Phylogenetics and Evolution, 31(3), pp.894–903. Mathews, J.P., 2009. Chicle: The chewing gum of the Americas, from the ancient Maya to William Wrigley. Tucson: University of Arizona Press. McLoughlin, S., 2001. The breakup history of Gondwana and its impact on preCenozoic floristic provincialism. Australian Journal of Botany, 49(3), pp.271–300. Miller, M.A. et al., 2015. A RESTful API for Access to Phylogenetic Tools via the CIPRES Science Gateway. Evolutionary Bioinformatics, 11, pp.43-48. Moller, M., & Cronk, Q., 1997. Origin and relationships of Saintpaulia(Gesneriaceae) based on ribosomal DNA internal transcribed spacer (ITS) sequences. American Journal of Botany, 84, pp.956 – 965. Morley, R. J., 2000. Origin and evolution of tropical rain forests. Chichester: John Wiley & Sons, Ltd. Morley, R.J., 2003. Interplate dispersal paths for megathermal angiosperms. Perspectives in Plant Ecology, Evolution and Systematics, 6(12), pp.5–20. 121 Morley, R.J., 1998. Palynological evidence for Tertiary plant dispersals in the SE Asian region in relation to plate tectonics and climate. Biogeography and Geological Evolution of SE Asia, pp.211–234. Moyle, R.G., 2004. Calibration of molecular clocks and the biogeographic history of Crypteroniaceae. Evolution; international journal of organic evolution, 58(8), pp.1871–1873; discussion 1874–1876. Muellner, A.N. et al., 2006. The mahogany family "out-of-Africa": divergence time estimation, global biogeographic patterns inferred from plastid rbcL DNA sequences, extant, and fossil distribution of diversity. Molecular Phylogenetics and Evolution, 40, pp.236–250. Nathan, R., et al., 2008. Mechanisms of long-distance seed dispersal. TRENDS in Ecology & Evolution, 23(11),pp.638- 647. Oliveira, A. C. M. & Ferrari, S. F., 2000. Seed dispersal by black-handed tamarins, Saguinus midas niger (Callitrichinae, Primates): implications for the regeneration of degraded forest habitats in eastern Amazonia. Journal of Tropical Ecology, 16,pp.709-716. Olmstead, R.G., Sweere, J. a & Spangler, R.E., 1999. Phylogeny and provisional classification of the Solanaceae based on chloroplast DNA. Solanaceae IV: Advances in biology and utilization, pp.111–137. Peng, D. & Wang, X. Q., 2008. Reticulate evolution in Thuja inferred from multiple gene sequences: Implications for the study of biogeographical disjunction between eastern Asia and North America. Molecular Phylogenetics and Evolution, 47(3),pp.1190-1202. Pennington, T. D., 1991. The genera of the Sapotaceae. Royal Botanic Gardens Kew & New York Botanical Garden. Plana, V. et al., 2004. Pleistocene and pre-Pleistocene Begonia speciation in Africa. Molecular Phylogenetics and Evolution, 31, pp.449–461. 122 Posada, D. et al., 2012. jModelTest 2: more models, new heuristics and parallel computing. Nature Methods, 9, pp.772–772. Praveen Karanth, K., 2006. Out-of-India Gondwanan origin of some tropical Asian biota. Current Science, 90(6), pp.789–792. Rambaut, A., 2009. FigTree, a graphical viewer of phylogenetic trees. Institute of Evolutionary Biology University of Edinburgh. Raven, P. H., & Axelrod, D.I., 1974. Angiosperm biogeography and past continental movements. Ann. Missouri Bot. Gard, 61,pp.539– 673. Ree, R.H. & Smith, S. a, 2008. Maximum likelihood inference of geographic range evolution by dispersal, local extinction, and cladogenesis. Systematic biology, 57(1), pp.4–14. Renner, S.S. et al., 2010. Biogeography of the Monimiaceae (Laurales): A role for East Gondwana and long-distance dispersal, but not West Gondwana. Journal of Biogeography, 37(7), pp.1227–1238. Richardson, J.E. et al., 2004. Historical biogeography of two cosmopolitan families of flowering plants: Annonaceae and Rhamnaceae. Philosophical transactions of the Royal Society of London. Series B, Biological sciences, 359(1450), pp.1495–1508. Richardson, J.E. et al., 2014. The influence of tectonics, sea-level changes and dispersal on migration and diversification of Isonandreae (Sapotaceae). Botanical Journal of the Linnean Society, 174(1), pp.130–140. Ronquist, F., 1997. Dispersal-vicariance analysis: A new approach to the quantification of historical biogeography. Systematic Biology, 46(l), pp.195–203. Ronquist, F. et al., 2012. MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space. Systematic Biology, 61, pp.539–42. 123 Rutschmann, F. & Eriksson, T., 2004. Did Crypteroniaceae really disperse out of India? Molecular dating evidence from rbcL, ndhF, and rpl16 intron sequences. International Journal of plant Science, 165(S4), pp.S69–S89. Samuei, M. A. et al., 1997. Field ob- servations from the Socotran platform: Their interpretation and cor- relation to Southern Oman. Mar. Petr. Geol, 14, pp.661– 673. Setoguchi, H. et al., 1998. Phylogenetic relationships within Araucariaceae based on rbcL gene sequences. American Journal of Botany, 85, pp.1507–1516. Skema, C., 2003. Phylogeny and biogeography of Andira. A case study in Neotropical speciation. M.Sc. Thesis, Edinburgh: University of Edinburgh. Smedmark, J.E.E. & Anderberg, A. a., 2007. Boreotropical migration explains hybridization between geographically distant lineages in the pantropical clade Sideroxyleae (Sapotaceae). American Journal of Botany, 94(9), pp.1491–1505. Smedmark, J.E.E., Swenson, U. & Anderberg, A. a., 2006. Accounting for variation of substitution rates through time in Bayesian phylogeny reconstruction of Sapotoideae (Sapotaceae). Molecular Phylogenetics and Evolution, 39(3), pp.706– 721. Snow, D. W. (1981). Tropical frugivorous birds and their food plants - a world survey. Biotropica, 13(1),pp.1-14. Soltis, D. E. et al., 1996. Discordance between ITS and chloroplast topologies in the Boykinia group (Saxifragaceae). Systematic Botany,21(2),pp.169-185. Soltis, D. E. & Kuzoff, R. K., 1995. Discordance between nuclear and chloroplast phylogenies in the Heuchera group (Saxifragaceae). Evolution, 49(4),pp.727-742. Soltis, D. E. et al., 1991. Chloroplast DNA variation within and among genera of the Heuchera group (Saxifragaceae) - Evidence for chloroplast transfer and paraphyly. American Journal of Botany, 78(8), pp.1091-1112. 124 Sosef, M.S.M., 1994. Refuge Begonias. Taxonomy, phylogeny and historical biogeography of Begonia sect. Loasibegonia and sect. Scutobegonia in relation to glacial forest refuges in Africa (Stud- ies in Begoniaceae 5). Wageningen Agric. Univ. Pap. 94, pp.1–306. Su, Y. C. F. & Saunders, R. M. K., 2009. Evolutionary divergence times in the Annonaceae: evidence of a late Miocene origin of Pseuduvaria in Sundaland with subsequent diversification in New Guinea. BMC Evolutionary Biology, 9(153). Swenson, U. & Anderberg, A.A., 2005. Phylogeny, character evolution, and classification of Sapotaceae (Ericales). Cladistics, 21, pp.101–130. Swenson, U., Backlund, A., McLoughlin, S. & Hill, R.S., 2001. Nothofagus biogeography revisited with special emphasis on the enigmatic distribution of subgenus Brassospora in New Caledonia. Cladistics, 17, pp.28–47. Swenson, U. et al., 2007. Phylogeny, diagnostic characters and generic limitation of Australasian Chrysophylloideae (Sapotaceae, Ericales): evidence from ITS sequence data and morphology. Cladistics, 23(3),pp.201-228. Swenson, U., Nylinder, S. & Munzinger, J., 2014. Sapotaceae biogeography supports New Caledonia being an old Darwinian island. Journal of Biogeography, 41, pp.797–809. Swenson, U., et al., 2008. Multi-gene phylogeny of the pantropical subfamily Chrysophylloideae (Sapotaceae): evidence of generic polyphyly and extensive morphological homoplasy. Cladistics, 24, pp.1006–1031. Takhtajan, A., 1986. Floristic Regions of the World. Berkeley, CA: University of California Press. Thomas, D.C. et al., 2012. West to east dispersal and subsequent rapid diversification of the mega-diverse genus Begonia (Begoniaceae) in the Malesian archipelago. Journal of Biogeography, 39(1), pp.98–113. 125 Uriarte, M. et al., 2005. Seedling recruitment in a hurricane-driven tropical forest: light limitation, density-dependence and the spatial distribution of parent trees. Journal of Ecology, 93(2),pp.291-304. Van Welzen, P. C., Slik, J. W. F., & Alahuhta, J., 2005. Plant distribution patterns and plate tectonics in Malesia. Kong. Dansk. Vidensk, 55, pp.199–217. Weeks, A., Daly, D.C. & Simpson, B.B., 2005. The phylogenetic history and biogeography of the frankincense and myrrh family (Burseraceae) based on nuclear and chloroplast sequence data. Molecular Phylogenetics and Evolution, 35, pp.85– 101. Weeks, A. & Simpson, B. B., 2007. Molecular phylogenetic analysis of Commiphora (Burseraceae) yields insight on the evolution of historical biogeography of an “impossible” genus. Molecular Phylogenetics and Evolution, 42,pp.62-79. Wendel, J. F. et al., 1995. An unusual ribosomal DNA sequence from Gossypium gossypioides reveals ancient, cryptic intergenomic introgression. Molecular Phylogenetics and Evolution,4(3),pp.298-313. White, L.T. & Lister, G.S., 2012. The collision of India with Asia. Journal of Geodynamics, 56-57, pp.7–17. Zhou, L. et al., 2012. “Out-of-Africa” dispersal of tropical floras during the Miocene climatic optimum: Evidence from Uvaria (Annonaceae). Journal of Biogeography, 39(2), pp.322–335. Zachos, J. et al., 2001. Trends, rhythms, and aberrations in global climate 65 Ma to present. Science (New York, N.Y.), 292(April), pp.686–693. 126 CHAPTER 5: Pantropical biogeography of Zingiberaceae Introduction Zingiberaceae is the largest family of the order Zingiberales, with 53 genera and over 2000 species (Kress et al. 2005; Kress et al. 2002). It is pantropical in distribution, with one genus Renealmia L.f. in the Neotropics, four genera (Aframomum K.Schum., Aulotandra Gagnep., Siphonochilus J.M.Wood & Franks and Renealmia L.f.) occurring in Africa, with the bulk of the remaining genera occurring in South and South East Asia (Kress et al. 2002, Särkinen et al. 2007). The family is later divided into four sub families and six tribes based on the nuclear ITS and chloroplast matK:Siphonochiloideae (Siphonochileae), Tamijioideae (Tamijieae), Alpinioideae (Alpinieae, Riedelieae), and Zingiberoideae (Zingibereae, Globbeae) (Kress et al. 2005). Costaceae, which was included in the family is now placed as the sister group to Zingiberaceae.Siphonochilus and Tamijia are the basal clade within Zingiberaceae (Kress et al. 2002). Table 5.1 Division of Zingiberaceae according to Kress et al. (2002) Subfamily Tribe Genera Siphonochiliodeae Siphonochileae Siphonochilus Tamijioideae Tamijieae Tamijia Alpinioideae Alpinieae Etlingera, Amomum, Alpinia, Hornstedia Riedelieae Riedelia, Burbidgea, Pleuranthodium Zingibereae Zingiber, Boesenbergia, Curcuma, Hedychium Globbeae Globba, Hemiorchis, Gagnepainia, Mantisia Zingiberoideae The genus Tamijia shows a restricted distribution in Nothern Borneo. The monotypic genus Siphonochilus is found in Tropical Africa and all other genera are shared among Alpinioideaea and Zingiberoideae. The subfamily Zingiberoideae are found only in Asia where as Alpinioideae are widely distributed in Africa, Neotropics and Asia. Within Alpinioideae, Renealmia are distributed in Neotropics while Aframomum in both Africa 127 and Neotropics. The Alpinioideae have their main centre of diversity in Malesia, while Zingiberoideae in Indochina (Kress et al. 2002). A phylogenetic analysis by Kress et al. (1995) on Zingiberales using both morphological and rbcL DNA sequence data could obtained a better resolution across the order by combining both data sets than either one alone. Later three molecular data sets, including two chloroplast genes (rbcL and atpB) and one nuclear gene (18S) were utilized in combination with morphological data, have further proven the effectiveness of combing both molecular and morphological data in phylogenetic studies (Kress et al. 2001). Within Alpinoideaea, two genera, Alpinia and Amomum are two large genera with wider distributions across South Asia, Indo china to Southeast Asia (Xia et al. 2004). Alpinia is the largest and most taxonomically complex genus in Zingiberaceae and parsimony and Bayesian analysis using Internal Transcribed Spacer (ITS) and chloroplast matK suggested six clades across the polyphyletic tribe Alpinieae (Kress et al. 2005). Phylogenetic analysis using same markers confirmed the polyphyly of the genus Amoumum with three well defined groups (Xia et al. 2004) and combined data sets resulted in more highly resolved phylogenies. Further, chloroplast matK have been used in various genera of the Zingiberaceae; Aframomum, Alpinia, Curcuma, Globba, Hedychium and Zingiber in order to find inter and intra-level relationships among them. Polymorphism assessment of matK has proven the suitability of the region for DNA barcoding of Zingiberaceae (Selvaraj et al. 2008). Studying the extant distribution patterns of tropical plant families and understanding the history of tropical floristic patterns is an emerging science. The Zingiberaceae is an ideal group for this purpose given their pantropical distribution and the availability of robust phylogenetic data (Kress & Specht 2006). The order Zingiberales originated ca. 124 Mya, followed by divergence events leading to the formation of the family lineages by ca. 110 Mya (Kress & Specht 2006). The ancestral area for the entire Zingiberales was reconstructed as America and Southeast Asia (Kress & Specht 2006) using a Dispersal-Vicariance analysis (DIVA;Ronquist 128 1997). Even though the age estimates for the order are congruent with a Gondwanan vicariance hypothesis, the lack of tropical climate during that time in Gondwana suggests a Northern Laurasian distribution for the early diverging lineages. Fossil records in America, Eurasia and Greenland during upper Cretaceous to early Tertiary further favours the hypothesis of an ancestral Laurasian distribution(Kress & Specht 2006). The Zingiberaceae-Costaceae group diversified and split into the currently recognised two families ca. 105 Mya. The DIVA analysis of Kress & Specht (2006) reconstructed a wide ancestral area hypothesis for the Zingiberaceae-Costaceae clade, effectively a Gondwanan distribution including Africa, America, Melanesia and Southeast Asia. The ancestral area for the Zingiberaceae was reconstructed as Africa and India. The Zingiberaceae is currently well represented in Africa, Madagascar, India and Sri Lanka. Fossil material found in the Intertrappean beds of the Deccan Plate which resembles Amomum (extant in Africa, tropical Asia and the Mascarenes) provides tantalising evidence for the migration of Zingiberaceae to tropical Asia via rafting on the Indian plate (Ashton & Gunatilleke 1987; Kress & Specht 2006). The Deccan plate, composed of India and Sri Lanka, moved close to the Malay Peninsula during its northward migration in the mid Eocene, thus could have facilitated the dispersal of lineages to Southeast Asia (Kress & Specht 2006; Aitchison et al. 2008). Sri Lanka had an intermittent land connection with India since the separation of the Deccan plate from Gondwanaland until colliding with Eurasia in Miocene(Ashton & Gunatilleke 1987). Thus, Sri Lanka is a key location for examining the route of entry of Zingiberaceae into Tropical Asia and to interpret the timing and diversification of the family in Asia. In Sri Lanka Zingiberaceae show an abundant distribution in the wet zone, mostly in lowland and mid montane primary forest, less so in secondary forests. Sri Lankan Zingiberaceae are composed of twelve genera, of which 12 out of 36 species are known to be endemic. The Sri Lankan Alpinia fax clade is of particular importance since it is one of the potential sister groups to neotropical Renealmia, based on nuclear data (Särkinen et al. 129 2007).The members, Alpinia fax (endemic) & Alpinia abundiflora (native), of this small clade occur in Sri Lanka and a small part of South West India and their close affinity with African/neotropical clades suggest past dispersal paths across the Indian ocean following the breakup of Gondwana (Kress et al. 2005). This study is the first to incorporate multiple Sri Lankan taxa in a global dated phylogeny and to infer molecular divergence age estimates for Sri Lankan Zingiberaceae. I aim to reconstruct the ancestral areas of distribution, focussing on Asian Zingiberaceae, using recent advances in biogeograhic analysis (BIOGEOBEARS, Matzke 2013), which allows us to compare models of goeographic range evolution. I used a finer scale area coding than in the studies of Särkinen et al. (2007) and Droop (2012) in order to obtain a more detailed picture of the origin and timing of diversification events during the early biogeographic evolution of Zingiberaceae. Materials and methods Taxon sampling In order to complement lacking South Asian Zingiberaceae the following 21 samples were added to the DNA alignment generated by Droop (2012) comprising 127 taxa: Zingiber cylindricum, Curcuma albiflora, Amomum graminiflolium, Amomum nemorale, Amomum acuminatum, Alpinia fax, Alpinia sp, Cyphostigma pulchellum, Amomum sp 1, 2, 3, 5, 6, 7 (Sri Lankan endemics 80%), Zingiber wightianum, Elettaria sp, Amomum pterocarpum, Amomum fulviceps, Amomum masticatorium, Alpinia abundiflora, Curcuma zedoaria (distributed in both in Sri Lanka and India, natives 65%). In order to obtain a better understanding of biogeographic affinities with African and neotropical gingers, taxonsampling was increased by adding sequences of Renealmia battenbergiana, Aframomum danielii, Aframomum sceptrum (all native to Africa) which were downloaded from Genebank. The final data set comprised 151 taxa representing broadly all tribes of Zingiberaceae; Alpinieae, Zingibereae, Globbeae, Siphonochilieae and Tamijieae. All the Asian species Zingiberaceae have been proved to be monophyletic and Tamijia flagellaris, Siphonochilus kirkii, S. Decorus and S. Aethiopicus (African) were identified as basally diverging lineages in the order Zingiberales in previous studies 130 (Särkinen et al. 2007; Kress et al. 2005).Thus the Asian species were assigned as monophyletic, allowing basally diverging African taxa to be the out group. The newly generated sequences are listed in table 5.2 and all the accessions used in the study are listed in Appendix 5. Table 5.2 Voucher information for the newly generated sequences for the present study Species name Voucher EDNA number Zingiber_cylindricum Moon Kumarage 7 (E) EDNA13-0033869 Amomum_pterocarpum Thwaites Kumarage 21 (E) EDNA13-0033871 Zingiber_wightianum Thwaites Kumarage 20 (E) EDNA13-0033870 Amomum_sp1 Kumarage 34 (E) EDNA13-0033867 Elettaria_sp Kumarage 5 (E) EDNA13-0033868 Curcuma_zedoaria Roxb. Kumarage 24 (E) EDNA13-0033872 Amomum_fulviceps Thwaites Kumarage 33 (E) EDNA13-0033873 Amomum_sp2 Kumarage 50 (E) EDNA13-0033874 Curcuma_albiflora Thwaites Kumarage 38 (E) EDNA13-0033875 Amomum_graminifolium Thwaites Kumarage 48 (E) EDNA13-0033876 Amomum_sp3 Kumarage 82 (E) EDNA13-0033878 Amomum_nemorale Benth. & Hook.f. Kumarage 52 (E) EDNA13-0033879 Cyphostigma_pulchellum Benth. Kumarage 54 (E) EDNA13-0033880 Amomum_acuminatum Thwaites Kumarage 55 (E) EDNA13-0033881 Amomum_masticatorium Thwaites Kumarage67 (E) EDNA13-0033882 Alpinia_abundiflora1 Burtt & R.M.Sm. Kumarage 70 (E) EDNA14-0035527 Amomum_sp5 Kumarage 86 (E) EDNA14-0035531 Amomum_sp6 Kumarage 88 (E) EDNA14-0035532 Amomum_sp7 Kumarage 89 (E) EDNA14-0035533 Alpinia_sp Kumarage 71 (E) EDNA14-0035528 Alpinia_fax Burtt & R.M.Sm. Kumarage 94 (E) EDNA14-0035529 131 DNA region sampling Many Studies on Zingiberaceae have used chloroplast markers such as matK and trnL-F in combination with ITS in order to obtain phylogenetic resolution (Kress et al. 2005; Kress et al. 2002). The Internal Transcribed Spacer (ITS) is a useful marker in resolving species level relationships in Zingiberaceae and the high copy number makes it easy to amplify. matK, which is located within the intron of trnK, is another marker with proven suitability in resolving species level relationships due to its high rate of substitution and low transition/transversion ratios. However, trnL-F does not provide highly resolved trees due to lack of variation, thus was not used in my study. DNA extraction, amplification and sequencing Total genomic data was extracted from silica gel dried material using DNeasy Plant Mini Kit (Qiagen, UK) according to the manufacturer’s protocols. 25 μl PCR reactions were setup for amplification of both ITS and matK regions. For ITS, each 25μl PCR reaction contained 16.125μl of ddH2O, 2.5μl of 10x reaction buffer, 1.25μl 25mM MgCl2, 2.5μl 2mM dNTPs, 0.75μl 10μM forward primer (ITS4), 0.75μl 10μM reverse primer (ITS5p), 0.125μl of Biotaq DNA polymerase (Bioline, UK) and 1μl of DNA of DNA template. The PCRtemperature profile was template denaturation at 94oC for 3 min followed by 29 cycles of denaturation at 94oC for 1 min, primer annealing at 55oC for 1 min, primer extension at 72oC for 1.5 min followed by a final extension step at 72oC for 5 min. The matK region was amplified in four separate reactions using the primer pairs trnK1 and mIR; mSP2F and m5R; mIF and m8R; and m8Fa and trnK2R (Table 5.3). Each 25μl PCR reactioncontained 15.25μl of ddH2O, 2.5μl of 10x reaction buffer, 1.25μl 25mM MgCl2, 2.5μl 2mM dNTPs, 0.75μl 10μM forward primer, 0.75μl 10μM reverse primer, 0.8μl 0.4% BSA, 0.2 μl of Biotaq DNA polymerase (Bioline, UK) and 1μl of DNA template. The temperature profile included an initial template denaturation step of 94oC for 5 minutes, followed by 35cycles of denaturation at 94oC for 30 s, primer annealing at 55oC for 45 Seconds, primer extension at 72oC for 1.5 minutes; and a final extension step at 72oC for 30 seconds. 132 Table 5.3 Nuclear (ITS) and chloroplast primer sequences used in Zingiberaceae DNA sequence generation. Region Primer name Primer sequence References ITS ITS4 TCCTCCGCTTATTGATATGC White et al. 1990 ITS5p GGAAGGAGAAGTCGTAACAAG White et al. 1990 trnK1F (F) CTCAACGGTAGAGTACTCG Manos & Steele 1997 mIR (R) CGT TTC ACA AGTACT GAA CTA Kress et al. 2002 mSP2F (F) TGG GTT AGA GAC GAA TGT GT Kress et al. 2002 m5R (R) AGG ATC CTT GAA AAT CCA TAG A Kress et al. 2002 mIF (F) GTTCAG TAC TTG TGA AAC GTT Kress et al. 2002 m8R (R) AGC ACA AGA AAG TCG AAG Kress et al. 2002 m8Fa (F) TAC TTC GAC TTT CCT GTG CC Kress et al. 2002 trnK2R (R) AACTAGTCGGATGGAGTAG Steele & Vilgalys 1994 matK Amplified products were run on1% Agarose TBE gel with Syber safe as the staining agent and visualized in a UV transilumintor. PCR purification was carried outusing EXOSAP IT in7 μl reactions. 5μl of PCR product was mixed with 2μl of EXOSAP and incubated for 37oC at 15 minutes followed by 80o C for 15 minutes. Sequencing PCR was carried out using purified PCR products in 10μl reactions using 5.68μl of ddH2O, 2μl of sequencing buffer, 0.32μl of primer, 1μl of Big dye and 1–Xμl of template. The sequencing PCR protocol was denaturation at 95oC for 30 sec, followed by 24 cycles of primer annealing at 50oC for 20sec, extension at 60oC for 4 min. Separate forward and reverse sequencing PCR s were carried out and products were sent to the Genepool facility at the University of Edinburgh (Genepool, UK) for analysis. Sequence editing and alignment Newly generated DNA sequences were edited in Geneious (7.1.4; Kerase et al. 2012). Amomum villosum sequences of the relevant marker from Droop (2012) were used as the reference consensus sequence for assembling. Sequences were aligned manually in 133 Bioedit 7.1.3 (Hall 1999) and checked for indels. Bases were excluded due to uncertainties in the alignment or missing data at the region ends, ITS dataset: 108-113, 513-518, 696-776. matK dataset: 3143-3149, 3171-3173. Phylogenetic analysis and Bayesian divergence time estimation Models of sequence evolution were determined using jmodeltest 2.1.3 (Posada et al. 2012). Maximum likelihood topologies were used to estimate the optimal evolutionary model and twenty four models were used under the Akaike Information Criterion (AICc) and the Bayesian Information Criterion (BIC). Both chloroplast and nuclear datasets were analyzed under Bayesian inference (BI) which was performed in Mrbayes 3.2.1 (Ronquist et al. 2012) in the CIPRES science gatewayV. 3.3 (Miller et al. 2010). The regions were treated as single partitions. MCMC runs were carried out for 10000000 generations and sampled every 1000 generations. 25% burn-in was set to discarded the initial trees and the remainder were summarised as a 50% majority rule consensus tree and were visualized in Fig tree (Rambaut 2009) and checked for congruence between regions. A combined matK-ITS data matrix was not used due to hard incongruence between two tree topologies. The ITS region was chosen for molecular dating in BEAST and ancestral area reconstructions in BIOGEOBEARS, due to the more complete sampling and more resolved phylogeny. Bayesian divergence time estimation was performed on the ITS alignment using BEAST v.1.8.0 (Drummond & Rambaut 2007). GTR was selected as nucleotide substitution model with Gamma + invariant site as the site heterogeneity model. The data set was treated as a single partition and uncorrelated relaxed lognormal clock model was selected in order to relax the assumption of a molecular clock and allow for rate heterogeneity between lineages. The tree prior was set to random birth death speciation process with a randomly started generating tree. Four separate Markov Chain Monte Carlo (MCMC) runs were carried out for 10000000 generations sampling every 1000 generations. Plots of the logged parameters for each run were visualised using Tracer v.1.5 (Drummond & Rambaut 2007) to confirm convergence between runs. Time series plots of all parameters were analyzed in tracer 134 to check for convergence and to confirm adequate effective sample sizes. Trees were combined in LOGCOMBINER (Drummond & Rambaut 2007) and burn-in was set for 25% for initial sample for each run. A single maximum clade creditability tree was obtained from Tree Annotator v.1.7.5 (Drummond & Rambaut 2007) and visualized in fig tree v.1.4.0 (Rambaut 2009). Fossil constraints and secondary calibrations The fossil taxon Zingiberopsis magnifolia (Knowlton) Hickey from North America and Eurasia was used in the temporal calibration of the Zingiberaceae phylogeny (Peppe et al. 2007). On the basis of the leaf venation pattern, it has been clearly ascribed to Zingiberaceae and have been used in biogeographic studies in order to assign the age of the crown node of Zingiberaceae (Särkinen et al. 2007; Kress & Specht 2006).The age of the fossil is re-estimated and now it is accepted the earliest occurance was in the Maastrichtian, beginning 71 Ma. The crown age of 69-71 Ma was used where the basal lineages Siphonochilous and Aulotandra splits from the remaining taxa in the phylogeny and a lognormal prior distribution was used in favour of older age estimates. Biogeographic analysis Geographical area delimitation and scoring Twelve areas were coded based on the extant distribution, areas of endemism and geological history. In order to obtain a clear picture of the biogeographic history of South Asian Zingiberaceae, Sri Lanka and India were coded as separate geographic areas. The areas are (1) America; (2) Africa; (3) Madagascar; (4) India; (5) Sri Lanka; (6) China; (7) Indochina; (8) Sunda shelf; (9) Sahul shelf; (10) Papua New Guinea; (11) Philippines; (12) Australia. A data matrix was prepared coding presence/absence in each of the area depending on the collection details and voucher specimens (Zingiberaceae Resource Centre, RBGE). The maximum number of areas in ancestral ranges was constrained to two since all the species found in a single area except Amomum petaloidium and Alpnia japonica found in China-Indochina and Amomum fulviceps, Amomum masticatorium, Zingiber wghitianum, Alpinia abundiflora and Elettaria sp found both in India and Sri Lanka. 135 Ancestral Area Reconstructions Ancestral areas within internal nodes were constructed using Biogeobears (Matzke 2013) in R package under four models; DIVA like (Ronquist 1997), DIVA LIKE+J, DEC (Ree & Smith 2008), DEC+J. DEC and DEC+J calculates maximum likelihood ratios of ancestral states at speciation events, in a method similar to LAGRANGE (Ree et al. 2005). In addition to two free parameters; d (dispersal), e (extinction) included in DEC model (Batalha-Filho et al. 2014; Matzke 2013) the additional parameter “J“ is added to the model to account for founder event speciation. Founder event speciation is important in lineage splitting especially in island systems. The J parameter controls the probability of two events during cladogenesis; founder event speciation versus sympatric and vicariant speciation (Matzke 2013). Log likelihood values for each model were compared and the model with highest value was chosen as the best model for inferring ancestral ranges at nodes. Results Phylogenetic Analysis Both ITS and matK dating were carried out, however alignments were not combined in a single analysis due to the hard incongruence of the datasets. The ITS alignment was used in biogeographic analysis as it gives a better resolution across the family and is compatible with published phylogenies (Särkinen et al. 2007; Kress & Specht 2006; Kress et al. 2005). The resulted from Bayesian and beast analysis of nuclear ITS region are shown in figures 5.1 and 5.2 respectively. Table 5.4 Descriptive statistics of nuclear ITS and matK datasets DNA region Amplicon length Alignment length Number of variable characters Number of informative characters ITS 776 695 432 (62.15%) 360 (51.79%) matK 3173 3170 882 (27.82%) 540 (17.03%) Within the Sri Lankan Zingiberaceae, the Alpinia fax clade which exhibits prominent incongruence between the trees derived from nuclear and chloroplast data (Figure 5.5). In the ITS tree, the Indo-Sri Lankan Alpinia fax clade is sister to Aframomum from 136 Africa, however it is not supported (PP=0.42). The Sri Lankan endemics Cyphostigma pulchellum and Amomum nemorale are sister to the Afro-American Renealmia with high support (PP=0.99). In the matK tree the Alpinia fax clade (Clade A) is sister to Cyphostigma pulchellum and Amomum nemorale (although with low support; PP=0.83), and Aframomum is reconstructed as sister to Renealmia (although with low support; PP=0.84). However this topological incongruence does not contribute to any biogeographic incongruence since for both data sets Sri Lanka is constructed as the ancestral area for both African Aframomum and Neotropical Renealmia. Considering the ITS trees, Tamijia flagellaris is constructed as the sister taxon to rest of the Zingiberaceae, which is strongly supported as monophyletic (Clade I, PP=1). The remaining taxa fall into clade II which comprises two strongly supported clades: the subfamilies Zingiberoideae (clade III, PP=1) and Alpiniodeae (Clade VI, PP=1). Clade III consists of Campandra, Gagnepainia, Hemiorchis, Curcuma, Distichichlamys, Hedychium and Zingiber. Within clade III the Sri Lankan endemic Curcuma albiflora is nested within clade IV (PP=1) comprising species from China and Indochina, with the Indochinese Curcuma roscoeana as the sister taxon. The genus Zingiber is strongly supported as monophyletic (Clade V, PP=0.98) with the Sri Lankan endemic Zingiber cylindricum reconstructed as sister taxon to Indo-Sri Lankan and Eastern Asian (China and Indochina) Zingiber. Clade VI comprises tribes Ridelieaeand Alpiniae. The Ridelieae are monophyletic with strong support (Siliquamomum/Burbidgia/Ridelia clade PP=1) while Alpinieae are nonmonophyletic, being spread amongst six clades with varying levels of support. Clade VIII is composed of Sri Lankan Amomum nemorale, Cyphostigma pulchellum, Alpinia species and Elettaria together with African Aframomum and Afro-American Renealmia, but is not supported (PP=0.27). This clade is nested within clade VII (also not supported; PP=0.42), with Alpinia from China, Indochina and Sunda Shelf constructed as the sister group. Sri Lankan Amomum are polyphyletic, with one species Amomum nemorale resolved in a clade with the monotypic Cyphostigma pulchellum (PP=1), sister to Afro-American Renealmia with a high support (PP=0.99). Other Sri 137 Lankan Amomums are nested within clades XII, XIII and XIV with varying statistical support. Sri Lankan Alpinia, Alpinia fax clade (Clade A) are paraphyletic, with indo Sri Lankan Elettaria nested within (PP=1), forming a sister clade to Aframomum from Madagascar with no support (PP=0.42). The Indo-Sri Lankan Amomum pterocarpum is nested within the Amomum/Eletariopsis clade (collapsed in Fig. 3) and forms a clade with Australian Amomum queenslandicum with high support (PP=1.0). Clade IX contains the remaining Alpinia species of Eastern and Southeast Asian origin along with three Chinese Amomum species, A. aff. paratsaoko, A. paratsaoko, A. coriandriodorum nested within (PP=0.52). The clade X (PP=1) is composed of species from Sunda shelf, Indo china, Papua New Guinea along withtwo strongly supported sub-clades of origins with Australian and Philippines. Indo Chinese Hornstedtia hainanensis and Sundanian Amomum centrocephalum are the sister group for rest of taxa. The Phillipine taxa form a monophyletic clade with strong support (P=0.96). The Australian Alpinia species are nested within clade X forming a monophyletic clade with high support value (PP=1). The clade XI is composed of all remaining Amomum species and is strongly supported as monophyletic (PP=1). Sri Lankan endemic Amomum species are polyphyletic, highly nested within clade XI, as are the Indo-Sri Lankan taxa A. masticatorium and A. fulviceps. Divergence time estimates The outgroup Siphonochilus separated from the remaining Zingiberaceae 72.5 Mya, and Tamijia is basal to the remaining Zingiberaceae, separating from them about 57.6 Mya. The mean divergence time for remaining Zingiberaceae is 57.1 (45.5-67.3) Mya. The divergence time for Zingiberoideae (Clade III) is 45.0 Mya and within clade III, Sri Lankan endemic Curcuma albiflora shows an origin dated back to 10.5 (5.0-16.8) Mya. The mean divergence time estimate for the genus Zingiber (Clade V) is 36.7 (26.5-48.7) Mya with the Sri Lankan endemic Zingiber cylindricum being the early diverging lineage at 25.7 Mya (10.1-28.4). Indo Sri Lankan Zingiber wightianum is nested within 138 with a divergence time of 19.1 (10.1-28.4) Mya. The Alpiniodeae (Clade VI) shows an initial diversification beginning 46.3 (35.0-57.6) Mya. The divergence time for clade VII is 35.3 Mya, and clade VIII consisting of Sri Lankan endemic and Afro-American taxa shows a divergence time of 27.3 Mya. Afro-American Renealmia separated from the Sri Lankan endemics Amomum nemorale and Cyphostigma pulchellum 21.5 (11.9-31.8) Mya. 139 Figure 5.1. Bayesian majority rule consensus tree based on nuclear ITS data. Bayesian posterior probability (PP) support values are indicated next to the nodes. 140 Figure 5.2 Maximum-clade-credibility chronogram of a relaxed molecular clock analysis of the ITS data set. Node heights indicate mean ages. Numbers at nodes represent clades in Table 5.2. Branches coloured according to their optimal range reconstructions under the DEC+J model in biogeobears. Pie charts show the relative probability of ancestral state reconstructions at selected nodes. Dotted lines indicate posterior clade probabilities less than 0.95. 141 Figure 5.3 Maximum clade credibility chronogram of the ITS dataset. Node heights indicate mean ages. Node bars indicate 95% highest posterior density date ranges. Numbers inside boxes at each node represent node numbers and values next to nodes are the posterior probability values for each node. 142 Figure 5.4 Maximum-clade-credibility chronogram of a relaxed molecular clock analysis of the matK data set. Values next to nodes are node age values. 143 Figure 5.5 Histogram representing the incongruence between ITS and matK phylogenies. Left: ITS right: matK. * indicate posterior clade probabilities greater than 0.9. 144 Sri Lankan Alpinia fax clade (Clade A) separates from African Aframomum 23.6 Mya and the Aframomum clade shows a recent diversification beginning 4.4 (0.9-11.8) Mya. The Indo-Sri Lankan elements show an initial diversification 13.9 (6.5-23.3) Mya followed by more recent origin of the endemic Alpinia fax 6.0 (1.8-12) Mya. The divergence time estimate for clade IX, consist of Alpinia species of Chinese origin, is 36.0 (26.5-47.9) Mya. Within clade X, which began to diversify 18.5 (12.2-25.7) Mya, Australian Alpinia date to 9.0 (4.4-14.3) Mya. Sri Lankan endemic Amomum are highly nested within clade XI, and show recent separate speciation events, such as Amomum sp 2 at 3.8 Ma and Amomum sp 5 & 6 more recently, 0.9 Ma. The Indo-Sri Lankan Amomum masticatorium is the sister lineage to the remainder of clade XIV which diversified 11.8 (6.4-19.5) Mya with a recent origin of 4.2 (1.2-8.6) Mya for 5 endemic Amomum species and the Indo Sri Lankan Amomum fulviceps. According to matK dating (Figure 5.4), African Aframomum and Neotropical Renealmia shows an initial diversification 22.67 (15.3-32.1) Mya.The divergence time for Aframomum is 6.7 (2.18-13.05), while for Renaelmia it is 10 (4.59-17.2) Mya. Sri Lankan Cyphostigma pulchellum, Alpinia species, Elettaria and Amomum clade shows an initial diversification 19.94 (10.73-29.43) Mya and later Cyphostigma and Amomum nemorale has separated from others 7.82 (2.43-16.15) Mya. Indo Sri Lankan Elettaria is the basal taxon for Alpinia clade which began to diversify 15.27 (7.0224.38) Mya. Southeast Asian Amomum cerasinum and Amomum mentawaiense clade resolves as sister to Sri Lankan Amomum which began to diversify 12.92(7.43-19.62) Mya. Indo Sri Lankan Amomum masticatorium shows and initial diversification 9.3(4.73-14.88) Mya and the rest of Amomum endemics show more recent evolution of 4.7 (1.86-8.66)Mya. Table 5.5 Posterioir probabilities, Divergance ages and Ancestral Area probabilities; E= Sri Lanka, F= China, G= Indo China, H= Sunda Shelf, I= Sahul Shelf, J= Papua New Guinea, K= Philippines. Clade no PP divergance age Area probabilities Clade I 1 70.04 (69.04-74.1) H= 0.64, GH=0.09, FH=0.08 Clade II 1 57.09 (45.52-67.31) H=0.33, G=0.18, F=0.16 145 CladeIII 1 44.96 (33.41-56.5) G= 0.97, E=0.02, EG= 0.01 CladeIV 1 18.74 (10.24-29.2) G=0.58, EG=0.37, E=0.04 CladeV 0.98 25.7 (15.63-36.72) F=0.46, H=0.41, FH=0.1 CladeVI 1 18.95 (10.75-28.85) H=0.65, HJ=0.32, J=0.02 CladeVII 0.42 35.25 F=0.98 CladeVIII 0.27 27.31 E=0.92, C=0.03, A=0.02 CladeIX 0.52 36.02 (26-49-47.9) F=0.50, I=0.22, H=0.13 CladeX 1 18.48 (12.19-25.65) H=0.99 CladeXI 1 22.91 (15.98-31.67) H=0.99 Clade XII 0.64 3.76 (1.34-7.13) F=0.65, E=0.26, EF=0.09 Clade XIII 1 3.82 (1.09-7.46) E=0.44, K=0.44, EK=0.13 Clade XIV 1 11.78 (6.35-19.46) H=0.53, E=0.31, EH=0.09 Biogeographic analysis and ancestral area reconstructions The DEC+J model resulted in a higher likelihood value compared to other models DEC, DIVA LIKE, DIVA LIKE+J tested in BGB, thus was chosen as the best fit model for my data (Table 5.6). The basal node of the core Zingiberaceae in clade I is reconstructed as being Asian to the Sunda Shelf Sunda Shelf (H)= 0.64, Indo China+Sunda Shelf (GH)= 0.09, China+Sunda Shelf (FH)= 0.08 with an age of at least 71 Ma, having split from African and Madagascan lineages in the Siphonochilieae Table 5.6 d (dispersal), e (extinction), j (J value, founder-event speciation) LnL (log likelihood) for each of the models compared in Biogeobears. Model d e j LnL DEC 0.0033 0.0048 0 -350.03 DEC+J 7e-04 0 0.0217 -297.63 DIVA LIKE 0.0031 4e-04 0 -332.48 DIVA LIKE+J 8e-04 0 0.0186 -298.87 The Zingiberoideae (clade III) have an Eastern Asian origin with Indochina constructed as the ancestral area with probabilities: Indo China (G)= 0.97, Sri Lanka (E)= 0.02, Sri Lanka+Indo China (EG)= 0.01. 146 The most likely geographic origin for clade VI is China or Sunda Shelf: China (F)= 0.46, Sunda Shelf (H)= 0.41, China+Sunda Shelf (F+H)= 0.1 and the ancestral area for clade VIII is constructed as Sri Lanka: Sri Lanka (E)= 0.92, Madagascar (C)= 0.03, America (A)= 0.02 with African Aframomum and Afro-American Renealmia nested within. The geographic origin for African Renealmia battenbergiana and other Renealmia from the Neotropics is Sri Lanka: Sri Lanka (E)=0.79, America+Sri Lanka (AE)=0.07, Africa+Sri Lanka (BE)= 0.07 during late Oligocene and the topology is highly supported. The African Aframomum have likely diverged from a Sri Lankan ancestor: Sri Lanka (E)= 0.84, Madagascar+Sri Lanka (CE)= 0.09, Madagascar (C)= 0.07 during the late Oligocene to early Miocene, although the topology at this node is not strongly supported. The ancestral area for the Alpinia fax clade is Sri Lanka (E=1.0). The ancestral area for clade IX is constructed as China with multiple dispersal events between Eastern Asia (China & Indo China), Sunda-Sahul and the Philippines being inferred. Clade X has an area of origin in Sunda shelf (H=0.99), a single dispersal to Australia during the mid Miocene followed by recent speciation in situ. The ancestral area for clade XI is constructed as Sunda Shelf and within clade XI, four dispersal events have occurred to Sri Lanka (Amomum sp 2, Amomum sp 5 & 6, and 2 dispersals within sub clade XIV). The geographic origin for the clade XIV is Sunda Shelf with one dispersal to Sri Lanka giving rise to endemic taxa during Pliocene, and one taxon shared between India and Sri Lanka (Amomum fulviceps) with a Pleistocene origin in Sri Lanka. Discussion Ploidy level, Chromosome numbers in the family Zingiberaceae shows a wide varation in chromosome numbers ranging from 2n=22 to 2n=96. Polyploidization and hybradization events have played a wide role evolution and diversification of ganera in the family Zingiberaceae. The small size and the large number of chromosome number arises in difficulties in carrying out karyological studies, however a recent study revelas the chromosome numbers of three species, 147 Alpinia zerumbet, Globba, and Hedychium. The somatic chromosome number of Alpinia zerumbet has been reported as 2n=52 and the size ranges fron 0.77-3.97μm while in Gobba marantiana 2n=52. The somatic chromosome number of Hedychium spicatum is reported as 2n=68 with length of chromosomes vary from 0.42-1.7μm (Bhadra & Bandyopadhyay 2016). Within Curcuma, the basic chromosome number has been identified as 2n=40, 42 and the chromosome size ranges from 0.5-2.1μm. However, studies have shown that chromosomes size is correlated with climatic conditions they exist, mainly tropical or sub tropical genera contains smaller chromosomes while large chromosomes have been reported in genera in temperate climates. The genus Curcuma is widely spreaded in tropical and sub tropical regions in the world with high species diversity in the Eastern, South Eastern and South Asian region possess a small chromosome like the other genera; Globba and Hedychium. Triploids (3n) are more common in Chinese Curcuma and can be resultdue to the fusion of reduced (n) and unreduced (2n) gamates of diploids within or between species. The triploides are sterile and they might have some type of competitive advantage to distribute in a wide area (Chen & Xia 2013). The basic chromosome number of Curcuma be x=7; some speices are polypolid such as Curcuma longa (a nonaploid (9x)2n=63) and Curcuma amada (a hexaploid (6x) with 2n=42 chromosomes. The genus Zingiber shows less variability in chromosome number and the basic chromosome number is reported as x=11. The diploid chromosome number of Zingiber officinale is 2n=22 while that ofZingiber zerumet is 2n=22 (Bhadra &Bandyopadhyay, 2016) within Indian Curcuma species. Variation of this basic chromosome number and resulting different ploidy levels may have contributed to species adapting to different ecological habitats they live in. ITS and chloroplast incongruence Chloroplast markers are usually single copies and rarely present problems of paralogy which can be problematic with the multi-copy genes such as ITS. However some chloroplast markers such as trnL-F are less informative in Zingiberaceae due slow rates of nucleotide substitution. Nuclear regions such as ITS aremore rapidly evolving in nature and have a proven wide utility in distinguishing between closely related taxa. 148 Hybridisation and uni-directionality of gene flow may result in incongruence between nuclear and plastid gene trees. It may also lead to extensive intra-individual ITS polymorphism (Zaveska et al. 2012; Rieseberg et al. 1996; Okuyama et al. 2005). In some genera of Zingiberaceae, hybridization along with the multi copy nature of the nuclear ITS region can result in paraloguous sequences being amplified in PCR and this may make phylogenetic interference difficult. Since direct sequencing may result in unreadable sequences, cloning is sometimes necessary for ITS region to reveal the broad range of polymorphisms within the individuals (Zaveska et al. 2012). In some genera like Curcuma, polymorphism within individuals has been observed, commonly resulting in unreadable DNA sequences for the ITS region. It has been observed that the polyploid origin, homoploid hybridization and mode of reproductionhave a high impact on this gene paralogy in Zingiberaceae (Zaveska et al. 2012). According to their results, the basal lineages in the phylogeny of Curcuma have very low variation and are represented by single terminal sequences in theirphylogeny, which suggests a single ITS paralogue prevails in the genome. Hybrids and polyploid speciation is rare in these basal groups and they mainly reproduce asexually. However, some terminal taxa showed high levels of ITS polymorphism and they are mainly hybrids or allopolyploids. That is likely the cause of the high intra specific variation and ITS paralogues in those individuals, however it is independent from the ploidy level of a particular taxa (Zaveska et al. 2012). In this study, the ITS sequences generated did not present any polymorphisms, and all were clear reads and did not highlight a need for cloning. It seems the problems of paralogy when using ITS in Zingiberaceae are mostly restricted to the Curcuma group. Origin of the pantropical distribution of Zingiberaceae The most basal lineages of Zingiberaceae are found in Africa (Siphonochilus) and Borneo (Tamijia) and Siphonochilus splits from the rest of Zingiberaceae 70.0 (69.074.1) Mya. Following the Cretaceous thermal maximum, global cooling 80-70 Mya resulted in the disappearance of megathermal taxa from the nothern hemisphere and subsequent dispersals of these taxa to the equatorial regions such as Southeast Asia. The greatest specific and generic diversity in the Zingiberaceae, and within the Alpinieae, is found in Continental Asia, to the west of Wallace’s Line and to the north of the Isthmus 149 of Kra. The subfamily Alpinoideae has its centre of diversity in the Malesian region while that of Zingiberoideae lies in the monsoonal areas of Indochina and Thailand, which have been called the evolutionary centre of the Zingiberaceae due to its exceptional diversity in those areas (Larsen 2005). Sunda-Sahul exchange The more recently diverged Alpinieae (Clade IX, Fig. 5.2) show a migration from west to east across Wallace’s Line, with many of these dispersals having occurred during the Miocene, however the topology at the base of the clade is poorly supported in both ITS and chloroplast trees. A similar pattern of dispersals can be seen in Begonia (Thomas et al. 2012), Cyrtandra (Gesneriaceae, Clark et al. 2008) and the Isonandreae (Sapotaceae, Richardson et al. 2014). Extensive land connections between Peninsular Malaysia and islands of Borneo, Java and Sumatra will likely have facilitated multiple transitions between these land masses (Voris 2000; Sathiamurthy & Voris 2006). Further, much of the Sunda Shelf and Penninsular Malaysia were covered by rain forests since the Miocene, thus should have provided suitable habitat for establishment and diversification of invading mega thermal taxa. The timing of the split of Alpinia oceanica and A. vittata at 9.2 (4.8-14.0) Mya and their occurrence in New Guinea to the east of Wallace’s line is consistent with a dispersal from Sunda shelf during the late Miocene, potentially facilitated by island hopping over the land masses that emerged in Sulawesi and New Guinea and the emergence of volcanic islands along the Sunda Arc, the Banda Arc and the Halmahera Arc (Hall 2001; Hall 2009). The Phillipine clade also shows Sunda Shelf origin 13.7 (8.6-19.4) Mya, and is best explained by long distance dispersal or island hopping via Palawan or the Sulu Archipelago between Borneo or Sangihe Arc between Sulawesi and the Philippines (Jones & Kennedy 2008). Australian Zingiberaceae are monophyletic and are nested within clade X which has an origin in Sunda Shelf during the middle Miocene. The collision of the Australian plate with Pacific, Philippine and Sunda plates during the Oligo-Miocene boundary resulted in the formation of New Guinea and the islands of the Banda Arc in Indonesia (Morley 2000; Morley 2003) during the Mid Miocene. These islands provide a potential route for flora to disperse across Wallace’s line, and the existence of flora with Australian 150 affinity in the Sunda shelf at present provide evidence for this dispersal route (Morley 2000). The majority of plant dispersals were into the Southeast Asian region from Australia rather than the opposite direction, such as Phormium or Dianella (Hemerocallidaceae) Dacrydium (Podocarpaceae) Camptostemon (Bombacaceae) Dacrycarpus (Podocarpaceae) (Morley 2003). However there are few examples of dispersals from the Southeast Asian area such as Stenochlaena palustris (Blechnaceae), Acacia, Caesalpinia and Crudia (Leguminosae) and Merremia (Convolvulaceae) observed during Miocene (Morley 2003). Warmer and wetter conditions prevailed in the Sunda Shelf, resulting in latitudinal expansion of rain forests and the decline of the mesic biome and extinction of rain forest lineages during cenozoic in Australia, providing niches available for immigrants from Sunda Shelf (Kershaw et al. 1994; Greenwood & Christophel 2005; Crayn et al. 2014). Further, Sahul-Sunda collision during the Mid Miocene resulted in extension of lands east of Wallaces line, enhancing pathways for stepping-stone migration of plant lineages between Sahul and Sunda which is evidenced by an increase in floristic exchange rate from 12 Mya (Crayn et al. 2014). My results provide supportive evidence for the contributions of immigrations from Sunda shelf to Australian tropical flora which is in contrast to earlier views on Gondwanan origin for Australian flora. According to the molecular age estimates, many of the Southeast Asian Alpinieae lineages in clades X and XI show a recent Plio-Pleistocene diversification associated with the climatic and sea level fluctuations and shifts in forest distribution (Cannon et al. 2009; Woodruff 2010). Origin and diversification of Sri Lankan Zingiberaceae My results strongly contradict a Gondwanan origin for Sri Lankan Zingiberaceae, which consist of much younger lineages ranging dating from 25.7 (15.6-36.7) Mya to 3.8 (1.37.1) Mya. The Sri Lankan species of Curcuma and Zingiber are nested within species from China and Indochina and have an ancestral area reconstructed as Eastern Asia during the Oligo-Miocene boundary approximately 25 Ma. The collision of the Deccan plate, composed of both India and Sri Lanka, with the southern coast of Laurasia during the 151 Eocene between 55-40 Mya (Briggs et al. 2003; Aitchison et al. 2008; Morley 2003) resulted in the the mixing of floras from Deccan Gondwana and Laurasia (Ashton & Gunatilleke 1987; Morley 2000). A hypothesis of a Laurasian migration of these Zingiberaceae lineages is congruent with the timing and geographic origins observed in this study (Figure 5.2). More seasonal, monsoonal climates prevailed in nothern India during the early Miocene, becoming more ever-wet in the mid-Miocene (Morley 2000). This would have provided habitat for the immigrants from China-Indochina to diversify in evergreen rain forests in the Indo-Sri Lankan region. My results confirm the Sundanian origin for Sri Lankan Amomumspecies. One hypothesis for this dispersals phase between Sunda Shelf and Indo-Sri Lankan region is believed to occur during Indian plate’s northward migration. The Indian plate moved extremely close to the Malay Peninsula during its northward migration during the mid Eocene (Hall 2001) and the Indian plate and parts of Southeast Asia would have been at similar latitudes during the mid-Miocene. The collision with Laurasian coast 35 Mya brought the northeastern corner of the sub-continent into glancing contact with Sumatra and to a same climatic belt giving the potential for floristic exchange (Ali & Aitchison 2008). However, the Indian flora was much more aggressive in migration, thus a sudden appearance of Indian Paleocene and Early Eocene of taxa are evidenced in palaeo floras which resulted in the depletion of local Southeast Asian flora (Morley 2003; Morley 2000). However, my date estimates do not support above dispersal phase, instead three independent long-distance dispersals occurred later are evidenced; ca 11.8 (6.3-19.4) Mya, 10.2 Mya, 3.8 (1-7.4) Mya and further diversification of endemics more recently associated with changing climatic conditions during glacial-interglacial periods during the Pliocene-Pleistocene. Amomum sp 2 shows a more recent (3.8 Mya) origin in China, thus a long distance dispersal or overland migration during Pliocene is inferred. Amomum species show a wide distribution in China- Indo China, India and Sri Lanka, thus, overland migration is a most plausible explanation for extant distributuion for the genus. The genus exhibits a diversity of fruits characters, smooth, echinate, winged, lobed and ridged. The method of dispersal of the fruits is not known for most Amomum species. However, the fleshy capsules and aromatic arils provide a hint for animal 152 dispersal, thus overland migration by animals may probably be the best explanation for extant distribution of the genus. Pollination and seed dispersal mechanisms Little is known about pollination of Amomum however, the majority of species possess characters which provide hints for animal pollination. Some trumpet shaped flowers with orange or yellow colouration like Amomum aculeatum are pollinated by bees. Some species such as the Sri Lankan endemic Amomum nemorale contains nectar guides for pollinators as sharp bright makings on labellum which facilitates bee landing. Amomum subulatum is a bumble bee pollinated species (Kishore et al. 2011) and Amomum maximum is pollinated by two species of honey bees, Apis dorsata and Apis cerana (Ren et al. 2007). Three pollination syndromes have been identified in Bornean Zingiberaceae; Amegilla bees are the only pollinators for Amomum calyptrum, Amomum gyrolophos and Amomum oligophyllum (Sakai et al. 1999). Amomum roseisquamosum has a long floral tube and is pollinated by spiderhunters, while Amomum polycarpum and other related species are pollinated by birds and various insects (Sakai et al. 1999). Large, dark red, showy bracts characterise the vertically standing inflorescences found inAmomum apiculatum and Amomum centrocephalum, which are pollinated by birds (Nagamasu & Sakai 1996). Many of the Sri Lankan Amomum species such as Amomum masticatorium and Amomum fulviceps possesses petaloid anther crests that might be involved in the attraction of as yet unknown pollinators. Two types of flower colours can be seen in Neotropical Renealmia; one with bright redpink flowers and the other group with white coloured flowers. Species such as Renealmia alpinia and Renealmia cernua have a tubular labellum, bright red-pink flowers and well developed nectary glands which are pollinated by hummingbirds (Maas, 1977) and it correlates with the high species diversity of hummingbirds in the Andes (Sarkinen et al. 2007). The other group of Renealmia with yellow/white flowers and poorly developed nectary glands are assumed to be pollinated by small insects such as bees (Maas 1977). The majority of the species in the genus Alpinia are yellowish white in colour and those are pollinated by large bees, however some species are pollinated by bats and birds (Zhang et al. 2003; Kress &Spetch2006). Different mechanisms have evolved to reduce 153 self-pollination, such as different levels of styles and different timings of anther dehiscence have been observed (Zhang et al. 2003). Since most of the species in Zingiberaceae flower near to the ground, animals are likely the most important vectors in fruit and seed dispersal (Pfeiffer et al. 2004). Fruits of Amomum are berries, size ranging from 4mm (Amomum oligophyllum) to 4.5 cm and covered with a fleshy aril. Some like Amomum pterocarpum have hairy fruits which can be trapped on bird’s legs and travel long distances. Fruits of Zingiber cylindricum are brightly red coloured with a fleshy aril and borne at the base of the plants. Seeds and fruits are eaten by small mammals like rodents and monkeys and can be transported to long distances with their gut contents. The genus Aframomum in African forests produces bright red, fleshy fruits with a sweet juicy pulp which is a reward for primates and other small mammals (Harris et al. 2000). Three different flower types: platform, tube and flag are present and pollination is mainly carried out by birds, bees and butterflies. An uncommon dispersal mechanism is exhibited by Aframomum psuedostipulare, which is dispersed by a fish that migrates to forest during flooding (Harris et al. 2000). Interestingly, it has been observed that the major seed dispersal in Globba in Borneo is occurred with the aid of ants. The fleshy aril provides food for ants and allows easy handling. Dispersal by one species of ant (Polyrhachis sp.) has been observed to carry seeds about 800cm(Pfeiffer et al. 2004). Sri Lanka’s influence on African and Afro-American Zingiberaceae Renealmia resolves as monophyletic (P=1) with the African species R. battenbergiana being the sister to other Renealmia species from the Neotropics. The genus forms a strongly supported clade (P=0.99) with the Sri Lankan endemics Amomum nemorale and Cyphostigma pulchellum. Renealmia was found by Sarkinen et al. (2007) to have the Alpinia fax clade as an unsupported sister group, with Aframomum or AmomumElettariopsis clades also as potential sister group to Renealmia due to poorly supported tree topology. This is the first instance that the endemics Cyphostigma pulchellum and Amomum nemorale have been incorporated in a phylogeny, and shows that the amphiatlantic Renealmia is nested within a clade with Sri Lanka as the ancestral area 154 (E=0.79, AE=0.07, BE= 0.07) The Afromadagascan genus Aframomum, the largest genus of African Zingiberaceae with approximately 80 species and one of the largest genera of African rainforest understory herbs (Harris et al. 2000), too exhibits a similar pattern, being nested within a clade which has Sri Lanka as the ancestral area. The hypothesis for these Sri Lanka/Africa exchanges being caused by early dispersal during rafting of the Deccan plate is refuted as the divergence between the Sri Lankan and African lineages are too young. A possible scenario for tropical disjunctions between Africa and Asia is overland migration after the Afro-Arabian plate collided with Asia approximately 25 Mya (Kulju et al. 2007; Li et al. 2009; Yuan et al. 2005). This dispersal route out of Africa in to Southeast Asia is evidenced in some plants such as Uvaria (Annonaceae) (Zhou et al. 2012) and in some Primates (Stewart & Disotell 1998; Zhou et al. 2012). The closing of the Tethys Sea and the formation of the Gomphotherium Land Bridge occurred later during 19 – 16 Ma, during which time the mid-Miocene Thermal Maximum would have facilitated the movement of megathermal species from Asia across the Arabian Peninsula into Africa (Zhou et al. 2012). The entrance of Renealmia in to Africa 21.5 (11.9-31.84) during the late Oligocene is congruent with this migratory path. Global temperatures were high during the Eocene climatic optimum 55-50 Mya (Zachos et al. 2001), and then cooler until late Oligocene, when temperatures started rising again, peaking at the mid-Miocene Thermal Maximum (17–15 Ma; Zachos et al. 2001) which potentially facilitated tropical and subtropical vegetation becaming more widespread in Africa (Andrews & van Couvering 1975; Coetzee 1978) thus providing favourable habitats for immigrants. However long distance dispersal from Sri Lanka directly to Africa is another potential explanation, which could occur at any point in geological time. Many African–Asian plant disjunctions have been explained by reference to long-distance dispersal, such as Exacum (Gentianaceae) (Yuan et al. 2005), Bridelia (Phyllanthaceae) (Li et al. 2009), Macaranga and Mallotus (Euphorbiaceae) (Kulju et al. 2007), Osbeckia (Melastomataceae) (Renner & Meyer 2001; Renner 2004). The African genus Renealmia hasbeen shown to have very recent diversification in the Neotropics (Särkinen et al. 2007) which is congruent with the Pliocene date in this study for the split of African and Neotropical Renealmia. A similar, recent scenario applies to 155 the diversification of African Aframomum, which also has evolved from a Sri Lankan ancestor during the late Oligocene. Aframomum are dispersed by primates and mammals which feed on fruits and seeds, leading to them potentially being carried for long distances (Wrangham et al. 1994) and they are well adapted in exploiting new habitats and changing forest environments during Pleistocene (Harris et al. 2000). Thus the rapid diversification in these rain forest taxa should be a result of speciation in response to Pleistocene climatic fluctuations, which were severe in Africa, also evident in some other families like Begoniaceae (Plana et al. 2004). Short distance dispersals of these species is facillitated by ants, however birds such as toucans and also mammals can transport seeds over longer distances (García-Robledo & Kuprewicz 2009). Such LDD events have more influence on ranges of taxa than numerous small steps available through local dispersals or range expansions after climate shifts (Nathan 2006). Renealmia is well represented both sides of the Atlantic with c. 15 species in Africa and 61 in the Neotropics (Maas & Maas 1987; Maas & Maas 1990). Due to the young ages of the lineages, its amphi Atlantic distribution does not favour either the Gondwanan vicariance, dispersal via amphi-Atlantic land-bridges or the boreotropical dispersal route which could have had occurred approximately 35 Mya (Särkinen et al. 2007). Thus, more recent oceanic long distance dispersal is the most plausible explanation for the dispersal from Africa to South America during the Miocene or Pliocene (15.8–2.7 Mya ago) and further speciation more recently during Pleistocene. The origin of the Sri Lankan Zingiberaceae post dates a Gondwanan origin and shows a more recent entry to the island and diversification of lineages. My results further support the current view on long distance dispersals in being pivotal to the evolution of the world’s tropical flora, in contrast to earlier views regarding the vicariance paradigm. Further it provides support for a growing body of evidence that the rainforests are youthful at the species level while they act as museums for higher level taxa to conserve traits over a long period of time. 156 References Aitchison, J.C., Ali, J.R. & Davis, A.M., 2008. When and where did India and Asia collide? Geological Bulletin of China, 27(9), pp.1351–1370. Ali, J.R. & Aitchison, J.C., 2008. Gondwana to Asia: Plate tectonics, paleogeography and the biological connectivity of the Indian sub-continent from the Middle Jurassic through latest Eocene (166-35??Ma). Earth-Science Reviews, 88(3-4), pp.145–166. Andrews, P. & van Couvering, J.A.H., 1975. Palaeoenviron ments in the East African Miocene. Approaches to primate paleobiology (ed. by X. Szalay), pp. 62–103. Contribution to Primate Biology, Vol. 5, Karger, Basel. Ashton, P.S. & Gunatilleke, C.V.S., 1987. New light on the plant geography of Ceylon I . Historical plant geography. Journal of Biogeography, 14(3), pp.249–285. Batalha-Filho, H. et al., 2014. Phylogeny and historical biogeography of gnateaters (Passeriformes, Conopophagidae) in the South America forests. Molecular Phylogenetics and Evolution, 79, pp.422–432. Bhadra, S. & Bandyopadhyay, M., 2016. New chromosome number counts and karyotype analysis in three important genera of Zingiberaceae. Nucleus, 59(1), pp.35-40. Briggs, J.C., Collections, Z. & Museum, G., 2003. The biogeographic and tectonic history of India. Journal of Biogeography, 30, pp.381–388. Cannon, C.H., Morley, R.J. & Bush, A.B.G., 2009. The current refugial rainforests of Sundaland are unrepresentative of their biogeographic past and highly vulnerable to disturbance. Proceedings of the National Academy of Sciences of the United States of America, 106(27), pp.11188–11193. Chen, X. & Xia, N., 2013. Chromosome Numbers and Ploidy Levels of Chinese Curcuma Species. HORTSCIENCE, 48(5), pp.525-530. 157 Clark, J.R. et al., 2008. A comparative study in ancestral range reconstruction methods: retracing the uncertain histories of insular lineages. Systematic biology, 57(5), pp.693–707. Coetzee, J.A., 1978. Climatic and biological changes in south- western Africa during the late Cainozoic. Palaeoecology of Africa, 10, pp.13–29. Crayn, D.M., Costion, C. & Harrington, M.G., 2014. The Sahul – Sunda floristic exchange : dated molecular phylogenies document Cenozoic intercontinental dispersal dynamics. Journal of Biogeography, 42, pp.11–24. Droop A.J. 2012. Systematic and biogeographic studies in the genus Amomum Roxb . ( Zingiberaceae ) in Sumatra. PhD thesis, University of Aberdeen and Royal Botanic Gardens Edinburgh, February 2012. Drummond, A.J. & Rambaut, A., 2007. BEAST: Bayesian evolutionary analysis by sampling trees. BMC evolutionary biology, 7, p.214. García-Robledo, C. & Kuprewicz, E.K., 2009. Vertebrate Fruit Removal and Ant Seed Dispersal in the Neotropical Ginger Renealmia alpinia (Zingiberaceae). Biotropica, 41(2), pp.209–214. Greenwood, D.R. & Christophel, D.C., 2005. The origins and Tertiary history of Australian “tropical” rainforests. Tropical rainforests: past, present, and future (ed. by E. Ber- mingham, C.W. Dick and C. Moritz), pp. 336–373. Chi- cago University Press, Chicago. Hall, R., 2001. Cenozoic reconstructions of SE Asia and the SW Pacific: Changing patterns of land and sea. Faunal and Floral Migrations and Evolution in Se AsiaAustralasia, (Figure 1), pp.35–56. Hall, R., 2009. Southeast Asia’s changing palaeogeography. Blumea - Biodiversity, Evolution and Biogeography of Plants, 54(1), pp.148–161. 158 Hall, T., 1999. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symposium Series, 41, pp.95–98. Harris, D. J., et al., 2000. Rapid radiation in Aframomum (Zingiberaceae): evidence from nuclear ribosomal DNA internal transcribed spacer (ITS) sequences. Edinburgh Journal of Botany, 57, pp.377–395. Jones, A.W. & Kennedy, R.S., 2008. Evolution in a tropical archipelago: comparative phylogeography of Philippine fauna and flora reveals complex patterns of colonization and diversification. Biological Journal of the Linnean Society, 95, pp.620–639. Kearse, M. et al., 2012. Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics, 28(12), pp.1647-1649. Kershaw, A.P. et al., 1994. The Neogene: a period of transition. History of the Australian vegetation: Cretaceous to Recent (ed. by R.S. Hill), pp. 299–327. Cambridge University Press, Cambridge, UK. Kishore, K. et al., 2011. Pollination studies in large cardamom (Amomum subulatum Roxb.) of Sikkim Himalayan region of India. Scientia Horticulturae, 129, pp.735– 74 Kress, W. J., 1995. Phylogeny of the Zingiberanae: Morphology and molecules. Pages 443–460 inMono- cotyledons: Systematics and evolution (P. Rudall, P. J. Cribb, D. F. Cutler, and C. J. Humphries, eds.). Royal Botanic Gardens, Kew, U.K. Kress, W.J. & Specht, C.D., 2006. The evolutionary and biogeographic origin and diversification of the tropical monocot Oder Zingiberales. Aliso, 22, pp.621–632. Kress, W.J. et al., 2001. Unraveling the evolutionary radiation of the families of the Zingiberales using morphological and molecular evidence. Systematic Biology, 50(6):926–944. 159 Kress, W.J. et al., 2005. The molecular phylogeny of Alpinia (Zingiberaceae): A complex and polyphyletic genus of gingers. American Journal of Botany, 92(1), pp.167–178. Kress, W.J., 2002. The phylogeny and a new classification of the gingers (Zingiberaceae): evidence from molecular data. Amer. J. Bot, 89, pp.1682–1696. Kulju, K.K.M. et al., 2007. Molecular phylogeny of Macaranga, Mallotus, and related genera (Euphorbiaceae s.s.): Insights from plastid and nuclear DNA sequence data. American Journal of Botany, 94(10), pp.1726–1743. Larsen, K., 2005. Distribution patterns and diversity centres of Zingiberaceae in SE Asia. Biologiske Skrifter, 55, pp.219–228, 2005. Li, Y. et al., 2009. More Miocene Dispersal Between Africa and Asia—the Case of Bridelia (Phyllanthaceae). Systematic Botany, 34(3), pp.521–529. Manos, P.S., & Steele, K.P., 1997. Phylogenetic analyses of ‘‘higher’’ Hamamelididae based on plastid sequence data. American Journal of Botany, 84, pp.1407–1419. Maas, P.J.M., 1977. Flora Neotropica Monograph No. 18: Renealmia (Zingigeraceae– Zingiberoideae) Costoideae (Additions) (Zingibera- ceae). New York Botanical Garden, New York. Maas, P.J.M., Maas, H., 1987. Notes on New World Zingiberaceae: III— Some new species in Renealmia. Notes Roy. Bot. Gard. Edinburgh, 44, pp.237–248. Maas, P.J.M., Maas, H., 1990. Notes on New World Zingiberaceae: IV— Some new species of Costus and Renealmia. Notes Roy. Bot. Gard. Edinburgh, pp. 46, 307– 320. Matzke, N.J., 2013. Probabilistic historical biogeography: new models for founder‐ event speciation, imperfect detection, and fossils allow im‐proved accuracy and model‐testing. Frontiers of biogeography, 5(4), pp.242-248. 160 Matzke, N.J., 2014. Model Selection in Historical Biogeography Reveals that FounderEvent Speciation Is a Crucial Process in Island Clades. Systematic Biology, 63(6), pp.951–970. Miller, M.A. et al., 2015. A RESTful API for Access to Phylogenetic Tools via the CIPRES Science Gateway. Evolutionary Bioinformatics, 11, pp,43-48. Morley, R. J., 2000. Origin and evolution of tropical rain forests. Chichester: John Wiley & Sons, Ltd. Morley, R.J., 2003. Interplate dispersal paths for megathermal angiosperms. Perspectives in Plant Ecology, Evolution and Systematics, 6(12), pp.5–20. Nagamasu, H. & Sakai, S., 1996. Amomum roseisquamosum (Zingiber- aceae), a new epiphytic ginger from Borneo. Edinburgh Journal of Botany, 53(1), pp.39–42. Nathan, R., 2006. Long-distance dispersal of plants. Science (New York, N.Y.), 313(August), pp.786–788. Okuyama, Y., et al., 2005. Nonuniform concerted evolution and chloroplast capture: Heterogeneity of observed introgression patterns in three molecular data partition phylogenies of Asian Mitella (Saxifragaceae). Molec. Biol. Evol. 22, pp. 285-296. Peppe, D.J. et al., 2007. Fossil leaf species from the Fox Hills Formation (Upper Cretaceous: North Dakota, USA) and their paleogeographic significance. Journal of Paleontology, 81(3), pp.550–567. Pfeiffer et al., 2004. Myrmecochory in the Zingiberaceae: seed removal of Globba franciscii and G. propinqua by ants (Hymenoptera-Formicidae) in rain forests on Borneo. Journal of Tropical Ecology, 20, pp.705-708. Plana, V. et al., 2004. Pleistocene and pre-Pleistocene Begonia speciation in Africa. Molecular Phylogenetics and Evolution, 31, pp.449–461. Posada, D. et al., 2012. jModelTest 2: more models, new heuristics and parallel computing. Nature Methods, 9, pp.772–772. 161 Rambaut, A., 2009. FigTree, a graphical viewer of phylogenetic trees. Institute of Evolutionary Biology University of Edinburgh. Ree, R.H. et al., 2005. A likelihood framework for inferring the evolution of geographic range on phylogenetic trees. Evolution; international journal of organic evolution, 59(11), pp.2299–2311. Ree, R.H. & Smith, S. a, 2008. Maximum likelihood inference of geographic range evolution by dispersal, local extinction, and cladogenesis. Systematic biology, 57(1), pp.4–14. Ren, P.Y. et al., 2007. An example of flexistyly in a wild cardamom species (Amomum maximum (Zingiberaceae). Plant Systematics and Evolution, 267, pp.147–154. Renner, S.S., 2004. Multiple Miocene Melastomataceae dispersal between Madagascar, Africa and India. Philosophical Transactions of the Royal Society B: Biological Sciences, 359(1450), pp.1485–1494. Renner, S.S. & Meyer, K., 2001. Melastomeae come full circle: biogeographic reconstruction and molecular clock dating. Evolution; international journal of organic evolution, 55(7), pp.1315–1324. Richardson, J.E. et al., 2014. The influence of tectonics, sea-level changes and dispersal on migration and diversification of Isonandreae (Sapotaceae). Botanical Journal of the Linnean Society, 174(1), pp.130–140. Rieseberg, W. et al., 1996. Molecular marker incongruence in plant hybrid zomes and phylogenetic trees. Acta Bot. Neerl. 45, pp. 243-262. Ronquist, F., 1997. Dispersal-vicariance analysis: A new approach to the quantification of historical biogeography. Systematic Biology, 46(l), pp.195–203. Ronquist, F. et al., 2012. MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space. Systematic Biology, 61, pp.539–42. 162 Sakai, S. et al., 1999. Three pollination guilds and variation in floral characteristics of Bornean gingers (Zingiberaceae and Costaceae). American Journal of Botany, 86(5), pp.646–658. Särkinen, T.E. et al., 2007. Recent oceanic long-distance dispersal and divergence in the amphi-Atlantic rain forest genus Renealmia L.f. (Zingiberaceae). Molecular Phylogenetics and Evolution, 44, pp.968–980. Sathiamurthy, E. & Voris, K.H., 2006. Maps of Holocene Sea Level Transgression and Submerged Lakes on the Sunda Shelf. Nat Hist J Chulalongkorn University, Supplement, 2, pp.1–43. Maps available at http://fmnh.org/research_collections/zoology/zoo_sites/seamaps/. Selvaraj et al., 2008. Phylogenetic analysis of chloroplast matK gene from Zingiberaceae for plant DNA barcoding. Bioinformation, 3(1), pp. 24-27. Steele & Vilgalys, 1994.Phylogenetic analyses of Polemoniaceae using nucleotide sequences of the Plastid gene matK.Systematic Botany, 19, pp.126-142. Stewart, C.B. & Disotell, T.R., 1998. Primate evolution - in and out of Africa. Current Biology, 8(16), pp.R582–588. Thomas, D.C. et al., 2012. West to east dispersal and subsequent rapid diversification of the mega-diverse genus Begonia (Begoniaceae) in the Malesian archipelago. Journal of Biogeography, 39(1), pp.98–113. Voris, H.K., 2000. Maps of Pleistocene sea levels in Southeast Asia: shorelines, river systems and time durations. Journal of Biogeography, 27, pp.1153–1167. White, T.J., 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis MA, Gelfand DH, Sninsky JJ, White TJ, eds. PCR protocols: a guide to methods and applications. London: Academic Press, pp.315322. 163 Woodruff, D.S., 2010. Biogeography and conservation in Southeast Asia: how 2.7 million years of repeated environmental fluctuations affect today’s patterns and the future of the remaining refugial-phase biodiversity. Biodiversity and Conservation, 19(4), pp.919–941. Xia, Y. M. et al., 2004. A phylogenetic analysis of Amomum (Alpinioideae: Zingiberaceae) using ITS and matK DNA sequence data. Syst. Bot. 29, pp.334– 344. Yuan, Y.-M. et al., 2005. Phylogeny and biogeography of exacum (gentianaceae): a disjunctive distribution in the Indian ocean basin resulted from long distance dispersal and extensive radiation. Systematic biology, 54(1), pp.21–34. Zachos, J. et al., 2001. Trends, rhythms, and aberrations in global climate 65 Ma to present. Science (New York, N.Y.), 292(April), pp.686–693. Zaveska, E. et al., 2012. Phylogeny of Curcuma (Zingiberaceae) based on plastid and nuclear sequences: Proposal of the new subgenus Ecomata. Taxon, 61(4), pp.747763. Zhang, L.B. & Renner, S., 2003. The deepest splits in Chloranthaceae as resolved by chloroplast sequences. Int. J. Plant Sci. 164 (5 Suppl.), S383–S392. Zhou, L. et al., 2012. “Out-of-Africa” dispersal of tropical floras during the Miocene climatic optimum: Evidence from Uvaria (Annonaceae). Journal of Biogeography, 39(2), pp.322–335. 164 CHAPTER 6: Summary and Conclusions During the present study 42 species (85.7% sampling from Sri Lankan representatives) were collected from three families from different pre identified localities in Sri Lanka. The collection list and the newly generated sequences are listed in tables 6.1 and 6.2 respectively. Table 6.1 Species collected in Begoniaceae, Sapotaceae and Zingiberaceae during the field work in Sri Lanka. Endemic species are indicated in bold. Family Genera Species present Species collected % sampling Begoniaceae Begonia Begonia tenera Begonia tenera Endemic-100% Begonia thwaitesii Begonia thwaitesii Native- 75% Begonia cordifolia Begonia cordifolia Begonia dipetala Begonia dipetala Begonia malabarica Begonia malabarica Begonia subpeltata Sapotaceae Palaquium Isonandra Madhuca Endemic-100% Palaquium pauciflorum Palaquium pauciflorum Palaquium thwaitesii Palaquium thwaitesii Palaquium grande Palaquium grande Palaquium hinmolpedda Palaquium hinmolpedda Palaquium zeylanicum Palaquium zeylanicum Palaquium rubiginosum Palaquium rubiginosum Palaquium petiolare Palaquium petiolare Palaquium laevifolium Palaquium laevifolium Palaquium canaliculatum Palaquium canaliculatum Isonandra lanceolata Isonandra lanceolata Isonandra compta Isonandra compta Isonandra montana Isonandra montana Isonandra zeylanica Isonandra zeylanica Madhuca fulva Madhuca fulva Endemic- 60% Madhuca moonii Madhuca clavata Native- 100% 165 Madhuca microphylla Madhuca neriifolia Madhuca clavata Madhuca longifolia Madhuca neriifolia Madhuca longifolia Zingiberaceae Manilkara Manilkara hexandra Manilkara hexandra Native- 100% Mimusops Mimusops elengi Mimusops elengi Native – 100% Zingiber Zingiber cylindricum Zingiber cylindricum Endemic-100% Zingiber wightianum Zingiber wightianum Native - 100% Curcuma albiflora Curcuma albiflora Endemic-100% Curcuma aromatica Curcuma zedoaria Native – 66% Curcuma zedoaria Curcuma oligantha Curcuma Curcuma oligantha Ammomum Amomum nemorale Amomum nemorale Endemic- 75% Amomum graminifolium Amomum graminifolium Native - 100% Amomum acuminatum Amomum acuminatum Amomum echinocarpum Amomum fulviceps Amomum fulviceps Amomum masticatorium Amomum masticatorium Amomum pterocarpum Amomum pterocarpum Alpinia Alpinia fax Alpinia fax Endemic - 66% Alpinia abundiflora Alpinia abundiflora Native - 100% Alpinia rufescence Alpinia nigra Alpinia nigra Alpinia malaccensis Alpinia malaccensis Elettaria Elettaria cardamomum Elettaria cardamomum Native 100% Hedichyum Hedichyum coronarium Hedichyum coronarium Native 50% Cyphostigma pulchellum Endemic-100% Hedichyum flavecens Cyphostigma Cyphostigma pulchellum 166 Table 6.2 Newly generated sequences for the present study for Begoniaceae, Sapotaceae and Zingiberaceae ndhA intron ndhF -rpl32 Rpl32 -trnL trnH psbA trnC trnD ndh F Family: Begoniaceae    B. cordifolia Thwaites    B. thwaitesii Hook    B. dipetala Graham    B. malabarica Lam.    B. tenera Dryand    B. albo-coccinea Hook    B. puspitae Ardi    B. sublobata Jack    B. spec.    B. sect. Reichenheimea    B. sect. Reichenheimea    B. sect. Reichenheimea    B. picta Sm.    B. tribenensis C.R.Rao    B. bryophila Ined.    B. rubella Ham. Ex D.Don    B. hatacoa Buch.Ham.    B. panchtharensis S.Rajbh.    B. flagellaris Hara    B. diocia Ham. Ex D.Don    Madhuca elmeri Merr. Ex. H.J.Lam    Madhuca laurifolia H.J.Lam    Madhuca motleyana Baehni    Species name ITS 5p-8p Family: Sapotaceae 167 ITS mat K Palaquium rostratum 2 Burck    Palaquium sericeum H.J.Lam    Payena maingayi 2 C.B.Clarke    Payena lucida 2A.d.Candolle    Madhuca sp.    Payena lucida 3A.d.Candolle    Madhuca longifolia J.F.Macbr.    Palaquium microphyllum King & Gamble    Payena obscura Burck    Madhuca kingiana 2 H.J.Lam    Payena leerii Kurz    Palaquium obovatum H.J.Lam    Palaquium sp.    Palaquium sp. 2    Payena sp    Madhuca sp.    Palaquium sericeum H.J.Lam    Palaquium clarkeanum King & Gamble    Palaquium gutta Baillon    Palaquium obovatum Engler    Palaquium oxleyanum Pierre    Palaquium xanthochymum Pierre ex Burck    Palaquium formosanum Hayata    Palaquium amboinense 3 Burck    Palaquium beccarianum van Royen    Payena leerii Kurz    168 Madhuca kingiana 1 H.J.Lam    Palaquium sumatranum Burck    Palaquium rostratum 1 Burck    Palaquium maingayi King & Gamble    Madhuca malaccensis H.J.Lam    Payena obscura Burck    Madhuca fulva J.F. Macber    Palaquium galactoxylum H.J.Lam    Palaquium sp. 1    Pouteria maclayana Baehni    Palaquium calophyllum Pierre ex Burck    Palaquium eriocalyx H.J.Lam    Palaquium rigidum Pierre ex Dubard    Palaquium leiocarpum 1 Boeriage    Palaquium pseudorostratum H.J.Lam    Palaquium ridleyi King & Gamble    Palaquium rufolanigerum P.Royen    Madhuca kuchingensis Yii & P.Chai    Palaquium walsurifolium Pierre ex Dubard    Madhuca sarawakensis H.J.Lam    Madhuca erythrophylla H.J.Lam    Madhuca kunstleri H.J.Lam    169 Palaquium calophyllum Pierre ex Burck    Palaquium herveyi King & Gamble    Palaquium hexandrum Baillon    Madhuca barbata T.D.Penn.    Palaquium cryptocariifolium P.Royen    Madhuca sericea H.J.Lam    Madhuca oblongifolia Merrill    Madhuca korthalsii H.J.Lam    Madhuca proxila Yii & P.Chai    Madhuca lancifolia H.J.Lam    Madhuca sp. nov    Palaquium amboinense 2 Burck    Madhuca pachyphylla(K.Krause) ined.    Burckella polymera P.Royen    Pouteria firma(Miq.) Baehni    Palaquium quercifolium Burck                            Palaquium_laevifolium_SL Isonandra_lanceolata_SL Wight Madhuca_sp_SL Madhuca_neriifolia_SL H.J.Lam Madhuca_longifolia2_SL J.F.macber Palaquium_thwaitesii_SL Trim. Palaquium_canaliculatum_SL Engl.  Palaquium_hinmolpedda_SL P.Royen 170 Palaquium_pauciflorum_SL                                                                 Engl. Palaquium_rubiginosum_SL Engl. Isonandra_montana_SL Gamble Isonandra_sp1_SL Palaquium_sp3_SL Palaquium_grande_SL Engl. Palaquium_petiolare_SL Engl. Palaquium_sp5_SL Isonandra_sp2_SL Isonandra_compta_SL Dubard Isonandra_zeylanica_SL Jeuken Manilkara_hexandra_SL Dubard Mimusops_elengi_SL L. Madhuca_clavata_SL Jayas. Palaquium_sp4_SL Isonandra_sp2_SL Family: Zingiberaceae Zingiber_cylindricum Moon   Amomum_pterocarpum Thwaites   Zingiber_wightianum Thwaites   Amomum_sp1   Elettaria_sp   Curcuma_zedoaria Roxb.   Amomum_fulviceps Thwaites   Amomum_sp2   Curcuma_albiflora Thwaites   Amomum_graminifolium   171 Thwaites Amomum_sp3   Amomum_nemorale Benth. &   Cyphostigma_pulchellum Benth.   Amomum_acuminatum Thwaites   Amomum_masticatorium     Amomum_sp5   Amomum_sp6   Amomum_sp7   Alpinia_sp   Alpinia_fax Burtt & R.M.Sm.   Hook.f. Thwaites Alpinia_abundiflora1 Burtt & R.M.Sm. The present study supports the importance of LDD over vicariance as the major cause for extant tropical plant disjunctions. The dated phylogenies show that the LDD events are not as infrequent as once thought, and that many tropical lineages have evolved recently. The dated phylogenies and divergence time estimates of the present study suggest recent (Oligocene to Pleistocene) origins and diversification of Sri Lankan families tested; Begoniaceae, Sapotaceae and Zingiberaceae post dating a Gondwanan origin. The family Begoniaceae shows an origin from an African ancestor and entry to Asia during the Miocene, and a point of entry tentatively reconstructed as India. The timing of entry to India via a dry Arabian corridor is not favored for Begonia, thus long distance dispersal is the acceptable hypothesis for the entry to Asia (Thomas et al. 2012). Overland migration explains dispersal from continental Asia to the South East Asian region and the predominant directional trend of dispersal is from west to East. However there is back dispersal from China-Indochina to Himalayan region which initiated during late Miocene with the onset of Asian monsoons at 7.4 Mya (Copeland 172 1997). Himalayan Begonia shows a further diversification event during 5.2 (3.1-7.9) Mya, and further back and forth dispersal could have occurred within the region. The position of Himalayan B. diocia in the phylogeny is enigmatic, it has a relatively old age (15.8 Mya) and potentially an area of origin in the region of the Indian Subcontinent. Indian ancestral lineages subsequently dispersed to Sri Lanka and Socotran islands, whose species form the western limits of Asian Begonia. Sri Lankan Begonia shows a more recent Pleistocene entry followed by more recent in-situ speciation of the endemics B. tenera and B. thwaitesii. Within the family Sapotaceae, sub family Chrysophylloideae shows an ancient origin in Africa during the Cretaceous and a single vicariance event between South America and Australia is postulated early in the evolution of the group (Bartish et al. 2011). However Long distance dispersal has played a major role in the assembly of the current hotspots of Chrysophylloideae, between Africa to the Neotropics, between Australia and New Caledonia and between Africa and Madagascar and a single dispersal back to Africa from the Neotropics (Bartish et al. 2011). Within the sub family Sapotoideae, the biogeographic reconstructions of the genus Sideroxylon suggest a Northern hemisphere origin for tribe Sideroxyleae in the early Tertiary and subsequent spread to extant distributions via the North Atlantic Land Bridge during the early Eocene (Smedmark & Anderberg 2007). The present study focused on the biogeographic reconstruction of the Sri Lankan Sapotaceae with a high emphasis for tribe Isonandrae, which is well represented in Sri Lanka. My results confirm the origin of Indo Sri Lankan Manilkara hexandra as a long distance dispersal from an African ancestor during the Miocene and Mimusops elengi from Africa to Indo-Sri Lanka with subsequent spread eastward into Malesia. 173 Figure 6.1Different geographic regions in world aand collision times between continental fragments. Excerpted from McLoughlin 2001. Figure 6.2 Present positions of continental plates and geographic boundaries. 174 Sri Lankan Isonandrae are of Sunda Shelf in origin, and have resulted from six independent long distance dispersals which occurred during the Eocene-Miocene period. The close proximity between the Malay Peninsula and the Deccan plate during its Northward migration provided many opportunities for floral exchange between two landmasses (Hall 2001). However the molecular age estimates post date this dispersal phase, thus probably have occurred after the Deccan plate collided with Laurasia during 45-50 Mya. The collision with Laurasian coast brought India and South East Asia to similar latitudes and within the same climatic belt, thus there is a sudden appearance of Indian taxa in South East Asian Palaeo floras (Morley 2003; Morley 2000). Further, there are two dispersals out of Sri Lanka evidenced, Madhuca longifloia to India and Madhuca hainanensis to Eastern Asia during early-Mid Oligocene. More rapid radiation of endemics have occurred as a result of the changing climatic conditions during glacial inter glacial periods during Pliocene-Pleistocene, giving rise to ten endemic species in Sri Lanka during the Pleistocene. Within Zingiberaceae, Siphonochilus (J. M. Wood & Franks) found in Africa while Tamijia (S. Sakai & Nagam.) in Borneo. Zingiberaceae have dispersed into equatorial regions in South East Asia as results of global cooling following cretaceous thermal maximum (Kress & Specht 2006). The continental Asia harbors the highest diversity with Alpinoideae having its centre of diversification in the Malesian region while Zingiberoideae has itsin the monsoonal areas of Indochina and Thailand (Larsen 2005). Within Sri Lankan Zingiberaceae, the species in the genera Curcuma, Zingiber and Alpinia shows an Eastern Asian origin during Oligo-Miocene boundary, thus provide an evidence for overland migration into Sri Lanka via India. Sri Lankan Amomum has an origin in Sunda shelf by three independent dispersals during mid-miocene-pliocene and Amomum endemics are more recent Pleistocene speciation within the island. Sri Lanka is constructed as the ancestral area for African/ Madagascan Aframomum andAfrican Renealmia, which shows an entry to Africa during Oligocene. The collision of the Afro-Arabian plate with Asia approximately 25 Mya (Kulju et al. 2007; Yuan et al. 2005) might have facilitated overland migration of taxa from Asia to Africa from a Sri Lankan ancestor; however long distance over water dispersal from Sri Lanka directly to Africa is another possibility which could have occurred at any point of time. 175 Neotropical Renealmia have originated from African ancestor via oceanic long distance dispersal during Miocene-Pleistocene and further speciation more recently during Pleistocene. Table 6.3A summary for reconstructed geographic origins for the tested families, and the number of dispersals during different geological epochs. Family Origin Era No of dispersals Age (Mya) Begoniaceae India Late Miocene 1 7.68 (3.91-12.62) Sapotaceae Africa Early Miocene 1 19.04 (11.3-25.52) Late Miocene 1 6.81 (3.47-11.12) Eocene 1 36.91 (27.76-47.25) Oligocene 4 31.93 (22.88-41.43) Sunda Shelf 28.49 33.14 (25-40.02) 31.53 Zingiberaceae Eastern Asia Mid Miocene 1 11.49 Oligocene 2 25.7 (15.63-36.72) 27.31 Sunda Shelf Early Miocene 1 18.74 (10.24-29.2) Pliocene 1 3.76 (1.34-7.13) Mid Miocene 1 11.8 (6.3-19.4) Late Miocene 1 10.2 Pliocene 1 3.8 (1-7.4) The dates of origin for Sri Lankan lineages considerably post-date the Gondwanan break up, instead suggesting the importance of more recent transoceanic long distance dispersal for their origin and diversification within the island. Among tested plant families, dispersals have occurred from Africa, India, Eastern Asia and Sunda Shelf into Sri Lanka. Much of the Sri Lankan flora shows an origin in the Sunda Shelf (53%), especially genera Madhuca, Palaquium, Isonandra in the family Sapotaceae and the genus Amomum in the family Zingiberaceae. During the Deccan plate’s northward migration, it moved very close to Peninsular Malaysia and had a glancing contact with Sumatra 176 during the late Paleocene. This could have facilitated plant dispersals between the Deccan plate and South East Asia. However Sri Lankan lineages post date this dispersal phase, as all have taken place after the Deccan plate collided with the Laurasian coast, which brought them to same altitudes and climatic belt. Africa India Eastern Asia Sunda shelf Figure 6.3 Reconstructed geographic origins for Sri Lankan flora and their relative contributions. Sri Lankan flora has also been influenced by the Eastern Asian flora (20%), especially with regard to reconstructing as the ancestral area for some genera in family Zingiberaceae. After the Deccan plate collided with the Laurasian coast during the Eocene there would have been many opportunities for mixing of the Laurasian flora with the Deccan Gondwanan flora. Himalayan uplift resulted in numerous climatic and geological changes thus had a profound effect in the flora of that region. Africa contributes 11% of the lineages observed to the Sri Lankan flora. My results post date the late Cretaceous dispersal phase, which is accepted as the major phase for AfroIndo Sri Lankan disjunctions evidenced in some families like Crypterionaceae (Conti et al. 2002; Ashton & Gunatilleke 1987) and Dipterocarpaceae. Instead they are much younger indicating more recent trans-oceanic long distance dispersal events from Africa to the Indo Sri Lankan region. India accounts for only 12% of the Sri Lankan families tested, despite the long proximities between the two land masses since Jurrasic times where they are separated by the narrow Palk Strait 32 km in length. The intervening continental shelf between Sri Lanka and India provided an intermittent land connection until the last sea level rise 177 6000 years ago during the Holocene (McLoughlin 2001; Ashton & Gunatilleke 1987) and has potential to act as a migratory path for floristic exchange between the two land masses. However as the endemic Sri Lanka Begonia are restricted to altitudes of 10001200m, the overland migration via a lowland land bridge seems unlikely. The Western Ghats of India and the sub montane forests in Sri Lanka could have act as archipelagolike systems, facilitating plant dispersals among them. Of the four regions, the Sunda Shelf is overwhelmingly the main source for Sri Lankan floral elements, despite the geographic proximity of India. It would seen that climatic similarity between the perhumid forests of Sri Lanka and those of the Sunda Shelf has a lot of influence on successful immigration. eocene oligocene miocene pliocene pleistocene Figure 6.4 Percentage of dispersals in terms of geological epochs. Among the families tested, dispersals have occurred stochastically, one during the Eocene, six during the Oligocene, seven during the Miocene, two during the Pliocene and one during the Pleistocene (Figure 6.4). The highest number of dispersals occurred during the Miocene when a warm climate was prevailing during the Miocene thermal maximum (Zachos et al. 2001). The Deccan plate, composed of India and Sri Lanka, has probably undergone more changes than any other tropical region since it began to drift away from Antarctica during Cretaceous (Morley 2000). During its northward journey, as it entered the southern hemisphere high pressure zone, its Gondwanan flora was replaced by a tropical African flora. During the Paleocene, moist equatorial climate resulted in the evolution 178 of new lineages, with a rapidly diversifying mega thermal flora being dominant. The plate was at the equator during the early Eocene and extremely wet climates could have provided optimum climatic conditions for angiosperm proliferation, thus fully covered by multistoried rain forests. Warm, perhumid conditions during Late Eocene to Oligocene further nourished rain forest taxa. It is accepted that humid, tropical climate persisted in the Deccan plate since it collided with Eurasia, which is comparable to climate in India and Sri Lanka at present (Ashton & Gunatilleke 1987; Morley 2000). The massive volcanism at the Cretaceous-Tertiary boundary during 65 Ma and extensive aridification during early Tertiary due to the uplift of Himalaya resulted in further impoverishment of allochthonous African elements from the Indian flora (Rutschmann & Eriksson 2004; Morley 2003; Conti et al. 2002; McLoughlin 2001), leaving the Western Ghats and Sri Lanka as refugial areas for those flora. Further, my results confirm that in situ speciation is an important contributor to the Sri Lankan flora. More rapid radiation of endemics has occurred during PliocenePleistocene; two endemics in Begoniaceae, ten endemics in Sapotaceae and ten endemics in Zingiberaceae have evolved in situ during this period. Sri Lanka will have been subjected to expansion and contraction of climatic and vegetation zones within the island during glacial and interglacial periods, potentially resulting in allopatric speciation. The results provide strong evidence for youthful tropics which is congruent with the view of Koenen et al. (2015) on the evolution of rain forest species in the Neotropics. Rain forests have been acted as museums for higher level taxa, conserving diverse traits within them ultimately giving rise to new species (Pennington et al. 2015). A similar scenario of recent speciation also evident in Africa (Harris et al. 2000) and Asia (Thomas et al. 2012). Among birds and mammals, a recent Pleistocene speciation has been evidenced at higher altitudes; however it is throughout Pre-Pleistocene to Pleistocene towards the equator (Weir et al. 2007). 179 Figure 6.5 A plot of the minimum ages of Sri Lankan clades (25) from the present study and other published studies on plants (Hortonia) and animal groups (toads) arranged from oldest (left) to youngest. The error bars indicate the uncertainty around the minimum age estimates. P., Pleistocene; Pl.,Pliocene; Paleo., Paleocene The vicariance paradigm has long been accepted as the major cause for tropical disjunctions and assembly of flora land masses like Sri Lanka which was a part of Gondwana a long time ago. One striking example is the Sri Lankan endemic monotypic genus Hortonia in Monimiaceae, being the oldest lineage among Sri Lankan flora. The other genera in the family occur in America, Africa, Madagascar and The Mascarenes, New Caledonia, Australia, New Zealand and the Malesian region (Renner et al. 2010). The dated phylogeny of Monimiaceae suggests a Gondwanan origin for Sri Lankan Hortonia which dates back to 71 (57-84) Mya during the late Cretaceous. One possibility of this ancient lineage of Hortonia in Sri Lanka is explained as rafting on the Deccan plate to Asia (Ashton & Gunatilleke 1987). The other explanation is long distance dispersal of Hortonia from Antarctica to Sri Lanka, however the huge distance 180 between two land masses ca 2100 Km (Aitchison et al. 2008), raises the question of ability to cross such long distance over water according to Renner et al. (2010). A classic Gondwanan distribution is observed among some animals, such as in the fresh water Cichlids (order Perciformes: family Cichlidae). They are mainly distributed Africa, South and Central America, Madagascar, and Indo/Sri Lanka. If they have originated in Africa and migrated into South America, Madagascar, and India via saltwater dispersal, the African clade should have been constructed as the sister clade for the rest. However, more recent dated phylogeny by Azuma et al. (2008) have found dates for major phylogenetic splits which are congruent with Gondwanan breakup of landmasses. The divergence time between Malagascan and Indo/Sri Lankan taxa is 87 Mya (69–106 MYA) which is congruent with time of separation between Madagascar and India during 90 Mya (Briggs et al. 2003; Conti et al. 2002; McLoughlin 2001). Further it supports the evidence to the time of separation between African and South American landmasses which probably occurred 100 Mya with a phylogenetic split occurring between African and neotropical clades around 89 Mya (72-108 Mya) (Azuma et al. 2008). However, the present study largely contributes to the growing body of evidence in support of recent transoceanic dispersals as being a major factor in the tropical plant disjunctions seen today. More importantly it reveals the disproportionate importance of LDD in the development of the flora of small islands like Sri Lanka. LDD can happen by means of vectors such as wind, water, birds, mammals and extreme events like tropical cyclones, hurricanes and tornadoes (Nathan 2006), more importantly these climatic events have much influence on land clearing territory for new immigrants. In most tropical tree diaspores transport via ocean currents is much more feasible than dispersal by wind and Houle (1998) demonstrated that the trans-oceanic migration times are different in different water bodies which depend on paleocurrents and direction of wind (Table 6.2). Trans-oceanic dispersal by birds is also possible, however the capability for dispersal over very long distances may probably be infrequent since they void gut contents frequently and their migratory route is dominantly north-south (Fukui 2003). Further, 181 for tiny seeds it is possible to get trapped on a bird’s foot, however for tropical trees which produce large seeds, the scenario seems unlikely. Table 6.4 Hypothesized rafting times across the Atlantic Ocean, Caribbean Sea and Southeast Indian Ocean during different periods throughout the Tertiary according to Houle (1998). Ocean System Rafting time 50 Mya Rafting time 40 Mya Rafting time 30 Mya Atlantic Ocean 5.2-7.7 days 7.3-10.8 days 10-14.7 days Caribbean Sea 11.2-18.2 days 10.2-16.6 days 9.3-15.1 days SE Indian Ocean 24.2-25.6 days 18.4-19.5 days 11.5-12.2 days Although such dispersals are considered as rare, long jumps available through LDD have much influence on ranges of taxa than numerous small steps available through local dispersals or range expansions after climate shifts. Numerous recent dated phylogenies evoke the importance of relatively recent transoceanic dispersals on shaping the world’s tropical flora such as: Dispersals between Asia and Africa: Macaranga and Mallotus (Euphorbiaceae) (Kulju et al. 2008), Bridelia (Phyllathaceae) (Li et al. 2009), Gaertnera (Rubiaceae) (Malcomber 2002)Exacum (Gentianaceae) (Yuan et al. 2005), Malphigiaceae (Davis 2002), Africa and Neotropics: Melastomataceae (Renner et al. 2001), Annona (Annonaceae) (Richardson et al. 2004), Commiphora (Burseraceae) (Weeks et al. 2007) Across pacific ocean Myrtaceae (Sytsma et al. 2004), Piper and Peperomia, Hernandia (Hernandeaceae) (Michalak et al. 2010). Some taxa like Piper and Peperomia (Piperaceae) have shown multiple dispersals between South America, Africa, Asia and the Pacific (Smith et al. 2008). Within Cucurbitaceae, forty-three successful longdistance dispersal events from Asia to Africa, North and South America and Australia have been reported over the last 60 Ma, an average of seven long-distance dispersals every 10 Ma (Schaefer et al. 2008). Thus, LDD’s are possibly not so infrequent as once thought before and my results add to the growing body of importance of LDD over vicariance. Further, my study provides strong evidence for a directional trend of dispersals from west to east, from the climatically similar Sunda Shelf. Future work 182 should concentrate on putting further Sri Lankan elements into dated phylogenies, and also focus on more intense sampling covering the South and eastern Asian flora to provide better geographic coverage. As a conclusion, long distance dispersals have played a prominent role in the evolution of the Sri Lankan flora. The young ages of Sri Lankan clades and also clades from other regions challenge the vicariant paradigm for their origin and provide further strong evidence for youthful tropics at the species level. References Aitchison, J.C., Ali, J.R. & Davis, A.M., 2008. When and where did India and Asia collide? Geological Bulletin of China, 27(9), pp.1351–1370. Ashton, P.S. & Gunatilleke, C.V.S., 1987. New light on the plant geography of Ceylon I . Historical plant geography. Journal of Biogeography, 14(3), pp.249–285. Azuma, Y. et al., 2008. Mitogenomic evaluation of the historical biogeography of cichlids toward reliable dating of teleostean divergences. BMC evolutionary biology, 8(1), p.215. Bartish, I. V. et al., 2011. Vicariance or long-distance dispersal: Historical biogeography of the pantropical subfamily Chrysophylloideae (Sapotaceae). Journal of Biogeography, 38(1), pp.177–190. Briggs, J.C., Collections, Z. & Museum, G., 2003. The biogeographic and tectonic history of India.Journal of Biogeography, 30, pp.381–388. Conti, E. et al., 2002. Early Tertiary out-of-India dispersal of Crypteroniaceae: evidence from phylogeny and molecular dating. Evolution; international journal of organic evolution, 56(10), pp.1931–1942. Copeland, P., 1997. The when and where of the growth of the Himalaya and the Tibetan Plateau. In: Ruddiman, W.F. (ed) Tectonic Uplift and Climate Change. New York: Plenum Press. 183 Davis, C.C. et al., 2002. Laurasian migration explains Gondwanan disjunctions: evidence from Malpighiaceae. Proceedings of the National Academy of Sciences, USA, 99, pp.6833–6837. Fukui, A., 2003. Relationship between seed retention time in bird’s gut and fruit characteristics. Ornithological Science, 2, pp.41-48. Hall, R., 2001. Cenozoic reconstructions of SE Asia and the SW Pacific: Changing patterns of land and sea. Faunal and Floral Migrations and Evolution in Se AsiaAustralasia, (Figure 1), pp.35–56. Harris, D. J. et al., 2000. Rapid radiation in Aframomum (Zin- giberaceae): evidence from nuclear ribosomal DNA internal transcribed spacer (ITS) sequences. Edinburgh Journal of Botany, 57, pp.377–395. Houle, A., 1998. Floating islands: a mode of long distance dispersal for small and medium-sized terrestrial vertebrates. Divers. Distrib, 4, pp.201–216. Koenen, E.J.M. et al., 2015. Recently evolved diversity and convergent radiations of rainforest mahoganies (Meliaceae) shed new light on the origins of rainforest hyperdiversity. New Phytol, 207, pp.327–339. Kress, J.W. & Specht, C.D., 2006. The evolutionary and biogeographic origin and diversification of the tropical monocot Oder Zingiberales. Aliso, 22, pp.621–632. Kulju, K.K.M. et al., 2007. Molecular phylogeny of Macaranga, Mallotus, and related genera (Euphorbiaceae s.s.): Insights from plastid and nuclear DNA sequence data. American Journal of Botany, 94(10), pp.1726–1743. Larsen, K., 2005. Distribution patterns and diversity centres of Zingiberaceae in SE Asia. Biologiske Skrifter, 55, pp.219–228, 2005. Li, Y. et al., 2009. More Miocene Dispersal Between Africa and Asia—the Case of Bridelia (Phyllanthaceae). Systematic Botany, 34(3), pp.521–529. 184 Malcomber, S.T., 2002. Phylogeny of Gaertnera Lam. (Rubiaceae) based on multiple DNA markers: evidence of a rapid radiation in a widespread, morphologically diverse genus. Evolution; international journal of organic evolution, 56(1), pp.42– 57. McLoughlin, S., 2001. The breakup history of Gondwana and its impact on preCenozoic floristic provincialism. Australian Journal of Botany, 49(3), pp.271–300. Michalak, I.,2010. Trans-Atlantic, trans-Pacific and trans- Indian Ocean dispersal in the small Gondwanan Laurales family Hernandiaceae. Journal of Biogeography, 37(7),pp.1214-1226. Morley, R. J., 2000. Origin and evolution of tropical rain forests. Chichester: John Wiley & Sons, Ltd. Morley, R.J., 2003. Interplate dispersal paths for megathermal angiosperms. Perspectives in Plant Ecology, Evolution and Systematics, 6(12), pp.5–20. Nathan, R., 2006. Long-distance dispersal of plants. Science (New York, N.Y.), 313(August), pp.786–788. Pennington, R.T., Hughes, M. & Moonlight, P.W., 2015. The Origins of Tropical Rainforest Hyperdiversity. Trends in Plant Science, 20(11), pp.693–695. Renner, S. S., and K. Meyer. 2001. Melastomeae come full circle: biogeographic reconstruction and molecular clock dating. Evolution, 55, pp.1315–1324. Renner, S.S. et al., 2010. Biogeography of the Monimiaceae (Laurales): A role for East Gondwana and long-distance dispersal, but not West Gondwana. Journal of Biogeography, 37(7), pp.1227–1238. Richardson, J.E. et al., 2004. Historical biogeography of two cosmopolitan families of flowering plants: Annonaceae and Rhamnaceae. Philosophical transactions of the Royal Society of London. Series B, Biological sciences, 359(1450), pp.1495–1508. 185 Rutschmann, F. & Eriksson, T., 2004. Did Crypteroniaceae really disperse out of India? Molecular dating evidence from rbcL, ndhF, and rpl16 intron sequences. International Journal of plant Science, 165(S4), pp.S69–S89. Schaefer, H. et al., 2008. a dated phylogeny reveals an Asian origin of the gourd family (Curcubitaceae) and numerous oversea dispersal events. Proceedings of the Royal Society Botany, Biological Sciences. 276, pp.843-851. Smedmark, J.E.E. & Anderberg, A. a., 2007. Boreotropical migration explains hybridization between geographically distant lineages in the pantropical clade Sideroxyleae (Sapotaceae). American Journal of Botany, 94(9), pp.1491–1505. Smith, J.F. et al., 2008. Placing the origin of two species-rich genera in the late Cretaceous with later species divergence in the Tertiary: a phylogenetic, biogeographic and molecular dating analysis of Piper and Peperomia (Piperaceae). Plant Systematics and Evolution, 275, pp.9–30. Sytsma, K.J. et al., 2004. Clades, clocks, and continents: historical and biogeographical analysis of Myrtaceae, Vochysiaceae, and relatives in the Southern Hemisphere. International Journal of Plant Sciences, 165(4 Suppl.):S85–S105. Thomas, D.C. et al., 2012. West to east dispersal and subsequent rapid diversification of the mega-diverse genus Begonia (Begoniaceae) in the Malesian archipelago. Journal of Biogeography, 39(1), pp.98–113. Weeks, A., Daly, D.C. & Simpson, B.B., 2005. The phylogenetic history and biogeography of the frankincense and myrrh family (Burseraceae) based on nuclear and chloroplast sequence data. Molecular Phylogenetics and Evolution, 35, pp.85– 101. Weir, J.T., Schluter, D. & Schl, D., 2007. The Latitudinal Gradient in Recent Rates of and Extinction Speciation Birds and Mammals. Science, 315(5818), pp.1574–1576. 186 Yuan, Y.-M. et al., 2005. Phylogeny and biogeography of exacum (gentianaceae): a disjunctive distribution in the Indian ocean basin resulted from long distance dispersal and extensive radiation. Systematic biology, 54(1), pp.21–34. Zachos, J. et al., 2001. Trends, rhythms, and aberrations in global climate 65 Ma to present. Science (New York, N.Y.), 292(April), pp.686–693. 187 Appendices Appendix1:Accessions used for the generation of DNA sequence data, Begoniaceae and EDNA numbers. Species indicated in bold are newly generated sequences for the present study. Species name Origin Voucher B. ningmingensis D.Fang, Philippine Peng P20322 (HAST) B. acuminatissima Merr. Philippine Rubite R321 (PNH) B. anisoptera Merr. Philippine Rubite R479 (PNH) B. biliranensis Merr. Philippine Rubite R311 (PNH) B. blancii M.Hughes & C.I. Philippine Peng P22545 (HAST) B. culasiensis Philippine Rubite R234 (PNH) B. calcicola Merr. Philippine Peng P20761 (HAST) B. caminguinensis Elmer Philippine Rubite R506 (PNH) B. castilloi Merr. Philippine Rubite R98 (PNH) B. chingipengii Rubite Philippine Peng P23368 (HAST) B. chloroneura P.Wilkie & Forrest, L.L. 128 (E) Sands Cultivated: RBGE (acc. num.: 19972555), Philippines, Luzon Island B. cleopatrae_1 Coyle Palawan Wilkie et al., 25373 (E) B. copelandii Merr. Luzon Adduru, M 259 (K) B. dinglensis Philippine Peng P23859 (HAST) B. elmeri Merr. Luzon Rubite R319 (PNH) B. elnidoensis ined. Philippine Peng P25308 (HAST) B. fenecis Merr. Philippine Peng P10794 (HAST) B. gabaldonensis ined. Philippine Peng P23356 (HAST) B. gitingensis Elmer Philippine Rubite R255 (PNH) B. gueritziana Gibbs Philippine Peng P22311 (HAST) Y.G. Wei & C.I.Peng Peng 188 B. gutierrezii Coyle Philippine Blanc s.n. (E) B. hernandioides_1 Merr. Cultivated: BGB (acc. num.: 00603589), Philippines Forrest, L.L. 129 (E) B. hernandioides_2 Merr. Philippine Peng P21006 (HAST) B. hughesii_1 ined. Philippine Peng P23466 (HAST) B. hughesii_2 ined. Philippine Peng P23475 (HAST) B. klemmei Merr. Philippine Rubite R182 (PNH) B. longiscapa_1 Warb. Philippine Rubite R298 (PNH) B. longiscapa_2 Warb. Philippine Rubite R309 (PNH) B. luzonensis_1 Warb. Luzon R316 (PNH) B. luzonensis_2 Warb. Luzon R420 (PNH) B. luzonensis_3 Warb. Luzon K030960 (HAST) B. manillensis_1 A.DC. Luzon Rubite R256 (PNH) B. merrilliana ined. Philippine Peng P23765 (HAST) B. mindorensis Merr. Palawan Rubite R326 (PNH) B. nigritarum_1 (Kamel) Palawan Rubite R419 (PNH) Palawan Rubite R406 (PNH) Palawan Peng P23855 (HAST) B. obtusifolia Merr. Panay Peng P23828 (HAST) B. oxysperma A.DC. Luzon Rubite R213 (PNH) B. rhombicarpa A.DC. Luzon Cuming H. 510 (E) B. rubitae M.Hughes Philippines Rubite R356 (PNH) B. rufipila Merr. Luzon Rubite R256 (PNH) Begonia sp_1 Luzon Peng P23566 (HAST) Begonia sp_2 Luzon Kokubagata GK71 (HAST) Steud. B. nigritarum_2 (Kamel) Steud. B. nigritarum_3 (Kamel) Steud. 189 B. subnummarifolia Merr. Sabah No voucher B. suborbiculata Merr. Palawan Rubite R353 (PNH) B. sykakiengii ined. Philippine Peng P23856 (HAST) B. tagbanua ined. Philippine Blanc s.n. (E) B. taraw ined. Philippine Blanc s.n. taraw2 (E) B. tayabensis Merr. Luzon Rubite R360 (PNH) B. trichocheila Warb. Luzon Peng P20764 (HAST) B. wadei Merr. & Quisumb. Palawan Rubite R699 (PNH) B. woodii Merr. Palawan Peng P23479 (HAST) B. dregei Otto & Dietr. Cultivated: RBGE (acc. num.: 20000902), South Africa McLellan, T. 415 (E) B. goudotii A.DC. Madagascar Plana, V. 120 (E) B. oxyloba Welw. ex. Hook.f. Cultivated: RBGE (acc. num.: 19982761), Tanzania Thomas, D.C. 08-141 (E) B. poculifera Hook.f. Cultivated: RBGE (acc. num.: 19923143), Cameroon Forrest, L.L. 234 (E) B. polygonoides Hook.f. Ivory Coast Van der Burg, W. J. 244 (WAG) B. samahensis M.Hughes & Thomas, D.C. 09-01 (E) A.G. Mill. Cultivated: RBGE (acc. num.: 19990412), Yemen, Socotra B. socotrana Hook.f. Yemen, Socotra Miller, A.G. 19210/10 (E) B. sutherlandii Hook.f. Cultivated: RBGE (acc, num.: 20010167), South Africa Thomas, D.C. 08-140 (E) B. boliviensis A.DC. Cultivated: GBG (acc.num.: 00801998) Bolivia No voucher available B. nelumbifolia Cham. & Cultivated: RBGE (acc. num.: 19791888), Mexico Hunt, D.R. 7516 (K) Schlecht 190 B. radicans Vell. Cultivated: GBG (acc. num.: 00908995), Brazil No voucher available B. tenera Dryand Sri Lanka Kumarage 68 (E) B. thwaitesii Hook Sri Lanka Kumarage 23 (E) B. cordifolia Thwaites Sri Lanka, India Kumarage 14 (E) B. dipetala Graham Sri Lanka, India Kumarage 25 (E) B. malabarica Lam. Sri Lanka, India Kumarage 28 (E) B. malabarica_India Lam. Cultivated: GBG (acc. num.: 00201896), India, Sri Lanka Forrest, L.L. 288 (E) B. albo-coccinea Hook India Photo voucher available B. floccifera Bedd Cultivated: GBG (acc. num.: 03009989), India Forrest, L.L. 238 (E) B. bryophila Ined. Nepal 6100 B. diocia Ham. Ex D.Don Nepal 13651 B. flagellaris Hara Nepal 6010 B. hatacoa Buch Ham. Nepal 5971 B. panchtharensis S.Rajbh. Nepal 5968 B. picta Sm. Nepal 5993 B. roxburghii A.DC. Cultivated: GBG (acc. num.: 01100797), India Thomas, D.C. 08-103 (E) B. rubella Ham. Ex D.Don Nepal 6000 B. sikkimensis A.DC. Cultivated: RBGE (acc, num.: 20051755), India Thomas, D.C. 08-144 (E) B. tribenensis C.R.Rao. Nepal 6043 B. aceroides Irmsch. Thailand Phutthai 243 (E) B. acetosella Craib Cultivated: GBG (acc, num.: 00107396), Vietnam Thomas, D.C. & Ardi, W.H. 08105 (E) B. aff. elisabethae Kiew Cultivated: RBGE (acc. Num.: 20020477), China Moller, M. 01-156B (E) 191 B. alicida C.B.Clarke Burma, Thailand Phutthai 139 (E) B. brandisiana Kurz Burma, Laos Brandis, D. 1327 (K) B. demissa Craib China Phutthai 221 (E) B. elisabethae Kiew Penisular Malesia Phutthai 239 (E) B. grandis Dryand. Cultivated: RBGE (acc. num.: 19521036), China Thomas, D.C. 08-145 (E) B. hymenophylla Gagnep. Cambodia Phutthai 232 (E) B. masoniana Irmsch. Ex Cultivated: RBGE (acc. num.: 19980075), China Thomas, D.C. 07-24 (E) B. morsei Irmsch. Cultivated: GBG (acc. num.: 19980076), China Unknown s.n. (E) B. obovoidea Craib. Thailand Phutthai 244 (E) B. palmata D.Don Cultivated: RBGE (acc. num.: 20020476), China Moller, M. 01-127 (E) B. silletensis A.DC. Cultivated: GBG (acc, num.: 00115295), China Thomas, D.C. 08-104 (E) B. sizemoreae Kiew Cultivated: GBG (acc, num.: 00101400), Vietnam Thomas, D.C. 08-111 (E) B. smithiae Geddes Thailand Chamchamroon 3662 (E) B. spec Cambodgia B. spec_China1 China Thomas 08-145 (E) B. spec_China2 China Forrest 31 (E) B. spec_Thailand1 Thailand Phutthai 195 (E) B. spec_Thailand2 Thailand Suddee 3375 (E) B. spec_Thailand3 Thailand Suddee 3371 (E) B. versicolor Irmsch. Cultivated: RBGE (acc, num.: 19980037), China Forrest, L.L. 2 (E) B. aff. congesta Ridl. Cultivated: SBG, Malaysia, Borneo Thomas, D.C. 09-05 (E) B. areolata Miq. Cultivated: BoBG, Thomas, D.C. & Ardi, W.H. 09- Ziesenh. 20100763 EDNA12-0025039 192 Indonesia, Java 137 (E) B. bracteata Jack Indonesia, Java Ardi, W.H. & Thomas, D.C. 25 (E) B. chlorosticta Sands Cultivated: SBG, Malaysia, Borneo Thomas, D.C. 09-04 (E) B. corrugata Kiew & S. Julia Cultivated: SBG, Malaysia, Borneo Thomas, D.C. 09-02 (E) B. decora Stapf Cultivated: RBGE (acc, num.: 20021608), Malaysia, Peninsul Malaysia Neale, S. 8C (E) B. fuscisetosa Sands Peninsular Malaysia 20120800 (E) B. goegoensis N.E.Br. Cultivated: GBG (acc. num.: 01112557), Indonesia, Sumatra Thomas, D.C. & Ardi, W.H. 08107 (E) B. harauensis Girm Peninsular Malaysia Thomas 09-134 (E) B. kingiana Irmsch Cultivated: GBG (acc. num.: 01807007), Malaysia, Peninsula Malaysia Thomas, D.C. 08-102 (E) B. laruei M.Hughes Indonesia, Sumatra Hughes, M. 1389 (E) B. longifolia Blume Cultivated: GBG (acc. num.: 20021848), Malaysia, Peninsula Malaysia Neale, S. 11C (E) B. multangula Blume Indonesia, Bali Thomas, D.C. & Ardi, W. H. 0890 (E) B. multijugata M. Hughes Indonesia, Sumatra Wilkie, P., Hughe, M., Sumadijaya, A., Rasnovi, S., Marlan & Suhardi 768 (E) B. muricata Blume Indonesia, Java Ardi, W.H. & Thomas, D.C. 27 (E) B. pavonina Ridl Cultivated: RBGE (acc. num.: 20021611), Malaysia, Peninsula Neale, S. 9C (E) 193 Malaysia B. puspitaeArdi Cultivated: RBGE Indonesia 20111539 B. resectaMiq. ex Koord. Indonesia Hughes 786 B. robusta Blume Cultivated: BaBG, Indonesia, Java Thomas, D.C. & Ardi, W.H. 08133 (E) B. spec_Borneo1 Cultivated: RBGE (acc. nom.: 20030131), Malaysia, Borneo Thomas, D. C. 07-1 (E) B. spec_Borneo2 Cultivated: BoBG, Indonesia, Borneo Thomas, D. C. & Ardi, W. H. 09-136 (E) B. spec_Sumatra Cultivated: BoBG, Indonesia, Sumatra Thomas, D. C. & Ardi, W. H. 08-132 (E) B. spec_Sumbawa1 Cultivated: BoBG, Indonesia, Sumbawa Thomas, D. C. & Ardi, W. H. 09-138 (E) B. spec sect Reichenheimea1 Cultivated: RBGE, Indonesia, Java 20111543 B. spec sect Reichenheimea3 Cultivated: RBGE, Indonesia, Java 20111545 B. spec sect Reichenheimea4 Cultivated: RBGE, Indonesia, Java 20112191 B. sublobata Jack Cultivated: RBGE, Indonesia, Java 20101649 B. sudjanae Jansson Cultivated: GBG (acc, num.: 02605499), Indonesia, Sumatra Thomas, D.C. & Ardi, W.H. 08109 (E) B. tenuifolia Dryand. Indonesia, Bali Thomas, D.C & Ardi, W.H. 0886 (E) B. venusta King Cultivated: RBGE (acc, num.: 20021596), Malaysia, Peninsula Malaysia Neale, Sophie 7 (E) B. verecunda M.Hughes Indonesia, Sumatra Wilkie, P., Hughes, M., Sumadijaya, A., Rasnovi, S., 194 Marlan & Suhardi 618 (E) B. aff. mekonggensis Gim & Indonesia, Sulawesi Thomas, D. C. & Ardi, W. H. 09-108 (E) B. aff. multangula Blume Indonesia, Sulawesi Thomas, D.C. & Ardi, W.H. 0985 (E) B. aptera Blume Indonesia, Sulawesi Smith, P. & Galloway, L. 67(E) B. argenteomarginata Tebbitt Cultivated: GBG (acc. Num.: 00803887), Papua New Guinea Forrest, L.L. 145 (E) B. bonthainensis Hemsl. Indonesia, Sulawesi Thomas, D.C. & Ardi, W.H. 0963 (E) B. brevirimosa Irmsch Cultivated: RBGE (acc. num.: 19821108, Papua New Guinea Forrest, L.L. 137 (E) B. capituliformis Irmsch Indonesia, Sulawesi Kinho, J. & Poulsen, A. 169 (E) B. chiasmogynaM.Hughes Cultivated: RBGE (acc. num.: 20021895) Indonesia, Sulawesi Thomas, D.C. 07-29 (E) B. cleopatrae_2 Coyle Philippines, Palawan Wilkie, P., Mendum, M., Argent, G.C.G., Cronk, Q., Middleton. D.J., Fuentes, R. & Chavez, R.V. 25373 (E) B. comestibilis D.C.Thomas Indonesia, Sulawesi Thomas, D. C. & Ardi, W. H. 09-62 (E) Indonesia, Sulawesi Thomas, D.C. & Ardi, W.H. 0877 (E) B. flacca Irmsch. Cultivated: BaBG, Indonesia, Sulawesi Thomas, D. C. & Ardi, W. H. 09-133 (E) B. guttapila D.C. Thomas & Ardi Cultivated: BaBG Indonesia, Sulawesi Thomas, D.C. & Ardi, W.H. 0881 (E) B. hekensis D.C.Thomas Indonesia, Sulawesi Thomas, D.C. & Ardi, W.H. 0843 (E) B. hispidissima Zipp. ex. Indonesia, Sulawesi Kinho, J. & Poulsen, A. 168 (E) Wiriad & Ardi B. didyma D.C. Thomas & Ardi Koord. 195 B. koordersii Warb.ex. Indonesia, Sulawesi Koorders, S.H. 16246B Indonesia, Sulawesi Thomas, D. C. & Ardi, W. H. 09-110 (E) B. macintyreana M.Hughes Cultivated: RBGE (acc. num.: 20021848), Indonesia, Sulawesi Thomas, D.C. 07-28 (E) B. masarangensis Irmsch Cultivated: PCHW Indonesia, Sulawesi Thomas, D.C. & Ardi, W.H. 09131 (E) B. mendumiae M.Hughes Cultivated: RBGE (acc. num.: 20021912), Indonesia, Sulawesi Thomas, D.C. 07-27 (E) B. negrosensis Elmer Philippines, Negros Wilkie, P. 76 (E) B. nigritarum Steud. Cultivated: RBGE (acc. num.: 19991994), Philippines, Luzon Island Thomas, D.C. 07-25 (E) B. nobiae spec Indonesia, Sulawesi Thomas 09-123 (E) B. ozotothrix D.C.Thomas Indonesia, Sulawesi Thomas, D.C. & Ardi, W.H. 0858 (E) B. poliloensis Indonesia, Sulawesi Forrest 139 (E) B. prionota D.C.Thomas & Indonesia, Sulawesi Thomas, D. C. & Ardi, W. H. 09-97 (E) B. pseudolateralis Warb. Indonesia, Sulawesi Thomas & Ardi 08-23 (E) B. rantemarioensis Indonesia, Sulawesi Thomas, D. C. & Ardi, W. H. 09-78 (E) Indonesia, Sulawesi Thomas, D. C. & Ardi, W. H. 09-125 (E) B. serratipetala Imrsch. Cultivated: RBGE (acc. num.: 19681637), Papua New Guinea Forrest, L.L. 135 (E) B. siccacaudata J.Door. Indonesia, Sulawesi Thomas, D.C. & Ardi, W.H. 0960 (E) B. spec_New Guinea Cultivated: BoBG, Thomas, D. C. & Ardi, W. H. L.B.Sm.et Wassh B. lasioura D.C.Thomas & Ardi Ardi D.C.Thomas & Ardi B. sanguineopilosa D.C.Thomas & Ardi 196 Indonesia, Papua 09-139 (E) B. spec_Papua1 Papua Armstrong, K. 351 (E) B. spec_Philippines Cultivated: RBGE (acc. nom.: 20080433), Philippines: Luzon Island Thomas, D. C. 08-146 (E) B. spec_Sumbawa2 Cultivated: BoBG, Indonesia, Sumbawa Thomas, D. C. & Ardi, W. H. 09-138 (E) B. steveiM.Hughes Cultivated: RBGE (acc, num.: 20040642), Indonesia, Sulawesi Thomas, D.C. 07-30 (E) B. strigosa (Warb.) Cultivated: GBG, Papua New Guinea Forrest, L.L. 143 (E) Forrest, L.L. 142 (E) & Holingsw. Cultivated: GBG (acc, num.: 00312793), Papua New Guinea B. torajanaD.C.Thomas & Indonesia, Sulawesi Thomas, D. C. & Ardi, W. H. 09-104 (E) B. varipeltata D.C. Thomas Indonesia, Sulawesi Thomas, D.C. & Ardi, W.H. 0851 (E) B. vermeulenii D.C.Thomas Indonesia, Sulawesi Vermeulen, J.J. 2301 (L) B. watuwilensis Girmansyah Cultivated: BaBG, Indonesia, Sulawesi Thomas, D.C. & Ardi, W.H. 0955 (E) B. weigallii Hemsl. Solomon Islands Pitisopa, F., Gardner, M. F., Herrington, S. 10 (E) L.L.Forrest & Hollingsw. B. symsanguineaL.L. Forrest Ardi & Ardi 197 Appendix 2: Presence/ Absence coding of each taxa of Begoniaceae in the biogeographic analysis in Biogeobears. A= America, B= Africa, C= Socotra, D= India, E= Sri Lanka, F= Himalaya, G= China, H= Indo China, I= Penninsular Malasia, J= Sumatra, K= Borneo, L= Sulawesi, M= Papua New Guinea, N= Negros, O= Java, P= Luzon, Q= Lanyu_Batan, R= Panay, S= Palawan, T= Mindanao, U= Sibuyan, V= Biliran Species Name A B C D E F G H I J K L M N O P Q R S T U V B. ningmingensis 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. acuminatissima 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 B. anisoptera 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 B. biliranensis 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 B. blancii 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 B. calasiensis 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 B. calcicola 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. caminguinensis 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 B. castilloi 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. chingipengii 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. chloroneura 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. cleopatrae_1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 B. copelandii 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 B. dinglensis 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 B. elmeri 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 B. elnidoensis 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 B. fenecis 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. gabaldonensis 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 198 B. gitingensis 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 B. gueritziana 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 B. gutierrezii 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 B. hernandioides_1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. hernandioides_2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. hughesii_1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 B. hughesii_2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 B. klemmei 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. longiscapa_1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. longiscapa_2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 B. luzonensis_1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. luzonensis_2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. luzonensis_3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. manillensis_1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. merrilliana 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 B. mindorensis 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 B. nigritarum_1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. nigritarum_2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 B. nigritarum_3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. obtusifolia 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 B. oxysperma 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. rhombicarpa 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. rubitae 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 199 B. rufipila 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 Begonia sp_1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 Begonia sp_2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. subnummarifolia 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 B. suborbiculata 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 B. sykakiengii 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 B. tagbanua 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 B. taraw 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 B. tayabensis 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. trichocheila 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. wadei 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 B. woodii 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 B. dregei 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. goudotii 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. oxyloba 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. poculifera 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. polygonoides 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. samahensis 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. socotrana 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. sutherlandii 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. boliviensis 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. nelumbifolia 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. radicans 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 200 B. tenera 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. thwaitesii 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. cordifolia 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. dipetala 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. malabarica 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. malabarica_India 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. albo-coccinea 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. floccifera 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. bryophila 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. diocia 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. flagellaris 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. hatacoa 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. panchtharensis 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. picta 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. roxburghii 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. rubella 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. sikkimensis 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. tribenensis 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. aceroides 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. acetosella 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. aff. elisabethae 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. alicida 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. brandisiana 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 201 B. demissa 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. elisabethae 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. grandis 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. hymenophylla 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. masoniana 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. morsei 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. obovoidea 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. palmata 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. silletensis 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. sizemoreae 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. smithiae 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. spec 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. spec_China1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. spec_China2 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. spec_Thailand1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. spec_Thailand2 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. spec_Thailand3 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. versicolor 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 B. aff. Congesta 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 B areolata 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 B. bracteata 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 B. chlorosticta 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 B. corrugata 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 202 B. decora 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 B. fuscisetosa 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 B. goegoensis 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 B. harauensis 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 B. kingiana 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 B. laruei 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 B. longifolia 0 0 0 0 0 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 B. multangula 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 B. multijugata 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 B. muricata 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 B. pavonina 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 B. puspitae 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 B. resecta 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 B. robusta 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 B. spec_Borneo1 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 B. spec_Kalimantan 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 B. spec_Sumatra 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 B. spec_Sumbawa1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 B. spec sect Reichenheimea1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 B. spec sect Reichenheimea3 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 B. spec sect Reichenheimea4 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 B. sublobata 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 B. sudjanae 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 203 B. tenuifolia 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 B. venusta 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 B. verecunda 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 B. aff. mekonggensis spec 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. aff. multangula 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. aptera 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. argenteomarginata 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 B. bonthainensis 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. brevirimosa 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 B. capituliformis 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. chiasmogyna 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. cleopatrae_2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 B. comestibilis spec 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. didyma 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. flacca 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. guttapila 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. hekensis 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. hispidissima 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. koordersii 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. lasiouraspec 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. macintyreana 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. masarangensis 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. mendumiae 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 204 B. negrosensis 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 B. nigritarum 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. nobiaespec 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. ozotothrix 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. poliloensis 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. prionotaspec 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. pseudolateralis 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. rantemarioensis 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. sanguineopilosaspec 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. serratipetala 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 B. siccacaudata 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. spec_NewGuinea 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 B. spec_Papua1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 B. spec_Philippines 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 B. spec_Sumbawa2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 B. stevei 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. strigosa 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 B. symsanguinea 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 B. torajana 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. varipeltata 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. vermeulenii 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. watuwilensis 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 B. weigallii 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 205 Appendix 3:List of taxa used in the biogeographic analysis of Sapotaceae and Voucher numbers. Species indicated in bold are newly generated sequences for the present study. Species name Origin EDNA/voucher Sarcosperma_laurinum Hook.f. Hong Kong, China Saunders 2000 (S) Eberhardtia_aurata Lecomte Malaysia G. Hao 534 (S) Xantolis_siamensis P.Royen Thailand Smitairi 1 (L) Sideroxylon_oxyacanthum Baillon Ethiopia, Arabia Wood Y/75/388 (BM) Palaquium_laevifolium_SL Engl. Sri Lanka Kumarage 60 (E) Isonandra_lanceolata_SL Wight Brunei, Indonesia, borneo, Sri Lanka Kumarage 47 (E) Nesoluma_polynesiacum Baillon Hawaii Degener 20770 (S) Sideroxylon_americanum T.D.Penn. Jamaica, Bahamas Gillis 11576 (B) Sideroxylon_horridumT.D.Pennington Cuba Gutiérrez & Nilsson 5 (S) Sideroxylon_obovatum H.J.Lam Venezuela, Aruba, virgin islands, antigua, Dutch west indies García et al. 5586 (S) Sideroxylon_obtusifolium Brazil,Venezuela,Ar gentina, Trinidad, Paraguay, Guatemala Alvarez et al. 28772 (B) Sideroxylon_picardae T.D.Pennington Dominican Republic, Haiti Ekman 15576 (S) Sideroxylon_occidentale baillon Mexico Carter & Sharsmith 4268 T.D.Pennington (B) Sideroxylon_angustum T.D.Penn. Cuba Ekman 4034 (S) Sideroxylon_lanuginosum Michx. Mexico, USA, Correll & Ogden 28456 (S) Sideroxylon_lycioides L. USA Radford et al. 11453 (B) Sideroxylon_persimile T.D.Pennington Mexico,Nicaragua,V enezuela, Guatemala Véliz 99.7038 (BM) 206 Sideroxylon_reclinatum Michx. USA Traverse 592 (GB) Sideroxylon_tenax L. USA Radford & Leonard 11519 (B) Sideroxylon_confertum C.Wright Cuba Ekman 17405 (S) Argania_spinosa Skeels Morocco Nordenstam 9325 ?(S) Sideroxylon_betsimisarakum Lecomte Madagascar Schönenberger et al. A102 (UPS) Sideroxylon_capiri Piettier Guatemala, Honduras, Mexico, Nicaragua,Costa Rica, Trinidad García 1848 (BM) Sideroxylon_cubense T.D.Pennington Cuba,Haiti,Dominica n Republic Beurton & Mory 927 (B) Sideroxylon_floribundum Griseb. Guatemala,Jamaica, Belize Lundell 20263 (BM) Sideroxylon_foetidissimum Jacq. Dominican Republic, Jamaica,Virgin Islands,Guadeloupe Lundin 638 (S) Sideroxylon_inerme L. Comoros, Tropical Africa, India, South Africa Nielsen s.n. (S) Sideroxylon_majus Baehni Mauritius, Port Louis Capuron 28185 (SF) Sideroxylon_marginatum Decne.ex Cape Verde Leyens CV-96-672 (B) Madeira, Canaries, Cape Verde Swenson & Fernandez Oman, Saudi Arabia,Ethiopia,Afga nistan,Yemen Thulin, Beier & Hussein Dominican Republic Greuter & Rankin 24954 Webb Sideroxylon_marmulano Banks ex Lowe Sideroxylon_mascatense T.D.Penn. Sideroxylon_repens T.D.Penn. 581 (S) 9774 (UPS) (S) Sideroxylon_salicifolium H.J.Lam Jamaica,Mexico,Beli ze,Barbuda,USA,Bah amas 207 Gutiérrez & Nilsson 14 (S) Sideroxylon_saxorum Lecomte Madagascar Jongkind 3500 (WAG) Sideroxylon_tepicense T.D.Penn. Mexico,Guatemala Ctentry 2931 (S) Sideroxylon_wightianum Hook.And Hong Kong, China Hao 532 (S) Jongkind 2351 (WAG) Mildbr. Ex Hutch.& Dalziel Ghana,Ivory coast, Congo Lecomtedoxa_klaineana Pierre ex Tropical Africa Veldhuizen 1509 (WAG) Tanzania C. K. 3483 (S) Poutera_firma (Miq.) Baehni Indonesia, West Papua Armstrong 305 (E) Pouteria_maclayana Baehni Indonesia,Malaysia,P NG Armstrong 316 (E) Pouteria_sp1 Malaysia Armstrong 317 (E) Neolemonniera_clitandrifolia Heine Ghana Jongkind, Schmidt & Arn Omphalocarpum_pachysteloides Dubard Englerophytum_natalense T.D.Pennington Abbiw 1777 (MO) Northia_seychellana Hook.F. Seychelles Chong-Seng s. n. (S) Capurodendron_androyense Madagascar Humbert 28855 (B) Madhuca_sp_SL Sri Lanka Kumarage 58 (E) Madhuca_fulva_SL J.F.Macber Sri Lanka Kumarage 62 (E) Madhuca_neriifolia_SL H.J.Lam Sri Lanka Kumarage 63 (E) Madhuca_longifolia2_SL J.F.Macbr. Sri Lanka Photo voucher available Madhuca_crassipes H.J.Lam Borneo, Indonesia, malaysia Jugah ak. Kudi 23757 (K) Madhuca_utilis H.J.Lam Malaysia Pennington & Asri 10209 Aubreville (K) Aulandra_longifolia H.J.Lam Malaysia, Sarawak Christensen 1720 (K) Palaquium_amboinense Burck Indonesia (native to New Guinea) Wilkie 813 (E) 208 Palaquium_beccarianum van Royen Malaysia,Papua, Borneo Wiijesundara s.n. (K) – native to West Papua and the Moluccas Diploknema_oligomera H.J.Lam Indonesia, Bogor Chase 1360 (K) Palaquium_brassii H.J.Lam Indonesia, West Papua Armstrong 311 (E) Palaquium_clarkeanum King & Malaysia Wilkie 501 (E) Palaquium_cryptocarifolium P.Royen Malaysia, Sarawak Wilkie 874 (E) Palaquium_dasyphyllum Pierre Malaysia Slik 9592 (L) Palaquium_eriocalyx H.J.Lam Indonesia Wilkie et al. 8/147 (E) Palaquium_formosanum Hayata Taiwan,Philippines Chung & Anderberg 1421 Gamble (HAST) Palaquium_galactoxylum H.J.Lam Australia Bartish and Jessup 9 (S) Palaquium_gutta Baillon Malaysia Wilkie 504 (E) Palaquium_herveyi King & Gamble Malaysia, Sarawak Wilkie 871 (E) Palaquium_impressionervium Ng Peninsular Malaysia Wilkie FRI52851 (E) Palaquium_leiocarpum Boeriage Malaysia, Sarawak Wilkie 870 (E) Palaquium_lobbianum Kurck Indonesia Armstrong 331 (E) Palaquium_hexandrum Baillon Malaysia, Sarawak Wilkie 872 (E) Palaquium_maingayi King & Gamble Peninsular Malaysia Wilkie 846 (E) Palaquium_microphyllum King & Malaysia Pennington, Kochummen Gamble & Wong (K) Palaquium_obovatum Engler Thailand Middleton 4387 (E) Palaquium_oxleyanum Pierre Malaysia Wilkie 527 (E) Palaquium_pseudorostratum H.J.Lam Malaysia, Sarawak Wilkie 857 (E) Palaquium_quercifolium Burck Indonesia, Kalimantan Slik CMF6780 (L) Palaquium_ridleyi King & Gamble Malaysia, Sarawak Wilkie 858 (E) 209 Palaquium_rigidum Pierre ex Dubard Malaysia Wilkie 878 (E) Palaquium_rostratum Burck Indonesia, Kalimantan Slik CMF9452 (L) Palaquium_rufolanigerum P.Royen Malaysia, Sarawak Wilkie 859 (E) Palaquium_sericeum H.J.Lam Indonesia, Kalimantan Slik CMF9737 (L) Palaquium_sp1 Solomon Islands Poulsen et al 2488 (E) Palaquium_sp2 Indonesia, Sulawesi Armstrong 370 (E) Palaquium_sumatranum Burck Indonesia, Java Wilkie 823 (E) Palaquium_walsurifolium Pierre ex Malaysia, Sarawak Wilkie 877 (E) Wilkie et al 1/477 (E) Burck Indonesia, Kalimantan Palaquium_xanthochymum Pierre ex Malaysia, Borneo Wilkie 544 (E) Palaquium_thwaitesii_SL Trim. Sri Lanka Kumarage 43 (E) Palaquium_canaliculatum_SL Engl. Sri Lanka Kumarage 45 (E) Palaquium_hinmolpedda_SL Sri Lanka Kumarage 64 (E) Palaquium_pauciflorum_SL Engl. Sri Lanka Kumarage 59 (E) Palaquium_rubiginosum_SL Engl. Sri Lanka Kumarage 65 (E) Isonandra_montana_SL Gamble Sri Lanka Kumarage 76 (E) Isonandra_sp1_SL Sri Lanka Kumarage 97 (E) Palaquium_sp3_SL Sri Lanka Kumarage 57 (E) Palaquium_grande_SL Engl. Sri Lanka Kumarage 46 (E) Palaquium_petiolare_SL Engl. Sri Lanka Kumarage 44 (E) Palaquium_sp5_SL Sri Lanka Kumarage 93 (E) Isonandra_sp2_SL Sri Lanka Kumarage 74 (E) Isonandra_compta_SL Dubard Sri Lanka Emanuelsson 3039 (S) Dubard Palaquium_calophyllum Pierre ex Burck P.Royen 210 Isonandra_lanceolata2 Wight malaysia, Brunei, Borneo, Sri Lanka Armstrong 314 (E) Isonandra_perakensis King & Gamble Malaysia Pennington & Wong 10227 (K) Isonandra_zeylanica_SL Jeuken Sri Lanka Kumarage 72 (E) Autranella_congolensis A.Chev. Congo Bokdam 4401 (WAG) Tieghemella_heckelii Pierre ex Dubard Ghana Jongkind 3936 (WAG) Vitellaria_paradoxa C.F.Gaertn. Africa Neumann 1512 (FR) Vitellariopsis_cuneata Aubreville Tropical Africa, Indonesia Thomas 3662 (S) Vitellariopsis_dispar Aubreville South Africa Balkwill & Balkwill (B) Vitellariopsis_marginata Aubreville Africa Chase 1122 (S) Faucherea_parvifolia Lecomte Madagascar Birkinshaw et al. 357(P) Inhambanella_henriquesii Dubard South Africa de Winter & Vahrmeijer 8536 (S) Baillonella_toxisperma Pierre Africa Cenarest (LBV) Labourdonnaisia_calophylloides Bojer Mauritius, Port Louis Capuron 28241-SF (P) Labourdonnaisia_revoluta Bojer Mauritius Bernardi 14717 (P) Labramia_costata Aubreville Madagascar Randrianmanalinarivo 577 (UPS) Labramia_mayottensis Labat, Madagascar Labat et al. 3309 (S) Letestua_durissima Lecomte Congo Normand s.n. (P) Manilkara_hexandra_SL Dubard Sri Photo voucher available Lanka,Hawaii,Vietn am,India,Thailand, Himalaya Manilkara_concolor Gerstner Africa Labat et al 3309 (S) Manilkara_discolor J.H.Hemsl. Africa K. Vollesen 2460 (S) Manilkara_kauki Dubard Thailand Chantaranothai 2341 M.Pignal & O.Pascal 211 (Khon Kaen Univ. Herbarium) Manilkara_obovata J.H.Hemsl. Africa Schmidt et al 3274 (S) Manilkara_zapota P.Royen Colombia, Jamaica,Costa Rica,Mexico J. Clayton 12 (S) Mimusops_caffra E.Mey ex Africa Swenson & Karis 636 (S) Mimusops_comorensis Engler Comoros Pignal & Ginguette 1065 Mimusops_elengi_SL L. Sri Lanka Kumarage 98 (E) Mimusops_obovata Sond. Africa, Madagascar Swenson & Karis 633 (S) Mimusops_zeyheri Sond. Africa Dahlstrand 6386 (S) Burckella_macropoda H.J.Lam Indonesia (native to New Guinea) Wilkie 818 (E) Burckella_polymera P.Royen Indonesia, West Papua Armstrong 326 (E) Burckella_sp India Armstrong 327 (E) Diploknema_butyracea H.J.Lam Nepal Polunin, Sykes & A.d.Candolle Williams 3975 (UPS) Diploknema_siamensis Fletcher China Middleton s.n. Madhuca_barbata T.D.Penn. Sarawak, Malaysia Wilkie 873 (E) Madhuca_curtisii Ridl. Peninsular Malaysia Wilkie 848 (E) Madhuca_elmeri Merr. Ex. H.J.Lam Indonesia, Kalimantan Wilkie P1 347 (E) Madhuca_erythrophylla H.J.Lam Malaysia, Sarawak Wilkie 867 (E) Madhuca_hainanensis Chun & China G. Hao 530 (S) Madhuca_kingiana H.J.Lam Malaysia, Sarawak Wilkie 856 (E) Madhuca_korthalsii H.J.Lam Malaysia, Sarawak Wilkie 876 (E) Madhuca_kuchingensis Yii & P.Chai Malaysia, Sarawak Wilkie 860 (E) F.C.Chow 212 Madhuca_kunstleri H.J.Lam Malaysia, Sarawak Wilkie 868 (E) Madhuca_lancifolia H.J.Lam Malaysia, Sarawak Wilkie 853 (E) Madhuca_laurifolia H.J.Lam Peninsular Malaysia Wilkie 843 (E) Madhuca_leucodermis H.J.Lam New Guinea Takeuchi et al. 17858 (S) Madhuca_malaccensis H.J.Lam Peninsular Malaysia Wilkie 832 (E) Madhuca_microphylla Alston Sri Lanka Fagerlind 4790 (S) Madhuca_motleyana Baehni Peninsular Malaysia Wilkie 837 (E) Madhuca_oblongifolia Merrill Sarawak, Malaysia Wilkie 861 (E) Madhuca_pachyphylla (K.Krause) Armstrong 313 (E) ined. Indonesia, West Papua Madhuca_palembanica Forman Indonesia Triono, Saman & Victobery 11 (K) Madhuca_palida Baehni Sarawak, Malaysia Wilkie 869 (E) Madhuca_prolixa Yii & P.Chai Sarawak, Malaysia Wilkie 875 (E) Madhuca_sarawakensis H.J.Lam Sarawak, Malaysia Wilkie 863 (E) Madhuca_sericea H.J.Lam Sarawak, Malaysia Wilkie 879 (E) Madhuca_sp Peninsular Malaysia Wilkie 834 (E) Payena_acuminata Pierre Indonesia, Java Chase 1368 (K) Madhuca_clavata_SL Jayas. Sri Lanka Kumarage 99 (E) Palaquium_sp4_SL Sri Lanka Kumarage 75 (E) Payena_lucida_1 A.de.Candolle Peninsular Malaysia Wilkie 845 (E) Payena_maingayi C.B.Clarke Peninsular Malaysia Wilkie 841 (E) Payena_leerii Kurz Malaysia Wilkie 811 (E) Payena_ferruginea J.T.Pereira Peninsular Malaysia Wilkie 866 (E) Payena_lucida_2 A.d.Candolle Borneo Ambri et al. AA1604 (L) Payena_obscura Burck Peninsular Malaysia Wilkie 880 (E) 213 Appendix 4:Presence/ Absence coding of each taxa of Sapotaceae in the biogeographic analysis in Biogeobears: A= America, B= Africa, C= Madagascar, D= Seychelles, E= Middle East, F= India, G= Sri Lanka, H= Himalaya, I= East Asia, J= Sunda Shelf, K= Sahul Shelf Species name A B C D E F G H I J K Sarcosperma_laurinum 0 0 0 0 0 0 0 0 1 0 0 Eberhardtia_aurata 0 0 0 0 0 0 0 0 0 1 0 Xantolis_siamensis 0 0 0 0 0 0 0 0 1 0 0 Sideroxylon_oxyacanthum 0 1 0 0 1 0 0 0 0 0 0 Palaquium_laevifolium_SL 0 0 0 0 0 0 1 0 0 0 0 Isonandra_lanceolata_SL 0 0 0 0 0 0 1 0 0 1 0 Nesoluma_polynesiacum 0 0 0 0 0 0 0 0 0 0 1 Sideroxylon_americanum 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_horridum 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_obovatum 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_obtusifolium 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_picardae 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_occidentale 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_angustum 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_lanuginosum 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_lycioides 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_persimile 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_reclinatum 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_tenax 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_confertum 1 0 0 0 0 0 0 0 0 0 0 Argania_spinosa 0 1 0 0 0 0 0 0 0 0 0 Sideroxylon_betsimisarakum 0 0 1 0 0 0 0 0 0 0 0 Sideroxylon_capiri 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_cubense 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_floribundum 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_foetidissimum 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_inerme 0 1 0 0 0 1 0 0 0 0 0 Sideroxylon_majus 0 1 0 0 0 0 0 0 0 0 0 214 Sideroxylon_marginatum 0 1 0 0 0 0 0 0 0 0 0 Sideroxylon_marmulano 0 1 0 0 0 0 0 0 0 0 0 Sideroxylon_mascatense 0 1 0 0 1 0 0 0 0 0 0 Sideroxylon_repens 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_salicifolium 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_saxorum 0 0 1 0 0 0 0 0 0 0 0 Sideroxylon_tepicense 1 0 0 0 0 0 0 0 0 0 0 Sideroxylon_wightianum 0 0 0 0 0 0 0 0 1 0 0 Omphalocarpum_pachysteloides 0 1 0 0 0 0 0 0 0 0 0 Lecomtedoxa_klaineana 0 1 0 0 0 0 0 0 0 0 0 Englerophytum_natalense 0 1 0 0 0 0 0 0 0 0 0 Poutera_firma 0 0 0 0 0 0 0 0 0 1 1 Pouteria_maclayana 0 0 0 0 0 0 0 0 0 1 1 Pouteria_sp1 0 0 0 0 0 0 0 0 0 1 0 Neolemonniera_clitandrifolia 0 1 0 0 0 0 0 0 0 0 0 Northia_seychellana 0 0 0 1 0 0 0 0 0 0 0 Capurodendron_androyense 0 0 1 0 0 0 0 0 0 0 0 Madhuca_sp_SL 0 0 0 0 0 0 1 0 0 0 0 Madhuca_fulva_SL 0 0 0 0 0 0 1 0 0 0 0 Madhuca_neriifolia_SL 0 0 0 0 0 0 1 0 0 0 0 Madhuca_longifolia2_SL 0 0 0 0 0 0 1 0 0 0 0 Madhuca_crassipes 0 0 0 0 0 0 0 0 0 1 0 Madhuca_utilis 0 0 0 0 0 0 0 0 0 1 0 Aulandra_longifolia 0 0 0 0 0 0 0 0 0 1 0 Palaquium_amboinense 0 0 0 0 0 0 0 0 0 1 1 Palaquium_beccarianum 0 0 0 0 0 0 0 0 0 1 1 Diploknema_oligomera 0 0 0 0 0 0 0 0 0 1 0 Palaquium_brassii 0 0 0 0 0 0 0 0 0 0 1 Palaquium_clarkeanum 0 0 0 0 0 0 0 0 0 1 0 Palaquium_cryptocarifolium 0 0 0 0 0 0 0 0 0 1 0 Palaquium_dasyphyllum 0 0 0 0 0 0 0 0 0 1 0 Palaquium_eriocalyx 0 0 0 0 0 0 0 0 0 1 0 Palaquium_formosanum 0 0 0 0 0 0 0 0 0 1 0 215 Palaquium_galactoxylum 0 0 0 0 0 0 0 0 0 0 1 Palaquium_gutta 0 0 0 0 0 0 0 0 0 1 0 Palaquium_herveyi 0 0 0 0 0 0 0 0 0 1 0 Palaquium_impressionervium 0 0 0 0 0 0 0 0 0 1 0 Palaquium_leiocarpum 0 0 0 0 0 0 0 0 0 1 0 Palaquium_lobbianum 0 0 0 0 0 0 0 0 0 1 0 Palaquium_hexandrum 0 0 0 0 0 0 0 0 0 1 0 Palaquium_maingayi 0 0 0 0 0 0 0 0 0 1 0 Palaquium_microphyllum 0 0 0 0 0 0 0 0 0 1 0 Palaquium_obovatum 0 0 0 0 0 0 0 0 1 0 0 Palaquium_oxleyanum 0 0 0 0 0 0 0 0 0 1 0 Palaquium_pseudorostratum 0 0 0 0 0 0 0 0 0 1 0 Palaquium_quercifolium 0 0 0 0 0 0 0 0 0 1 0 Palaquium_ridleyi 0 0 0 0 0 0 0 0 0 1 0 Palaquium_rigidum 0 0 0 0 0 0 0 0 0 1 0 Palaquium_rostratum 0 0 0 0 0 0 0 0 0 1 0 Palaquium_rufolanigerum 0 0 0 0 0 0 0 0 0 1 0 Palaquium_sericeum 0 0 0 0 0 0 0 0 0 1 0 Palaquium_sp1 0 0 0 0 0 0 0 0 0 0 1 Palaquium_sp2 0 0 0 0 0 0 0 0 0 1 0 Palaquium_sumatranum 0 0 0 0 0 0 0 0 0 1 0 Palaquium_walsurifolium 0 0 0 0 0 0 0 0 0 1 0 Palaquium_calophyllum 0 0 0 0 0 0 0 0 0 1 0 Palaquium_xanthochymum 0 0 0 0 0 0 0 0 0 1 0 Palaquium_thwaitesii_SL 0 0 0 0 0 0 1 0 0 0 0 Palaquium_canaliculatum_SL 0 0 0 0 0 0 1 0 0 0 0 Palaquium_hinmolpedda_SL 0 0 0 0 0 0 1 0 0 0 0 Palaquium_pauciflorum_SL 0 0 0 0 0 0 1 0 0 0 0 Palaquium_rubiginosum_SL 0 0 0 0 0 0 1 0 0 0 0 Isonandra_montana_SL 0 0 0 0 0 0 1 0 0 0 0 Isonandra_sp1_SL 0 0 0 0 0 0 1 0 0 0 0 Palaquium_sp3_SL 0 0 0 0 0 0 1 0 0 0 0 Palaquium_grande_SL 0 0 0 0 0 0 1 0 0 0 0 216 Palaquium_petiolare_SL 0 0 0 0 0 0 1 0 0 0 0 Palaquium_sp5_SL 0 0 0 0 0 0 1 0 0 0 0 Isonandra_sp2_SL 0 0 0 0 0 0 1 0 0 0 0 Isonandra_compta_SL 0 0 0 0 0 0 1 0 0 0 0 Isonandra_lanceolata2 0 0 0 0 0 0 0 0 0 0 1 Isonandra_perakensis 0 0 0 0 0 0 0 0 0 1 0 Isonandra_zeylanica_SL 0 0 0 0 0 0 1 0 0 0 0 Autranella_congolensis 0 1 0 0 0 0 0 0 0 0 0 Tieghemella_heckelii 0 1 0 0 0 0 0 0 0 0 0 Vitellaria_paradoxa 0 1 0 0 0 0 0 0 0 0 0 Vitellariopsis_cuneata 0 1 0 0 0 0 0 0 0 1 0 Vitellariopsis_dispar 0 1 0 0 0 0 0 0 0 0 0 Vitellariopsis_marginata 0 1 0 0 0 0 0 0 0 0 0 Faucherea_parvifolia 0 0 1 0 0 0 0 0 0 0 0 Inhambanella_henriquesii 0 1 0 0 0 0 0 0 0 0 0 Baillonella_toxisperma 0 1 0 0 0 0 0 0 0 0 0 Labourdonnaisia_calophylloides 0 1 0 0 0 0 0 0 0 0 0 Labourdonnaisia_revoluta 0 1 0 0 0 0 0 0 0 0 0 Labramia_costata 0 0 1 0 0 0 0 0 0 0 0 Labramia_mayottensis 0 0 1 0 0 0 0 0 0 0 0 Letestua_durissima 0 1 0 0 0 0 0 0 0 0 0 Manilkara_hexandra_SL 0 0 0 0 0 1 1 0 0 0 0 Manilkara_concolor 0 1 0 0 0 0 0 0 0 0 0 Manilkara_discolor 0 1 0 0 0 0 0 0 0 0 0 Manilkara_kauki 0 0 0 0 0 0 0 0 1 0 0 Manilkara_obovata 0 1 0 0 0 0 0 0 0 0 0 Manilkara_zapota 1 0 0 0 0 0 0 0 0 0 0 Mimusops_caffra 0 1 0 0 0 0 0 0 0 0 0 Mimusops_comorensis 0 1 0 0 0 0 0 0 0 0 0 Mimusops_elengi_SL 0 0 0 0 0 1 1 0 0 0 0 Mimusops_obovata 0 1 1 0 0 0 0 0 0 0 0 Mimusops_zeyheri 0 1 0 0 0 0 0 0 0 0 0 Burckella_macropoda 0 0 0 0 0 0 0 0 0 1 1 217 Burckella_polymera 0 0 0 0 0 0 0 0 0 1 1 Burckella_sp 0 0 0 0 0 1 0 0 0 0 0 Diploknema_butyracea 0 0 0 0 0 0 0 1 0 0 0 Diploknema_siamensis 0 0 0 0 0 0 0 0 1 0 0 Madhuca_barbata 0 0 0 0 0 0 0 0 0 1 0 Madhuca_curtisii 0 0 0 0 0 0 0 0 0 1 0 Madhuca_elmeri 0 0 0 0 0 0 0 0 0 1 0 Madhuca_erythrophylla 0 0 0 0 0 0 0 0 0 1 0 Madhuca_hainanensis 0 0 0 0 0 0 0 0 1 0 0 Madhuca_kingiana 0 0 0 0 0 0 0 0 0 1 0 Madhuca_korthalsii 0 0 0 0 0 0 0 0 0 1 0 Madhuca_kuchingensis 0 0 0 0 0 0 0 0 0 1 0 Madhuca_kunstleri 0 0 0 0 0 0 0 0 0 1 0 Madhuca_lancifolia 0 0 0 0 0 0 0 0 0 1 0 Madhuca_laurifolia 0 0 0 0 0 0 0 0 0 1 0 Madhuca_leucodermis 0 0 0 0 0 0 0 0 0 1 0 Madhuca_malaccensis 0 0 0 0 0 0 0 0 0 1 0 Madhuca_microphylla 0 0 0 0 0 0 1 0 0 0 0 Madhuca_motleyana 0 0 0 0 0 0 0 0 0 1 0 Madhuca_oblongifolia 0 0 0 0 0 0 0 0 0 1 0 Madhuca_pachyphylla 0 0 0 0 0 0 0 0 0 1 1 Madhuca_palembanica 0 0 0 0 0 0 0 0 0 1 0 Madhuca_palida 0 0 0 0 0 0 0 0 0 1 0 Madhuca_prolixa 0 0 0 0 0 0 0 0 0 1 0 Madhuca_sarawakensis 0 0 0 0 0 0 0 0 0 1 0 Madhuca_sericea 0 0 0 0 0 0 0 0 0 1 0 Madhuca_sp 0 0 0 0 0 0 0 0 0 1 0 Payena_acuminata 0 0 0 0 0 0 0 0 0 1 0 Madhuca_clavata_SL 0 0 0 0 0 0 1 0 0 0 0 Palaquium_sp4_SL 0 0 0 0 0 0 1 0 0 0 0 Payena_lucida_1 0 0 0 0 0 0 0 0 0 1 0 Payena_maingayi 0 0 0 0 0 0 0 0 0 1 0 Payena_leerii 0 0 0 0 0 0 0 0 0 1 0 218 Payena_ferruginea 0 0 0 0 0 0 0 0 0 1 0 Payena_lucida_2 0 0 0 0 0 0 0 0 0 1 0 Payena_obscura 0 0 0 0 0 0 0 0 0 1 0 219 Appendix 5: List of taxa used in the biogeographic analysis of Zingiberaceae, Voucher number and GenBank accession number of gene sequences. Species indicated in bold are newly generated sequences for the present study. Species name Voucher ITS matK Siphonochilus_aethiopicus2 GH 00-134 (US) AF478792 AF478892 GH 00-135 (US) AF478793 AF478894 Kress 92-3404 AF478794 AF478895 Kazuyuki S55 (KYO) AF478797 AF478898 Distichochlamys_sp Kress 01-6848 (E) AF478745 AF478844 Alpinia_ligulata1 K.Schum. L-98.0560 (LYON) AY472361 AY742415 Alpinia_nieuwenhuizii Valeton L-92.0474 (KES(LYON)) AY742366 AY742421 Alpinia_guangdongensis S.J.Chen & Kress 99-6372 (US) AY742352 AY742408 Alpinia_maclurei Merr. Kress 95-5540 (US)A AY742362 AY742416 Alpinia_rubricaulis K.Schum. Sul 02-251 (E) AY742378 AY742432 Alpinia_zerumbet B.L.Burtt & R.M. Liao 020704 (SCIB) AY742389 AY742443 Alpinia_oxyphylla Miq. Liao 020707 (SCIB) AY742372 AY742425 Alpinia_brevilabris (Valeton) Kress 94-5335 (US) AY742341 AY742399 Kress 94-3744 AY745692 AY742427 (Schweinf.) B.L.Burtt Siphonochilus_decorus1 (Drutten) Lock Siphonochilus_kirkii (Hook. F.) B.L. Burtt Tamijia_flagellaris S.Sakai & Nagam. Z.Y.Chen Sm. R.M.Sm. Alpinia_polyantha D. Fang (NMNH(SI)) Alpinia_nutans Roscoe L-91.0066 (CS (LYON)) AY742369 AY742423 Alpinia_foxworthyi Ridl. Kress 98-6293 (US) AF478714 AF478814 Alpinia_rosea Elmer Cronk 25436 (E) AY742377 AY742431 Gagnepainia_harmandii K. Schum. Williams 02-601 (DUKE) AY339740 AY341103 Curcuma_roscoeana Wall. Kress 98-6253 (US) AF478739 AF478839 Hedychium_greenei W.W. Sm. Kress 94-3721 (US) AF478759 AF478858 Hedychium_villosum Wall. Kress 00-6603 (US) AF478762 AF478861 220 Zingiber_sulphureum I. Theilade Kress 00-6719 (US) AF478801 AF478904 Zingiber_wrayii Ridl Kress 00-6721 (US) AF478802 AF478905 Gagnepainia_thoreliana K. Schum Williams 00-199 (DUKE) AF478752 AF478851 Hemiorchis_burmanica Kurz Kress 94-3721 (US) AF478764 AF478862 Hemiorchis_rhodorrhachis K.Schum Newman 861 (E) AF478763 AF478863 Alpinia_sibuyanensis Elmer L-99.0098 (LYON) AY742381 AY742434 Alpinia_vulcanica Elmer Wilkie 29129 (E) AY742387 AY742441 Alpinia_flabellata Ridl. None given AY742349 AY742406 Alpinia_japonica (Thunb.)Miq. Kress 99-6360 (HLA AY742357 AY742412 AY742365 AY742420 Alpinia_coriacea T.L.Wu & S.J.Chen Kress 95-5539 (US) AY742344 AY742402 Alpinia_rugosaJ.P.Liao,ined. Liao 020702 (SCIB) AY742379 AY742433 Alpinia_shimadae Hayata Liu 1140-1 (TNU) FJ496775 FJ496766 Alpinia_pumila Hook.f. Kress 97-6119 (US) AF478719 AF478819 Alpinia_formosana K.Schum Kress 94-5336 (US) AY742350 AY742407 Alpinia_intermedia Gagnep. Kress 97-5780 (US) AF478716 AF478816 Alpinia_coeruleoviridis K.Schum. Johansson et al 240 (E) AY742343 AY742401 Camptandra_parvula1 (Baker) Kress 99-6328 (US) AF478730 AF478830 Alpinia_monopleura K.Schum. SUL 143 (E) AY742363 AY742419 Alpinia_cylindrocephala K.Schum. SUL 02-60 (E) AY742345 AY724203 Alpinia_aenea B.L.Burtt & R.M. Argent 0016 (E) AY742351 AY742394 Alpinia_carolinensis Koidz. Kress 94-3657 (US) AF478711 AF478811 Alpinia_eremochlamys K.Schum SUL 02-68 (E) AY742346 AY742404 Renealmia_cernua (Roem. & Schult) Kress 99-6414 (US) AF478780 AF478881 Renealmia_alpinia1 (Rottb.) Maas. Kress 99-6407 (US) AF478778 AF478879 Renealmia_thyrsoidea (R.&P.) Kress 99-6406 (LYON) AF478783 AF478884 Burbidgea_nitida Hook f. Kress 97-5781 (US) AF478728 AF478828 Burbidgea_schizocheila1 Hackett Kress 01-6867 (US) AF478729 AF478829 (LYON) Alpinia_mutica Roxb. L-94.0307 (CS (LYON)) Ridl. Sm.Koidz. Macbride Poepp. & Endl. 221 Riedelia_sp GH 96-281 (US) AF478785 AF478886 Pleuranthodium_trichocalyx1 L-94.0315 (KMN) AY742332 AY742392 L-78-0460 (KMN) AY742333 AY742391 Kress 00-6725 (US) AF478775 AF478876 (Valeeton) R.M.Sm. Pleuranthodium_floccosum (Valeton) R.M.Sm. Pleuranthodium_schlechteri1 (K.Schum.)R.M.Sm. Amomum_mentawaiense Droop ADP 2249 (E) Amomum_cerasinum Ridl. AJD 160 (E) Amomum_aculeatum Roxb. AJD 96 (E) Amomum_pierreanum2 Gagnep. JKA 222 (E) Amomum_coriandriodorum Xia 721 (HITBC) AY351987 AY352016 Kress 00-6791 AY351997 AY352027 Kress 98-6197 (US) AY351998 AY352028 AY742370 AY742424 AY352003 AY352033 S.Q.Tong et Y.M.Xia Amomum_paratsaoko S.Q.Tong et Y.M.Xia Amomum_aff._paratsaoko S.Q.Tong et Y.M.Xia Amomum_xanthophlebiumBaker AJD 81 (E) Alpinia_oceanica Burkill L-94.0309 (LYON) Amomum_staminidivum ADP 2113 A.I.Gobilik,A.L. Lamb & A.D.Poulsen Amomum_quadratolaminare Xia 729 (HITBC) S.Q.Tong Amomum_krervanh2 Pierre ex AJD 10 (E) Gagnep. Amomum_krervanh3 Pierre ex AJD 16 (E) Gagnep. Amomum_villosumLour. Kress 01-6978 (US) AF478724 AF478824 Amomum_sp10 Kress 99-6373 (US) AF478723 AF478823 Siliquamomum_tonkinenseBaill. Kress 00-6802 (US) AF478791 AF478892 Alpinia_rafflesiana Baker Rangsiruji 52 (E) AY742376 AY742430 Amomum_uliginosum2 J.Konig. 99-0474 (LYON) AY352008 AY352038 Amomum_compactum3 Sol. Xia 720 (HITBC) AY351986 AY352016 222 Ex.Maton Alpinia_purpurata (Viell.) K.Schum. Rangsiruji 9 (LYON) AY742375 AY742429 Alpinia_arctiflora (F.Muell.) Benth. Rangsiruji 48 (E) AY742336 AY742395 Alpinia_caerulea (R.Br.)Benth. L-870285 (NSW) AY742342 AY742400 Alpinia_vittata W.Bull. Kress 99-6415 (US) AF478720 AF478820 Geocharis_sp2 AJD 19 (E) Amomum_laxesquamosum K.Schum. LYON 2000.0388b(HLA AY351994 AY352024 LYON) Amomum_oligophyllum Droop AJD 155 (E) Amomum_hastilabium2 Ridl. ADP 2262 (E) Amomum_hastilabium1 Ridl. AJD 76 (E) Amomum_sp9 AJD 92 (E) Amomum_pseudofoetensValeton ADP 2284 (E) Alpinia_modesta (F.mMuell.) K. 234226 (L-86.0459) (NSW) AY742364 AY742418 L-95.00014 (NSW) AY742338 AY742397 Schum. Alpinia_arundelliana (F.M.Bailey)K.Schum. Geocharis_sp1 AJD 106 (E) Vanoverberghia_sepulchrei (Merr.) Kress 95-5562 (US) AF478798 478899 Etlingera_elatior1 (Jack) R.M.Sm. Kress 94-4251 (US) AF478749 AF478848 Etlingera_triorgyalis Ped. & Joh. 1065e (C) AF414475 AF434864 Amomum_propinquum Ridl. 93-0558 (LYON) AY351999 AY352029 Amomum_koenigii J.F.Gmel. Xia 723 (HITBC) AY351991 AY352021 Amomum_yunnanense S.Q.Tong. Xia 737 (HITBC) AY352012 AY352042 Amomum_sericeum Roxb. Xia 730 (HITBC) AY352005 AY352035 Elettariopsis_smithiae Kress 99-6313 (US) AY352013 AY352043 Amomum_menglaense S.Q.Tong Xia 726 (HITBC) AY351996 AY352026 Elettariopsis_sp Kress 00-6720 (US) AF478747 AF478846 Amomum_putrescens D.Fang. Xia 728 (HITBC) AY352002 AY352032 Amomum_queenslandicum R.M.Sm. Kmn 1428 (HTBC) AY352004 AY352034 Amomum_austrosinenseD.Fang. Xia 719 (HITBC) AY351985 AY352015 Amomum_purpureorubrum S.Q.Tong Xia 727 (HITBC) AY352000 AY352030 (Baker)R.M.Sm. 223 et Y.M.Xia Amomum_maximum Roxb. Xia 725 (HITBC) AY351995 AY352025 Amomum_glabrum2 S.Q.Tong Kress 00-6715 (US) AF478721 AF478821 Aframomum_angustifolium (Sonn.) Kress 92-3403 (US) AF478704 AF478804 Alpinia_galanga2 (L.)Wild. Kress 02-7213d (US) AF478715 AF478815 Alpinia_conchigera Griff. Kress 00-6706 (US) AF478712 AF478812 Elettariopsis_kerbyi R.M. Sm. Kress 96-5746 (US) AF478746 AF478845 Elettariopsis_stenosiphon B.l.Burtt & Kress 01-6847 (US) AF478748 AF478847 AY769826 AY769789 K.Schum. R.M. Sm. Amomum_petaloideum (S.Q.Tong) Kress95-5508 (US) T.L.Wu Zingiber_cylindricum Moon Kumarage 7 (E) Amomum_pterocarpum Thwaites Kumarage 21 (E) Zingiber_wightianum Thwaites Kumarage 20 (E) Amomum_sp1 Kumarage 34 (E) Elettaria_sp Kumarage 5 (E) Curcuma_zedoaria Roxb. Kumarage 24 (E) Amomum_fulviceps Thwaites Kumarage 33 (E) Amomum_sp2 Kumarage 50 (E) Curcuma_albiflora Thwaites Kumarage 38 (E) Amomum_graminifolium Thwaites Kumarage 48 (E) Amomum_sp3 Kumarage 82 (E) Amomum_nemorale Benth. & Kumarage 52 (E) Hook.f. Cyphostigma_pulchellum Benth. Kumarage 54 (E) Amomum_acuminatum Thwaites Kumarage 55 (E) Amomum_masticatorium Thwaites Kumarage67 (E) Alpinia_abundiflora1 Burtt & Kumarage 70 (E) R.M.Sm. Amomum_sp5 Kumarage 86 (E) Amomum_sp6 Kumarage 88 (E) Amomum_sp7 Kumarage 89 (E) 224 Alpinia_sp Kumarage 71 (E) Alpinia_fax Burtt & R.M.Sm. Kumarage 94 (E) Alpinia_hookeriana Valeton L-91.0064 (CS (LYON)) AY742356 AY742411 Alpinia_latilabris Ridl. L-80.0710 (LYON) AY742360 AY742414 Alpinia_argentea R.M.Sm. L-95-0364 (US) AY742337 AY742396 Alpinia_warburgii K.Schum. Sul 02-169 (E) AY742388 AY742442 Alpinia_tonkinensis Gagnep. Liao 020709 (US) AF478720 AF478820 Alpinia_pricei Hayata Kress 01-6860 (LYON) AY742374 AY742428 Alpinia_calcarata Roscoe Kress 94-3657 AF478710 AF278810 Alpinia_bilamellata Makino L-97-0268 (LYON) AY742339 AY742398 Alpinia_nigra (Gaertn.) B.L.Burtt Kress 01-6880 AY742444 AY742445 Leptosolena_haenkei C.Presl. Funakoshi & Co 2006 AY742331 AY742390 AY742373 AY742426 AF478766 AF478865 AF478717 AF478817 (US) Amomum_lappaceum1 Ridl. ADP2353 Alpinia_pinetorum(Ridl.) Loes. L-87.0607 (LYON) Amomum_centrocephalum1 AJD 29 (E) A.D.Poulsen Hornstedtia_hainanensis T.L.Wu & Kress 97-5769 (US) S.J.Chen Elettariopsis_sp5 AJD 65 (E) Alpinia_luteocarpa Elmer Kress 99-6403 (US) Amomum_compactum2 Sol. Ex AJD 59 Maton Alpinia_elegans K.Schum. Kress 99-6412 (US) AF478713 AF478813 Amomum_uliginosum1 J.Konig. 99-0474 (LYON) AY352008 AY352038 Aframomum_daniellii Kress 99-6375 (US) AF478705 AF478805 Aframomum_sceptrum K.Schum. Carle 94-9097 AF478706 AF478806 Renealmia_battenbergiana Cummins Kress 94-5277(US) AF478779 AF478880 (Hook.f.)K.Schum. ex Baker 225 Appendix 6: Presence/ Absence coding of each taxa of Zingiberaceae ceae in the biogeographic analysis in Biogeobears: A= America, B= Africa, C= Madagascar, D= India, E= Sri Lanka, F= China, G= Indo China, H= Sunda Shelf, I= Sahul Shelf, J= Papua New Guinea, K= Philippines, L= Australia Species name A B C D E F G H I J K L Siphonochilus_aethiopicus2 0 1 0 0 0 0 0 0 0 0 0 0 Siphonochilus_decorus1 0 1 0 0 0 0 0 0 0 0 0 0 Siphonochilus_kirkii 0 0 1 0 0 0 0 0 0 0 0 0 Tamijia_flagellaris 0 0 0 0 0 0 0 1 0 0 0 0 Distichochlamys_sp 0 0 0 0 0 0 1 0 0 0 0 0 Alpinia_ligulata1 0 0 0 0 0 0 0 1 0 0 0 0 Alpinia_nieuwenhuizii 0 0 0 0 0 0 0 1 0 0 0 0 Alpinia_hookeriana 0 0 0 0 0 0 0 1 0 0 0 0 Alpinia_latilabris 0 0 0 0 0 0 0 1 0 0 0 0 Alpinia_guangdongensis 0 0 0 0 0 1 0 0 0 0 0 0 Alpinia_maclurei 0 0 0 0 0 1 0 0 0 0 0 0 Alpinia_rubricaulis 0 0 0 0 0 0 0 0 1 0 0 0 Alpinia_argentea 0 0 0 0 0 0 0 1 0 0 0 0 Alpinia_warburgii 0 0 0 0 0 0 0 0 1 0 0 0 Alpinia_tonkinensis 0 0 0 0 0 0 1 0 0 0 0 0 Alpinia_zerumbet 0 0 0 0 0 1 0 0 0 0 0 0 Alpinia_oxyphylla 0 0 0 0 0 1 0 0 0 0 0 0 Alpinia_brevilabris 0 0 0 0 0 0 0 0 0 0 1 0 Alpinia_polyantha 0 0 0 0 0 1 0 0 0 0 0 0 Alpinia_nutans 0 0 0 0 0 0 0 1 0 0 0 0 Alpinia_foxworthyi 0 0 0 0 0 0 0 0 0 0 1 0 Alpinia_rosea 0 0 0 0 0 0 0 0 0 0 1 0 Gagnepainia_harmandii 0 0 0 0 0 1 1 0 0 0 0 0 Curcuma_roscoeana 0 0 0 0 0 0 1 0 0 0 0 0 Hedychium_greenei 0 0 0 1 0 0 0 0 0 0 0 0 Hedychium_villosum 0 0 0 0 0 1 0 0 0 0 0 0 Zingiber_sulphureum 0 0 0 0 0 0 0 1 0 0 0 0 Zingiber_wrayii 0 0 0 0 0 0 1 0 0 0 0 0 226 Gagnepainia_thoreliana 0 0 0 0 0 0 1 0 0 0 0 0 Hemiorchis_burmanica 0 0 0 0 0 0 1 0 0 0 0 0 Hemiorchis_rhodorrhachis 0 0 0 0 0 0 1 0 0 0 0 0 Alpinia_sibuyanensis 0 0 0 0 0 0 0 0 0 0 1 0 Alpinia_vulcanica 0 0 0 0 0 0 0 0 1 0 0 0 Alpinia_flabellata 0 0 0 0 0 0 0 0 0 0 1 0 Alpinia_japonica 0 0 0 0 0 1 1 0 0 0 0 0 Alpinia_pricei 0 0 0 0 0 1 0 0 0 0 0 0 Alpinia_mutica 0 0 0 0 0 0 0 1 0 0 0 0 Alpinia_coriacea 0 0 0 0 0 1 0 0 0 0 0 0 Alpinia_rugosa 0 0 0 0 0 1 0 0 0 0 0 0 Alpinia_calcarata 0 0 0 0 0 1 0 0 0 0 0 0 Alpinia_shimadae 0 0 0 0 0 1 0 0 0 0 0 0 Alpinia_pumila 0 0 0 0 0 1 0 0 0 0 0 0 Alpinia_formosana 0 0 0 0 0 1 0 0 0 0 0 0 Alpinia_intermedia 0 0 0 0 0 1 0 0 0 0 0 0 Alpinia_bilamellata 0 0 0 0 0 1 0 0 0 0 0 0 Alpinia_nigra 0 0 0 0 0 0 1 0 0 0 0 0 Alpinia_coeruleoviridis 0 0 0 0 0 0 0 0 1 0 0 0 Camptandra_parvula1 0 0 0 0 0 0 1 0 0 0 0 0 Alpinia_monopleura 0 0 0 0 0 0 0 0 1 0 0 0 Alpinia_cylindrocephala 0 0 0 0 0 0 0 0 1 0 0 0 Alpinia_aenea 0 0 0 0 0 0 0 0 1 0 0 0 Alpinia_carolinensis 0 0 0 0 0 0 0 0 1 0 0 0 Alpinia_eremochlamys 0 0 0 0 0 0 0 0 1 0 0 0 Renealmia_cernua 1 0 0 0 0 0 0 0 0 0 0 0 Renealmia_alpinia1 1 0 0 0 0 0 0 0 0 0 0 0 Renealmia_thyrsoidea 1 0 0 0 0 0 0 0 0 0 0 0 Burbidgea_nitida 0 0 0 0 0 0 0 1 0 0 0 0 Burbidgea_schizocheila1 0 0 0 0 0 0 0 1 0 0 0 0 Riedelia_sp 0 0 0 0 0 0 0 0 0 1 0 0 Pleuranthodium_trichocalyx1 0 0 0 0 0 0 0 0 0 1 0 0 Pleuranthodium_floccosum 0 0 0 0 0 0 0 0 0 1 0 0 227 Pleuranthodium_schlechteri1 0 0 0 0 0 0 0 0 0 1 0 0 Leptosolena_haenkei 0 0 0 0 0 0 0 0 0 0 1 0 Amomum_mentawaiense 0 0 0 0 0 0 0 1 0 0 0 0 Amomum_cerasinum 0 0 0 0 0 0 0 1 0 0 0 0 Amomum_aculeatum 0 0 0 0 0 0 0 1 0 0 0 0 Amomum_pierreanum2 0 0 0 0 0 0 1 0 0 0 0 0 Amomum_coriandriodorum 0 0 0 0 0 1 0 0 0 0 0 0 Amomum_paratsaoko 0 0 0 0 0 1 0 0 0 0 0 0 Amomum_aff._paratsaoko 0 0 0 0 0 1 0 0 0 0 0 0 Amomum_xanthophlebium 0 0 0 0 0 0 0 1 0 0 0 0 Amomum_lappaceum1 0 0 0 0 0 0 0 1 0 0 0 0 Alpinia_pinetorum 0 0 0 0 0 0 0 0 0 0 1 0 Alpinia_oceanica 0 0 0 0 0 0 0 0 0 1 0 0 Amomum_centrocephalum1 0 0 0 0 0 0 0 1 0 0 0 0 Hornstedtia_hainanensis 0 0 0 0 0 1 0 0 0 0 0 0 Amomum_staminidivum 0 0 0 0 0 0 0 1 0 0 0 0 Amomum_quadratolaminare 0 0 0 0 0 1 0 0 0 0 0 0 Elettariopsis_sp5 0 0 0 0 0 0 0 1 0 0 0 0 Amomum_krervanh2 0 0 0 0 0 0 0 1 0 0 0 0 Amomum_krervanh3 0 0 0 0 0 0 0 0 1 0 0 0 Amomum_villosum 0 0 0 0 0 1 0 0 0 0 0 0 Amomum_sp10 0 0 0 0 0 0 0 0 1 0 0 0 Siliquamomum_tonkinense 0 0 0 0 0 1 0 0 0 0 0 0 Alpinia_rafflesiana 0 0 0 0 0 0 0 1 0 0 0 0 Amomum_uliginosum2 0 0 0 0 0 0 1 0 0 0 0 0 Amomum_compactum3 0 0 0 0 0 0 0 1 0 0 0 0 Alpinia_purpurata 0 0 0 0 0 0 0 1 0 0 0 0 Alpinia_arctiflora 0 0 0 0 0 0 0 0 0 0 0 1 Alpinia_caerulea 0 0 0 0 0 0 0 0 0 0 0 1 Alpinia_vittata 0 0 0 0 0 0 0 0 0 1 0 0 Alpinia_luteocarpa 0 0 0 0 0 0 0 0 0 0 1 0 Amomum_laxesquamosum 0 0 0 0 0 0 0 1 0 0 0 0 Amomum_oligophyllum 0 0 0 0 0 0 0 1 0 0 0 0 228 Amomum_hastilabium2 0 0 0 0 0 0 0 1 0 0 0 0 Amomum_hastilabium1 0 0 0 0 0 0 0 0 1 0 0 0 Amomum_sp9 0 0 0 0 0 0 0 1 0 0 0 0 Amomum_compactum2 0 0 0 0 0 0 0 1 0 0 0 0 Amomum_pseudofoetens 0 0 0 0 0 0 0 1 0 0 0 0 Alpinia_elegans 0 0 0 0 0 0 0 0 0 0 1 0 Alpinia_modesta 0 0 0 0 0 0 0 0 0 0 0 1 Alpinia_arundelliana 0 0 0 0 0 0 0 0 0 0 0 1 Geocharis_sp1 0 0 0 0 0 0 0 1 0 0 0 0 Vanoverberghia_sepulchrei 0 0 0 0 0 0 0 0 0 0 1 0 Etlingera_elatior1 0 0 0 0 0 0 0 1 0 0 0 0 Etlingera_triorgyalis 0 0 0 0 0 0 0 1 0 0 0 0 Amomum_propinquum 0 0 0 0 0 0 0 0 0 0 1 0 Amomum_koenigii 0 0 0 0 0 1 0 0 0 0 0 0 Amomum_yunnanense 0 0 0 0 0 1 0 0 0 0 0 0 Amomum_sericeum 0 0 0 0 0 1 0 0 0 0 0 0 Elettariopsis_smithiae 0 0 0 0 0 0 1 0 0 0 0 0 Amomum_menglaense 0 0 0 0 0 1 0 0 0 0 0 0 Amomum_uliginosum1 0 0 0 0 0 0 0 1 0 0 0 0 Elettariopsis_sp 0 0 0 0 0 0 0 1 0 0 0 0 Amomum_putrescens 0 0 0 0 0 1 0 0 0 0 0 0 Amomum_queenslandicum 0 0 0 0 0 0 0 0 0 0 0 1 Amomum_austrosinense 0 0 0 0 0 1 0 0 0 0 0 0 Amomum_purpureorubrum 0 0 0 0 0 1 0 0 0 0 0 0 Amomum_maximum 0 0 0 0 0 1 0 0 0 0 0 0 Amomum_glabrum2 0 0 0 0 0 1 0 0 0 0 0 0 Aframomum_angustifolium 0 0 1 0 0 0 0 0 0 0 0 0 Alpinia_galanga2 0 0 0 0 0 0 0 1 0 0 0 0 Alpinia_conchigera 0 0 0 0 0 1 0 0 0 0 0 0 Elettariopsis_kerbyi 0 0 0 0 0 0 0 1 0 0 0 0 Elettariopsis_stenosiphon 0 0 0 0 0 0 0 1 0 0 0 0 Amomum_petaloideum 0 0 0 0 0 1 1 0 0 0 0 0 Zingiber_cylindricum 0 0 0 0 1 0 0 0 0 0 0 0 229 Amomum_pterocarpum 0 0 0 1 1 0 0 0 0 0 0 0 Zingiber_wightianum 0 0 0 1 1 0 0 0 0 0 0 0 Amomum_sp1 0 0 0 0 1 0 0 0 0 0 0 0 Elettaria_sp 0 0 0 1 1 0 0 0 0 0 0 0 Curcuma_zedoaria 0 0 0 1 1 0 0 0 0 0 0 0 Amomum_fulviceps 0 0 0 1 1 0 0 0 0 0 0 0 Amomum_sp2 0 0 0 0 1 0 0 0 0 0 0 0 Curcuma_albiflora 0 0 0 0 1 0 0 0 0 0 0 0 Amomum_graminifolium 0 0 0 0 1 0 0 0 0 0 0 0 Amomum_sp3 0 0 0 0 1 0 0 0 0 0 0 0 Amomum_nemorale 0 0 0 0 1 0 0 0 0 0 0 0 Cyphostigma_pulchellum 0 0 0 0 1 0 0 0 0 0 0 0 Amomum_acuminatum 0 0 0 0 1 0 0 0 0 0 0 0 Amomum_masticatorium 0 0 0 1 1 0 0 0 0 0 0 0 Alpinia_abundiflora1 0 0 0 1 1 0 0 0 0 0 0 0 Amomum_sp5 0 0 0 0 1 0 0 0 0 0 0 0 Amomum_sp6 0 0 0 0 1 0 0 0 0 0 0 0 Amomum_sp7 0 0 0 0 1 0 0 0 0 0 0 0 Alpinia_sp 0 0 0 0 1 0 0 0 0 0 0 0 Aframomum_daniellii 0 0 1 0 0 0 0 0 0 0 0 0 Aframomum_sceptrum 0 0 1 0 0 0 0 0 0 0 0 0 Renealmia_battenbergiana 0 1 0 0 0 0 0 0 0 0 0 0 Geocharis_sp2 0 0 0 0 0 0 0 1 0 0 0 0 Alpinia_fax 0 0 0 0 1 0 0 0 0 0 0 0 230 Appendix 7:Example script used in Biogeobears in ancestral area reconstructions. ####################################################### # This is an introductory example script for the # R package "BioGeoBEARS" by Nick Matzke # # All scripts are copyright Nicholas J. Matzke, # please cite if you use. License: GPL-3 # http://cran.r-project.org/web/licenses/GPL-3 # # I am happy to answer questions at matzke@nimbios.org, but # I am more happy to answer questions on the # BioGeoBEARS google group # # The package is designed for ML and Bayesian inference # of # # (a) ancestral geographic ranges, and # # (b) perhaps more importantly, models for the # evolution of geographic range across a phylogeny. # # The example below implements and compares: # # (1) The standard 2-parameter DEC model implemented in # the program LAGRANGE (Ree & Smith 2008); users will # notice that the ML parameter inference and log- # likelihoods are identical # # (2) A DEC+J model implemented in BioGeoBEARS, wherein # a third parameter, j, is added, representing the # relative per-event weight of founder-event / jump # speciation events at cladogenesis events. The # higher j is, the more probability these events have, # and the less probability the standard LAGRANGE 231 # cladogenesis events have. # # (3) Some standard model-testing (LRT and AIC) is # implemented at the end so that users may compare models # # (4) The script does similar tests of a DIVA-like model (Ronquist 1997) # and a BAYAREA-like model (Landis, Matzke, Moore, & Huelsenbeck, 2013) # ####################################################### ####################################################### # Installing BioGeoBEARS ####################################################### # Uncomment this command to get everything # Please use the "0-cloud" R repository at "http://cran.rstudio.com" as it is # the only one that keeps download statistics ####################################################### # # # # Install BioGeoBEARS from CRAN 0-cloud: # install.packages("BioGeoBEARS", dependencies=TRUE, repos="http://cran.rstudio.com") # ####################################################### ####################################################### # SETUP -- libraries/BioGeoBEARS updates ####################################################### # Load the package (after installation, see above). library(optimx) # You need to have some version of optimx available # as it is a BioGeoBEARS dependency; however, if you # don't want to use optimx, and use optim() (from R core) # you can set: # BioGeoBEARS_run_object$use_optimx = FALSE # ...everything should work either way -- NJM 2014-01-08 232 library(FD) library(snow) # for FD::maxent() (make sure this is up-to-date) # (if you want to use multicore functionality; some systems/R versions prefer library(parallel), try either) library(parallel) library(BioGeoBEARS) ######################################################## # TO GET THE OPTIMX/OPTIM FIX, AND THE UPPASS FIX, # SOURCE THE REVISED FUNCTIONS WITH THESE COMMANDS # # CRUCIAL CRUCIAL CRUCIAL: # YOU HAVE TO RUN THE SOURCE COMMANDS AFTER # *EVERY TIME* YOU DO library(BioGeoBEARS). THE CHANGES ARE NOT "PERMANENT", # THEY HAVE TO BE MADE EACH TIME. IF YOU ARE GOING TO BE OFFLINE, # YOU CAN DOWNLOAD EACH .R FILE TO YOUR HARD DRIVE AND REFER THE source() # COMMANDS TO THE FULL PATH AND FILENAME OF EACH FILE ON YOUR # LOCAL SYSTEM INSTEAD. ######################################################## library(BioGeoBEARS) source("http://phylo.wdfiles.com/local--files/biogeobears/BioGeoBEARS_basics_v1.R") source("http://phylo.wdfiles.com/local--files/biogeobears/BioGeoBEARS_generics_v1.R") source("http://phylo.wdfiles.com/local--files/biogeobears/BioGeoBEARS_readwrite_v1.R") source("http://phylo.wdfiles.com/local--files/biogeobears/BioGeoBEARS_classes_v1.R") source("http://phylo.wdfiles.com/local-files/biogeobears/BioGeoBEARS_calc_transition_matrices_v1.R") source("http://phylo.wdfiles.com/local--files/biogeobears/BioGeoBEARS_detection_v1.R") source("http://phylo.wdfiles.com/local--files/biogeobears/BioGeoBEARS_models_v1.R") source("http://phylo.wdfiles.com/local-files/biogeobears/BioGeoBEARS_extract_Qmat_COOmat_v1.R") source("http://phylo.wdfiles.com/local--files/biogeobears/BioGeoBEARS_univ_model_v1.R") source("http://phylo.wdfiles.com/local--files/biogeobears/BioGeoBEARS_stratified_v1.R") source("http://phylo.wdfiles.com/local--files/biogeobears/BioGeoBEARS_plots_v1.R") source("http://phylo.wdfiles.com/local--files/biogeobears/BioGeoBEARS_simulate_v1.R") 233 source("http://phylo.wdfiles.com/local--files/biogeobears/calc_loglike_sp_v01.R") source("http://phylo.wdfiles.com/local--files/biogeobears/calc_uppass_probs_v1.R") source("http://phylo.wdfiles.com/local-files/biogeobears/get_stratified_subbranch_top_downpass_likelihoods_v1.R") calc_loglike_sp = compiler::cmpfun(calc_loglike_sp_prebyte) # crucial to fix bug in uppass calculations calc_independent_likelihoods_on_each_branch = compiler::cmpfun(calc_independent_likelihoods_on_each_branch_prebyte) # slight speedup hopefully ####################################################### # SETUP: YOUR WORKING DIRECTORY ####################################################### # You will need to set your working directory to match your local system # Note these very handy functions! # Command "setwd(x)" sets your working directory # Command "getwd()" gets your working directory and tells you what it is. # Command "list.files()" lists the files in your working directory # To get help on any command, use "?". E.g., "?list.files" # Set your working directory for output files # default here is your home directory ("~") # Change this as you like wd = np("C:/BGB ") setwd(wd) # Double-check your working directory with getwd() getwd() ####################################################### # SETUP: Extension data directory ####################################################### # When R packages contain extra files, they are stored in the "extdata" directory # inside the installed package. 234 # # BioGeoBEARS contains various example files and scripts in its extdata directory. # # Each computer operating system might install BioGeoBEARS in a different place, # depending on your OS and settings. # # However, you can find the extdata directory like this: extdata_dir = np(system.file("extdata", package="BioGeoBEARS")) extdata_dir list.files(extdata_dir) # "system.file" looks in the directory of a specified package (in this case BioGeoBEARS) # The function "np" is just a shortcut for normalizePath(), which converts the # path to the format appropriate for your system (e.g., Mac/Linux use "/", but # Windows uses "\\", if memory serves). # Even when using your own data files, you should KEEP these commands in your # script, since the plot_BioGeoBEARS_results function needs a script from the # extdata directory to calculate the positions of "corners" on the plot. This cannot # be made into a straight up BioGeoBEARS function because it uses C routines # from the package APE which do not pass R CMD check for some reason. ####################################################### # SETUP: YOUR TREE FILE AND GEOGRAPHY FILE ####################################################### # Example files are given below. To run your own data, # make the below lines point to your own files, e.g. # trfn = "/mydata/frogs/frogBGB/tree.newick" # geogfn = "/mydata/frogs/frogBGB/geog.data" ####################################################### # Phylogeny file # Notes: # 1. Must be binary/bifurcating: no polytomies 235 # 2. No negative branchlengths (e.g. BEAST MCC consensus trees sometimes have negative branchlengths) # 3. Be careful of very short branches, as BioGeoBEARS will interpret ultrashort branches as direct ancestors # 4. You can use non-ultrametric trees, but BioGeoBEARS will interpret any tips significantly below the # top of the tree as fossils! This is only a good idea if you actually do have fossils in your tree, # as in e.g. Wood, Matzke et al. (2013), Systematic Biology. # 5. The default settings of BioGeoBEARS make sense for trees where the branchlengths are in units of # millions of years, and the tree is 1-1000 units tall. If you have a tree with a total height of # e.g. 0.00001, you will need to adjust e.g. the max values of d and e, or (simpler) multiply all # your branchlengths to get them into reasonable units. # 6. DON'T USE SPACES IN SPECIES NAMES, USE E.G. "_" ####################################################### # This is the example Newick file for Hawaiian Psychotria # (from Ree & Smith 2008) # "trfn" = "example.newick" trfn = "example.newick" # Look at the raw Newick file: moref(trfn) # Look at your phylogeny: tr = read.tree(trfn) tr plot(tr) title("Example Psychotria phylogeny from Ree & Smith (2008)") axisPhylo() # plots timescale ####################################################### # Geography file # Notes: # 1. This is a PHLYIP-formatted file. This means that in the 236 # first line, # - the 1st number equals the number of rows (species) # - the 2nd number equals the number of columns (number of areas) # 2. This is the same format used for C++ LAGRANGE geography files. # 3. All names in the geography file must match names in the phylogeny file. # 4. DON'T USE SPACES IN SPECIES NAMES, USE E.G. "_" # 5. Operational taxonomic units (OTUs) should ideally be phylogenetic lineages, # i.e. genetically isolated populations. These may or may not be identical # with species. You would NOT want to just use specimens, as each specimen # automatically can only live in 1 area, which will typically favor DEC+J # models. This is fine if the species/lineages really do live in single areas, # but you wouldn't want to assume this without thinking about it at least. # In summary, you should collapse multiple specimens into species/lineages if # data indicates they are the same genetic population. ###################################################### # This is the example geography file for Hawaiian Psychotria # (from Ree & Smith 2008) geogfn = "example.txt" # Look at the raw geography text file: moref(geogfn) # Look at your geographic range data: tipranges = getranges_from_LagrangePHYLIP(lgdata_fn=geogfn) tipranges # Set the maximum number of areas any species may occupy; this cannot be larger # than the number of areas you set up, but it can be smaller. max_range_size = 2 #################################################### #################################################### # KEY HINT: The number of states (= number of different possible geographic ranges) # depends on (a) the number of areas and (b) max_range_size. 237 # If you have more than about 500-600 states, the calculations will get REALLY slow, # since the program has to exponentiate a matrix of e.g. 600x600. Often the computer # will just sit there and crunch, and never get through the calculation of the first # likelihood. # # (this is also what is usually happening when LAGRANGE hangs: you have too many states!) # # To check the number of states for a given number of ranges, try: numstates_from_numareas(numareas=4, maxareas=4, include_null_range=TRUE) numstates_from_numareas(numareas=4, maxareas=4, include_null_range=FALSE) numstates_from_numareas(numareas=4, maxareas=3, include_null_range=TRUE) numstates_from_numareas(numareas=4, maxareas=2, include_null_range=TRUE) # Large numbers of areas have problems: numstates_from_numareas(numareas=10, maxareas=10, include_null_range=TRUE) # ...unless you limit the max_range_size: numstates_from_numareas(numareas=10, maxareas=2, include_null_range=TRUE) #################################################### #################################################### ####################################################### # DEC AND DEC+J ANALYSIS ####################################################### ####################################################### # Run DEC ####################################################### BioGeoBEARS_run_object = define_BioGeoBEARS_run() BioGeoBEARS_run_object$force_sparse=FALSE # sparse=FALSE causes pathology & isn't much faster at this scale BioGeoBEARS_run_object$speedup=TRUE # shorcuts to speed ML search; use FALSE if worried (e.g. >3 params) BioGeoBEARS_run_object$use_optimx = TRUE 238 BioGeoBEARS_run_object$calc_ancprobs=TRUE # get ancestral states from optim run # Set up a time-stratified analysis # (un-comment to use; see example files in extdata_dir, # and BioGeoBEARS google group posts for further hints) #BioGeoBEARS_run_object$timesfn = "timeperiods.txt" #BioGeoBEARS_run_object$dispersal_multipliers_fn = "manual_dispersal_multipliers.txt" #BioGeoBEARS_run_object$areas_allowed_fn = "areas_allowed.txt" #BioGeoBEARS_run_object$areas_adjacency_fn = "areas_adjacency.txt" #BioGeoBEARS_run_object$distsfn = "distances_matrix.txt" # See notes on the distances model on PhyloWiki's BioGeoBEARS updates page. # Input the maximum range size BioGeoBEARS_run_object$max_range_size = max_range_size # Multicore processing if desired BioGeoBEARS_run_object$num_cores_to_use=1 # (use more cores to speed it up; this requires # library(parallel) and/or library(snow). Parellel, # is default on Macs in R 3.0+, but apparently still # has to be typed on Windows machines. Note: apparently # parallel works on Mac command-line R, but not R.app. # BioGeoBEARS checks for this and resets to 1 # core with R.app) # Sparse matrix exponentiation is an option for huge numbers of ranges/states (600+) # I have experimented with sparse matrix exponentiation in EXPOKIT/rexpokit, # but the results are imprecise and so I haven't explored it further. # In a Bayesian analysis, it might work OK, but the ML point estimates are # not identical. # Also, I have not implemented all functions to work with force_sparse=TRUE. # Volunteers are welcome to work on it!! BioGeoBEARS_run_object$force_sparse=FALSE # Give BioGeoBEARS the location of the geography text file 239 BioGeoBEARS_run_object$geogfn = geogfn # Give BioGeoBEARS the location of the phylogeny Newick file BioGeoBEARS_run_object$trfn = trfn # This function loads the dispersal multiplier matrix etc. from the text files into the model object. Required for these to work! # (It also runs some checks on these inputs for certain errors.) BioGeoBEARS_run_object = readfiles_BioGeoBEARS_run(BioGeoBEARS_run_object) # Divide the tree up by timeperiods/strata (uncomment this for stratified analysis) #BioGeoBEARS_run_object = section_the_tree(inputs=BioGeoBEARS_run_object, make_master_table=TRUE, plot_pieces=FALSE) # The stratified tree is described in this table: #BioGeoBEARS_run_object$master_table # Good default settings to get ancestral states BioGeoBEARS_run_object$return_condlikes_table = TRUE BioGeoBEARS_run_object$calc_TTL_loglike_from_condlikes_table = TRUE BioGeoBEARS_run_object$calc_ancprobs = TRUE # Set up DEC model # (nothing to do; defaults) # Look at the BioGeoBEARS_run_object; it's just a list of settings etc. BioGeoBEARS_run_object # This contains the model object BioGeoBEARS_run_object$BioGeoBEARS_model_object # This table contains the parameters of the model BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table # Run this to check inputs. Read the error messages if you get them! check_BioGeoBEARS_run(BioGeoBEARS_run_object) 240 # For a slow analysis, run once, then set runslow=FALSE to just # load the saved result. runslow = TRUE resfn = "Psychotria_DEC_M0_unconstrained_v1.Rdata" if (runslow) { res = bears_optim_run(BioGeoBEARS_run_object) res save(res, file=resfn) resDEC = res } else { # Loads to "res" load(resfn) resDEC = res } ####################################################### # Run DEC+J ####################################################### BioGeoBEARS_run_object = define_BioGeoBEARS_run() BioGeoBEARS_run_object$trfn = trfn BioGeoBEARS_run_object$geogfn = geogfn BioGeoBEARS_run_object$max_range_size = max_range_size # Set up the stratified part #BioGeoBEARS_run_object$timesfn = "timeperiods.txt" #BioGeoBEARS_run_object$dispersal_multipliers_fn = "manual_dispersal_multipliers.txt" #BioGeoBEARS_run_object$areas_allowed_fn = "areas_allowed.txt" #BioGeoBEARS_run_object$areas_adjacency_fn = "areas_adjacency.txt" #BioGeoBEARS_run_object$distsfn = "distances_matrix.txt" # See notes on the distances model on PhyloWiki's BioGeoBEARS updates page. 241 BioGeoBEARS_run_object$speedup=TRUE # shorcuts to speed ML search; use FALSE if worried (e.g. >3 params) BioGeoBEARS_run_object$use_optimx = TRUE BioGeoBEARS_run_object$num_cores_to_use=1 BioGeoBEARS_run_object$force_sparse=FALSE # sparse=FALSE causes pathology & isn't much faster at this scale BioGeoBEARS_run_object$calc_ancprobs=TRUE # get ancestral states from optim run # This function loads the dispersal multiplier matrix etc. from the text files into the model object. Required for these to work! # (It also runs some checks on these inputs for certain errors.) BioGeoBEARS_run_object = readfiles_BioGeoBEARS_run(BioGeoBEARS_run_object) # Divide the tree up by timeperiods/strata (uncomment this for stratified analysis) #BioGeoBEARS_run_object = section_the_tree(inputs=BioGeoBEARS_run_object, make_master_table=TRUE, plot_pieces=FALSE) # The stratified tree is described in this table: #BioGeoBEARS_run_object$master_table # Good default settings to get ancestral states BioGeoBEARS_run_object$return_condlikes_table = TRUE BioGeoBEARS_run_object$calc_TTL_loglike_from_condlikes_table = TRUE BioGeoBEARS_run_object$calc_ancprobs = TRUE # Set up DEC+J model # Get the ML parameter values from the 2-parameter nested model # (this will ensure that the 3-parameter model always does at least as good) dstart = resDEC$outputs@params_table["d","est"] estart = resDEC$outputs@params_table["e","est"] jstart = 0.0001 # Input starting values for d, e BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["d","init"] = dstart BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["d","est"] = dstart BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["e","init"] = estart 242 BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["e","est"] = estart # Add j as a free parameter BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["j","type"] = "free" BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["j","init"] = jstart BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["j","est"] = jstart check_BioGeoBEARS_run(BioGeoBEARS_run_object) resfn = "Psychotria_DEC+J_M0_unconstrained_v1.Rdata" runslow = TRUE if (runslow) { #sourceall("/Dropbox/_njm/__packages/BioGeoBEARS_setup/") res = bears_optim_run(BioGeoBEARS_run_object) res save(res, file=resfn) resDECj = res } else { # Loads to "res" load(resfn) resDECj = res } ####################################################### # PDF plots ####################################################### pdffn = "Psychotria_DEC_vs_DEC+J_M0_unconstrained_v1.pdf" pdf(pdffn, width=34, height=44) ####################################################### # Plot ancestral states - DEC 243 ####################################################### analysis_titletxt ="BioGeoBEARS DEC on Psychotria M0_unconstrained" # Setup results_object = resDEC scriptdir = np(system.file("extdata/a_scripts", package="BioGeoBEARS")) # States res2 = plot_BioGeoBEARS_results(results_object, analysis_titletxt, addl_params=list("j"), plotwhat="text", label.offset=0.45, tipcex=0.7, statecex=0.7, splitcex=0.6, titlecex=0.8, plotsplits=TRUE, cornercoords_loc=scriptdir, include_null_range=TRUE, tr=tr, tipranges=tipranges) # Pie chart plot_BioGeoBEARS_results(results_object, analysis_titletxt, addl_params=list("j"), plotwhat="pie", label.offset=0.45, tipcex=0.2, statecex=0.1, splitcex=0.1, titlecex=0.2, plotsplits=TRUE, cornercoords_loc=scriptdir, include_null_range=TRUE, tr=tr, tipranges=tipranges) ####################################################### # Plot ancestral states - DECJ ####################################################### analysis_titletxt ="BioGeoBEARS DEC+J on Psychotria M0_unconstrained" # Setup results_object = resDECj scriptdir = np(system.file("extdata/a_scripts", package="BioGeoBEARS")) # States res1 = plot_BioGeoBEARS_results(results_object, analysis_titletxt, addl_params=list("j"), plotwhat="text", label.offset=0.45, tipcex=0.7, statecex=0.7, splitcex=0.6, titlecex=0.8, plotsplits=TRUE, cornercoords_loc=scriptdir, include_null_range=TRUE, tr=tr, tipranges=tipranges) # Pie chart 244 plot_BioGeoBEARS_results(results_object, analysis_titletxt, addl_params=list("j"), plotwhat="pie", label.offset=0.45, tipcex=0.2, statecex=0.1, splitcex=0.1, titlecex=0.2, plotsplits=TRUE, cornercoords_loc=scriptdir, include_null_range=TRUE, tr=tr, tipranges=tipranges) dev.off() # Turn off PDF cmdstr = paste("open ", pdffn, sep="") system(cmdstr) # Plot it ####################################################### # DIVALIKE AND DIVALIKE+J ANALYSIS ####################################################### # NOTE: The BioGeoBEARS "DIVA" model is not identical with # Ronquist (1997)'s parsimony DIVA. It is a likelihood # interpretation of DIVA, constructed by modelling DIVA's # processes the way DEC does, but only allowing the # processes DIVA allows (widespread vicariance: yes; subset # sympatry: no; see Ronquist & Sanmartin 2011, Figure 4). # # I thus now call the model "DIVALIKE", and you should also. ;-) # ####################################################### # Run DIVALIKE ####################################################### BioGeoBEARS_run_object = define_BioGeoBEARS_run() BioGeoBEARS_run_object$trfn = trfn BioGeoBEARS_run_object$geogfn = geogfn BioGeoBEARS_run_object$max_range_size = max_range_size # Set up the stratified part #BioGeoBEARS_run_object$timesfn = "timeperiods.txt" #BioGeoBEARS_run_object$dispersal_multipliers_fn = "manual_dispersal_multipliers.txt" #BioGeoBEARS_run_object$areas_allowed_fn = "areas_allowed.txt" 245 #BioGeoBEARS_run_object$areas_adjacency_fn = "areas_adjacency.txt" #BioGeoBEARS_run_object$distsfn = "distances_matrix.txt" # See notes on the distances model on PhyloWiki's BioGeoBEARS updates page. BioGeoBEARS_run_object$use_optimx = TRUE BioGeoBEARS_run_object$num_cores_to_use=1 BioGeoBEARS_run_object$force_sparse=FALSE # sparse=FALSE causes pathology & isn't much faster at this scale BioGeoBEARS_run_object$speedup=TRUE # shorcuts to speed ML search; use FALSE if worried (e.g. >3 params) BioGeoBEARS_run_object$calc_ancprobs=TRUE # get ancestral states from optim run # This function loads the dispersal multiplier matrix etc. from the text files into the model object. Required for these to work! # (It also runs some checks on these inputs for certain errors.) BioGeoBEARS_run_object = readfiles_BioGeoBEARS_run(BioGeoBEARS_run_object) # Divide the tree up by timeperiods/strata (uncomment this for stratified analysis) #BioGeoBEARS_run_object = section_the_tree(inputs=BioGeoBEARS_run_object, make_master_table=TRUE, plot_pieces=FALSE) # The stratified tree is described in this table: #BioGeoBEARS_run_object$master_table # Good default settings to get ancestral states BioGeoBEARS_run_object$return_condlikes_table = TRUE BioGeoBEARS_run_object$calc_TTL_loglike_from_condlikes_table = TRUE BioGeoBEARS_run_object$calc_ancprobs = TRUE # Set up DIVALIKE model # Remove subset-sympatry BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["s","type"] = "fixed" BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["s","init"] = 0.0 BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["s","est"] = 0.0 246 BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["ysv","type"] = "2j" BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["ys","type"] = "ysv*1/2" BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["y","type"] = "ysv*1/2" BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["v","type"] = "ysv*1/2" # Allow classic, widespread vicariance; all events equiprobable BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["mx01v","type"] = "fixed" BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["mx01v","init"] = 0.5 BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["mx01v","est"] = 0.5 # No jump dispersal/founder-event speciation # BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["j","type"] = "free" # BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["j","init"] = 0.01 # BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["j","est"] = 0.01 check_BioGeoBEARS_run(BioGeoBEARS_run_object) runslow = TRUE resfn = "Psychotria_DIVALIKE_M0_unconstrained_v1.Rdata" if (runslow) { res = bears_optim_run(BioGeoBEARS_run_object) res save(res, file=resfn) resDIVALIKE = res 247 } else { # Loads to "res" load(resfn) resDIVALIKE = res } ####################################################### # Run DIVALIKE+J ####################################################### BioGeoBEARS_run_object = define_BioGeoBEARS_run() BioGeoBEARS_run_object$trfn = trfn BioGeoBEARS_run_object$geogfn = geogfn BioGeoBEARS_run_object$max_range_size = max_range_size # Set up the stratified part #BioGeoBEARS_run_object$timesfn = "timeperiods.txt" #BioGeoBEARS_run_object$dispersal_multipliers_fn = "manual_dispersal_multipliers.txt" #BioGeoBEARS_run_object$areas_allowed_fn = "areas_allowed.txt" #BioGeoBEARS_run_object$areas_adjacency_fn = "areas_adjacency.txt" #BioGeoBEARS_run_object$distsfn = "distances_matrix.txt" # See notes on the distances model on PhyloWiki's BioGeoBEARS updates page. BioGeoBEARS_run_object$use_optimx = TRUE BioGeoBEARS_run_object$num_cores_to_use=1 BioGeoBEARS_run_object$force_sparse=FALSE # sparse=FALSE causes pathology & isn't much faster at this scale BioGeoBEARS_run_object$speedup=TRUE # shorcuts to speed ML search; use FALSE if worried (e.g. >3 params) BioGeoBEARS_run_object$calc_ancprobs=TRUE # get ancestral states from optim run # This function loads the dispersal multiplier matrix etc. from the text files into the model object. Required for these to work! # (It also runs some checks on these inputs for certain errors.) BioGeoBEARS_run_object = readfiles_BioGeoBEARS_run(BioGeoBEARS_run_object) 248 # Divide the tree up by timeperiods/strata (uncomment this for stratified analysis) #BioGeoBEARS_run_object = section_the_tree(inputs=BioGeoBEARS_run_object, make_master_table=TRUE, plot_pieces=FALSE) # The stratified tree is described in this table: #BioGeoBEARS_run_object$master_table # Good default settings to get ancestral states BioGeoBEARS_run_object$return_condlikes_table = TRUE BioGeoBEARS_run_object$calc_TTL_loglike_from_condlikes_table = TRUE BioGeoBEARS_run_object$calc_ancprobs = TRUE # Set up DIVALIKE+J model # Get the ML parameter values from the 2-parameter nested model # (this will ensure that the 3-parameter model always does at least as good) dstart = resDIVALIKE$outputs@params_table["d","est"] estart = resDIVALIKE$outputs@params_table["e","est"] jstart = 0.0001 # Input starting values for d, e BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["d","init"] = dstart BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["d","est"] = dstart BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["e","init"] = estart BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["e","est"] = estart # Remove subset-sympatry BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["s","type"] = "fixed" BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["s","init"] = 0.0 BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["s","est"] = 0.0 BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["ysv","type"] = "2j" BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["ys","type"] = "ysv*1/2" 249 BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["y","type"] = "ysv*1/2" BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["v","type"] = "ysv*1/2" # Allow classic, widespread vicariance; all events equiprobable BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["mx01v","type"] = "fixed" BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["mx01v","init"] = 0.5 BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["mx01v","est"] = 0.5 # Add jump dispersal/founder-event speciation BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["j","type"] = "free" BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["j","init"] = jstart BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["j","est"] = jstart # Under DIVALIKE+J, the max of "j" should be 2, not 3 (as is default in DEC+J) BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["j","min"] = 0.00001 BioGeoBEARS_run_object$BioGeoBEARS_model_object@params_table["j","max"] = 1.99999 check_BioGeoBEARS_run(BioGeoBEARS_run_object) resfn = "Psychotria_DIVALIKE+J_M0_unconstrained_v1.Rdata" runslow = TRUE if (runslow) { #sourceall("/Dropbox/_njm/__packages/BioGeoBEARS_setup/") res = bears_optim_run(BioGeoBEARS_run_object) res 250 save(res, file=resfn) resDIVALIKEj = res } else { # Loads to "res" load(resfn) resDIVALIKEj = res } pdffn = "Psychotria_DIVALIKE_vs_DIVALIKE+J_M0_unconstrained_v1.pdf" pdf(pdffn, width=24, height=24) ####################################################### # Plot ancestral states - DIVALIKE ####################################################### analysis_titletxt ="BioGeoBEARS DIVALIKE on Psychotria M0_unconstrained" # Setup results_object = resDIVALIKE scriptdir = np(system.file("extdata/a_scripts", package="BioGeoBEARS")) # States res2 = plot_BioGeoBEARS_results(results_object, analysis_titletxt, addl_params=list("j"), plotwhat="text", label.offset=0.45, tipcex=0.7, statecex=0.7, splitcex=0.6, titlecex=0.8, plotsplits=TRUE, cornercoords_loc=scriptdir, include_null_range=TRUE, tr=tr, tipranges=tipranges) # Pie chart plot_BioGeoBEARS_results(results_object, analysis_titletxt, addl_params=list("j"), plotwhat="pie", label.offset=0.45, tipcex=0.2, statecex=0.1, splitcex=0.1, titlecex=0.2, plotsplits=TRUE, cornercoords_loc=scriptdir, include_null_range=TRUE, tr=tr, tipranges=tipranges) ####################################################### # Plot ancestral states - DIVALIKE+J 251 ####################################################### analysis_titletxt ="BioGeoBEARS DIVALIKE+J on Psychotria M0_unconstrained" # Setup results_object = resDIVALIKEj scriptdir = np(system.file("extdata/a_scripts", package="BioGeoBEARS")) # States res1 = plot_BioGeoBEARS_results(results_object, analysis_titletxt, addl_params=list("j"), plotwhat="text", label.offset=0.45, tipcex=0.7, statecex=0.7, splitcex=0.6, titlecex=0.8, plotsplits=TRUE, cornercoords_loc=scriptdir, include_null_range=TRUE, tr=tr, tipranges=tipranges) # Pie chart plot_BioGeoBEARS_results(results_object, analysis_titletxt, addl_params=list("j"), plotwhat="pie", label.offset=0.45, tipcex=0.2, statecex=0.1, splitcex=0.1, titlecex=0.2, plotsplits=TRUE, cornercoords_loc=scriptdir, include_null_range=TRUE, tr=tr, tipranges=tipranges) dev.off() cmdstr = paste("open ", pdffn, sep="") system(cmdstr) ####################################################### # SETUP: YOUR TREE FILE AND GEOGRAPHY FILE ####################################################### # Setup library(ape) library(BioGeoBEARS) # for get_lagrange_nodenums(), postorder_nodes_phylo4_return_table() library(phylobase) # required for postorder_nodes_phylo4_return_table() # Fix to postorder_nodes_phylo4_return_table source("http://phylo.wdfiles.com/local--files/biogeobears/BioGeoBEARS_readwrite_v1.R") 252 # Set your working directory to put the PDF in (pick your own) # Default getwd() # Example files are given below. To run your own data, # make the below lines point to your own files, e.g. trfn = "C:/BGB/example.newick" geogfn = "C:/BGB/example.txt" # Look at your phylogeny: tr = read.tree(trfn) tr plot(tr, cex=0.5) title("Example_BGB") axisPhylo() # plots timescale ################################################### # Plot APE/BioGeoBEARS node numbers ################################################### ntips = length(tr$tip.label) Rnodenums = (ntips+1):(ntips+tr$Nnode) tipnums = 1:ntips plot(tr, label.offset=0.25, cex=1.25) axisPhylo() tiplabels(cex=1.5) nodelabels(text=Rnodenums, node=Rnodenums, cex=0.5) title("APE/BioGeoBEARS node numbers") # END PDF dev.off() cmdstr = paste("open ", pdffn, sep="") system(cmdstr) prt(tr) names(res) 253 [1] "computed_likelihoods_at_each_node" [2] "relative_probs_of_each_state_at_branch_top_AT_node_DOWNPASS" [3] "condlikes_of_each_state" [4] "relative_probs_of_each_state_at_branch_bottom_below_node_DOWNPASS" [5] "relative_probs_of_each_state_at_branch_bottom_below_node_UPPASS" [6] "relative_probs_of_each_state_at_branch_top_AT_node_UPPASS" [7] "ML_marginal_prob_each_state_at_branch_bottom_below_node" [8] "ML_marginal_prob_each_state_at_branch_top_AT_node" [9] "relative_probs_of_each_state_at_bottom_of_root_branch" [10] "total_loglikelihood" [11] "inputs" [12] "outputs" [13] "optim_result" res$computed_likelihoods_at_each_node write.table(res$ computed_likelihoods_at_each_node, file="sap1.txt", sep="XTABX") res$relative_probs_of_each_state_at_branch_top_AT_node_DOWNPASS write.table(res$relative_probs_of_each_state_at_branch_top_AT_node_DOWNPASS , file="sap2.txt", sep="XTABX") res$condlikes_of_each_state write.table(res$condlikes_of_each_state, file="sap3.txt", sep="XTABX") res$relative_probs_of_each_state_at_branch_bottom_below_node_DOWNPASS write.table(res$relative_probs_of_each_state_at_branch_bottom_below_node_DOWNPASS , file="sap4.txt", sep="XTABX") res$relative_probs_of_each_state_at_branch_bottom_below_node_UPPASS 254 write.table(res$relative_probs_of_each_state_at_branch_bottom_below_node_UPPASS, file="sap5.txt", sep="XTABX") res$relative_probs_of_each_state_at_branch_top_AT_node_UPPASS write.table(res$relative_probs_of_each_state_at_branch_top_AT_node_UPPASS , file="sap6.txt", sep="XTABX") res$ML_marginal_prob_each_state_at_branch_bottom_below_node write.table(res$ML_marginal_prob_each_state_at_branch_bottom_below_node, file="sap7.txt", sep="XTABX") res$ML_marginal_prob_each_state_at_branch_top_AT_node write.table(res$ML_marginal_prob_each_state_at_branch_top_AT_node , file="sap8.txt", sep="XTABX") res$relative_probs_of_each_state_at_bottom_of_root_branch write.table(res$relative_probs_of_each_state_at_bottom_of_root_branch , file="sap9.txt", sep="XTABX") res$total_loglikelihood write.table(res$ML_marginal_prob_each_state_at_branch_top_AT_node, file="B.txt", sep="XTABX") res$inputs write.table(res$ML_marginal_prob_each_state_at_branch_top_AT_node, file="B.txt", sep="XTABX") res$outputs write.table(res$ML_marginal_prob_each_state_at_branch_top_AT_node, file="B.txt", sep="XTABX") res$optim_result write.table(res$ML_marginal_prob_each_state_at_branch_top_AT_node, file="B.txt", sep="XTABX") 255 Appendix 8: DNA sequence alignments and Biogeobears result output for Begoniaceae, Sapotaceae, Zingiberaceae Files on the accompanying DVD comprise DNA sequence data and sequence alignments in FASTA format and Biogeobears ancestral area probabilities for each family in excel format File Contents Chapter 3 Reference alignmet 179 taxa Begoniaceae_cpDNA.fas Alignment of concatenated ndhA intron, ndhF-rpl32, rpl32-trnL sequences of 179 taxa in family Begoniaceae Chapter 4 Reference alignmet 163 taxa Sapotaceae_ITS.fas Alignment of ITS sequences of 163 taxa in family Sapotaceae Chapter 4 Reference alignmet 136 taxa Sapotaceae_cpDNA.fas Alignment of concatenated trnH-psbA spacer, trnCtrnD, trnC-psbM, psbM-trnD, ndhF sequences of 136 taxa in family Sapotaceae Chapter 5 Reference alignmet 151 taxa Zingiberaceae_ITS.fas Alignment of ITS sequences of 151 taxa in family Zingiberaceae Chapter 5 Reference alignmet 151 taxa Zingiberaceae_cpDNA.fas Alignment of ITS sequences of 151 taxa in family Zingiberaceae Chapter 3 Ancestral area probabilities_Begoniaceae Ancestral area probabilities for each state resulted in the biogeographic analysis of Begoniaceae in Biogeobears Chapter 4 Ancestral area probabilities_Sapotaceae Ancestral area probabilities for each state resulted in the biogeographic analysis of Sapotaceae in Biogeobears Chapter 5 Ancestral area probabilities_Zingiberaceae Ancestral area probabilities for each state resulted in the biogeographic analysis of Zingiberaceae in Biogeobears . 256