Fungal Diversity A new species complex including Claviceps fusiformis and Claviceps hirtella Pažoutová, S.1*, Kolařík, M.1, Odvody, G.N.2, Frederickson, D.E.3, Olšovská, J.1, and Man, P.1 1 Institute of Microbiology ASCR, Vídeňská 1083, 142 20 Prague 4, Czech Republic Corpus Christi Research & Extension Center, 10345 Agnes Street, Corpus Christi, TX 78406-1412 3 Formerly INTSORMIL, c/o ICRISAT Bulawayo Centre, Bulawayo, Zimbabwe 2 Pažoutová, S., Kola ík, M., Olšovská, J., Odvody, G.N., and Frederickson, D.E. (2008). A new species complex including Claviceps fusiformis and Claviceps hirtella. Fungal Diversity 31: 95-110. Isolates of Claviceps species with lunate to fusiform macroconidia were collected from panicoid grasses in Texas and Zimbabwe and described as new species based on anamorphs since no teleomorphs were available. Characterization was based upon morphology and partial sequences of rDNA and β-tubulin. The isolates grouped into two stronglysupported clades. The first clade contained ancestral C. hirtella and C. fusiformis from pearl millet (Pennisetum glaucum) in clade terminal position with Texas isolates from native cup grass (Eriochloa sericea) and pearl millet grouped between them. The second clade consisted of African isolates from Urochloa and Eragrostis. The isolates from Texas from pearl millet and buffel grass (Cenchrus ciliaris) and isolates from E. sericea were described as new species, Sphacelia texensis and Sphacelia eriochloae, respectively. Both species had morphology, DNA markers, and alkaloid production that was intermediate between those features exhibited in C. fusiformis and C. hirtella. The African isolates from Urochloa and Eragrostis were also described as a new species, Sphacelia lovelessii. In shaken cultures, C. hirtella readily produced a whole range of clavines with agroclavine and festuclavine predominating, but ergometrine was also detected. Claviceps fusiformis produced mainly agroclavine and elymoclavine, S. eriochloae produced mainly agroclavine, elymoclavin and festuclavine and the cultures of S. texensis contained small amounts of agroclavine and festuclavine. Only traces of clavines were found in cultures of S. lovelessii of the second clade. The alkaloid content of infected florets in the sphacelial (honeydew) developmental stage was also measured. Only C. fusiformis and S. eriochloae produced alkaloids in planta at this early stage. Key words: alkaloids, β-tubulin, Cenchrus, Claviceps, clavine, conidia, Eragrostis, Eriochloa, pearl millet, Pennisetum, phylogeny, Sphacelia, rDNA, Urochloa Article Information Received 28 May 2007 Accepted 18 October 2007 Published online 31 July 2008 * Corresponding author: S. Pažoutová; e-mail: pazouto@biomed.cas.cz Introduction Since 1997, infections by Claviceps species with lunate to fusiform macroconidia have been found in Texas on heads of introduced pearl millet (Pennisetum glaucum), buffel grass (Cenchrus ciliaris), and native cupgrass (Eriochloa sericea). San Martín et al. (1997) and Velásquez-Valle et al. (1998) detected similar infections in Mexico. However, no teleomorphs of the parasites were observed, mostly due to fungal hyperinfection, precluding species description by traditional sclerotial germination studies. Ergot disease of C. ciliaris was observed in the late 1980’s in Texas (Craig and Hignight, 1991) but the authors did not describe the causal fungus. Morphologically similar Claviceps anamorphs (sphacelia) with fusiform to lunate conidia of unknown species were also collected from Urochloa spp. and Eragrostis sp. in Zimbabwe during 2000-2001. In this paper we report the identity of these fungi. Only two species of Claviceps with markedly fusiform to lunate conidia have been formally described, both colonizing grasses from Paniceae: C. fusiformis and the Australian endemite C. hirtella. In addition, there are numerous records of sphacelial forms with lunate conidia from various grasses and locations in India (Table 1). C. fusiformis has been described as the main ergot pathogen of pearl millet or bajra (Pennisetum typhoideum, now P. glaucum) in Africa and India (Loveless, 95 1967; Thakur et al., 1984). The oldest specimen of C. fusiformis came from Ghana in 1925 (Loveless, 1967). In Africa, C. fusiformis was the typical ergot parasite of Pennisetum and Cenchrus (Loveless, 1964a, 1967). In India, the picture is more confusing. Pearl millet was not initially recorded as a host plant of ergot fungi with fusiform or lunate conidia, but such fungi were recorded, although not identified, on other panicoid grasses well before the first documented epiphytotics of C. fusiformis. The first record of an ergot with long conidia (20.7 × 6.1 µm), later identified as C. fusiformis (Siddiqui and Khan, 1973), refers to pearl millet in Banaskanatha district in Bombay State in 1955 (Shinde and Bhide, 1958). Shinde and Bhide (1958) also noted that the asci and conidia of the fungus from pearl millet resembled those observed by Thomas et al. (1945b) on Pennisetum hohenackeri. The following year, fully developed ergot infection at severe incidence was recorded in Bombay and Mysore states (Bhide and Hegde, 1957). Pearl millet ergot remained little known in other regions until 1966 when, with the introduction of the first generation of pearl millet hybrids (HB 1 and HB 2), epiphytotics developed in other states (Sundaram et al., 1969). In addition to Cenchrus ciliaris, Cenchrus setigerus and Pennisetum hohenackeri, the unidentified Indian ergot anamorphs originated from Panicum antidotale, Urochloa panicoides, Urochloa ramosa, and Paspalidium flavidum (now Setaria flavida) (Ramakrishnan, 1937; Thomas et al., 1945a,b; Adyanthaya, 1946; Ramakrishnan, 1947). Cross-infection experiments have shown that some of these fungi possessed polygeneric host ranges. Ramakrishnan (1947) infected Urochloa ramosa with conidia collected from Cenchrus setigerus, whereas Thakur and Kanwar (1978) succeeded in infecting pearl millet with an ergot pathogen from Panicum antidotale. The question of whether the pearl millet ergot epiphytotics in India were caused by a change in host of an indigenous population of C. fusiformis already present e.g. that from Cenchrus, or by introduction of an African C. fusiformis more virulent to Pennisetum spp. remained open. Similarities between the alkaloid composition of African and Indian isolates 96 do not preclude an introduction (Kumar and Arya, 1978). Langdon (1952) noted the similarity of conidial morphology and host specificity of Claviceps anamorphs with lunate conidia in the Indian records to those of Claviceps hirtella (Langdon, 1942). C. hirtella is the only Australian species colonizing Urochloa spp. Queensland Plant Pathology Herbarium harbours specimens of C. hirtella from species of Urochloa, Paspalidium, Eriochloa and Entolasia. Since 1976, C. hirtella has been repeatedly found on Cenchrus ciliaris (supposedly introduced in 1870-1880), presenting thus a problem for the seed industry. In former Rhodesia (now Zimbabwe), Loveless (1964a) defined 13 groups of Claviceps sphacelial anamorphs on the basis of conidial morphology and host species. Six of them were assigned to teleomorphs already described (C. paspali, C. digitariae, C. sulcata, C. maximensis, C. pusilla and C. cynodontis). Since then, three new species have been identified with the corresponding morphological groups: C. rhynchelytri, group No. 1 (Herd and Loveless, 1965); C. fusiformis, group No. 7 (Loveless, 1967); and C. africana, group No. 10 (Loveless, 1964b; Herd and Loveless, 1965; Frederickson et al., 1991). Two of the original groups (No. 11 from Hyparrhenia spp. and No. 13 from Loudetia spp.) are still present in Zimbabwe (Frederickson, 1990; Pažoutová and Frederickson, 2005). No records of anamorphs with lunate conidia on Eragrostis in Zimbabwe, and only one for Urochloa and Brachiaria in South Africa, 1988 (conidia measuring 13 µm × 4.4 µm and identity suggested to be Claviceps sulcata), were mentioned in these studies. Sclerotia of African and Indian isolates of C. fusiformis from pearl millet contain agroclavine and elymoclavine as the main alkaloid components; minor or trace components are chanoclavine, setoclavine, penniclavine and, occasionally, festuclavine. Both also readily produce clavines in submerged culture in vitro (Banks et al., 1974; Bhat et al., 1976; Singh and Husain, 1977; Kumar and Arya, 1978). However, Janardhanan et al. (1982) detected festuclavine as the major component, accompanied by agroclavine, and chanoclavine, in sclerotia of Claviceps sp. with lunate conidia Fungal Diversity Table 1. A survey of Paleotropic ergot fungi with fusiform to falcate conidia. Unnamed records: Urochloa ramosa Urochloa panicoides Urochloa distachya Setaria flavida Panicum antidotale Pennisetum orientale x P. purpureum Cenchrus setigerus Cenchrus ciliaris Described species: Claviceps fusiformis Claviceps fusiformis Claviceps hirtella Sclerotia Conidial shape Conidial size (µm) Citation unknown falcate, pointed fusiform to lunate 14.6 < 19.8 < 29.2 × 4.4 < 5.8 < 7.3 12.8 < 15.4 < 19 × 3.2 < 5.1 < 6.4 (Ramakrishnan, 1937) arcuate, slightly pointed lunate 12.6-19 × 3.8-6.3 (Thirumalachar, 1945) 12.8 < 16 < 20.8 × 4.8 < 5 < 6.4 (Ramakrishnan, 1947) 11.4 < 21 < 28.6 × 3.9 < 5.3 < 5.8 (Janardhanan et al., 1982) (Sundaram et al., 1969) 12.8 < 17.9 < 26 × 3.2 < 4.8 < 6.4 (Ramakrishnan, 1947) 14 < 18.4 < 25.2 × 5.0 < 5.96 < 8.4 (Adyanthaya, 1946) fusiform to lunate 9.5 < 15.8 < 22.5 × 3 < 3.6 < 5 (Loveless, 1967) fusiform 12 < 15.9 < 26.4 × 2.4 < 3.9 < 6 (Thakur et al., 1984) arcuate 11-16.5 × 4.5-6.5 (Langdon, 1942) globose to oblong, brown, 1.5 × 0.6 mm subglobose to spherical curved, brown 4-5 × 1 mm unknown unknown unknown unknown obpyriform, slightly protruding elongated to round, brown subglobose, yellowish brown fusiform fusiform fusiform to lunate lunate from Panicum antidotale. The same alkaloids, plus elymoclavine, were found in submerged culture isolates from this source, suggesting that the pathogen may not have been C. fusiformis. No information is available about alkaloids of C. hirtella. Crous and Groenewald (2005) stated that taxonomic names based solely on fungal phenotype often represent species complexes or cryptic species, and not operational units. The objective of this study was to assess the relationships of the existing species C. fusiformis and C. hirtella to the anamorphs with lunate conidia to determine if they represent such a species complex. Characterization of the isolates was based on rDNA and partial β-tubulin sequences, morphological markers and alkaloid biosynthesis. An erroneous rDNA sequence, AJ133392 for C. fusiformis, from a previous study (Pažoutová, 2001) was also corrected. Materials and methods Herbarium specimens and isolates The origin of specimens and isolates is given in Tables 2 and 3. Specimens of Clavi- (Ramakrishnan, 1947) ceps hirtella, BRIP 16635 and 13544, and monosporic isolates of C. hirtella were obtained from Queensland Plant Pathology Herbarium (BRIP, courtesy of Dr Roger Shivas). Strain C. fusiformis 47A was isolated in 1957 from Pennisetum material originating from the Ivory Coast (Tyler, 1958; Pažoutová and Tudzynski, 1999). Strain C. fusiformis F27 was obtained from the former Farmitalia collection from Italy in 1983, but it originated from an African Pennisetum specimen collected in the 1960’s. Pure cultures of Claviceps specimens from Africa and Texas were isolated by plating honeydew drops from florets of infected pearl millet onto T2 agar and subsequent transfer of agar plugs with germinating macroconidia (Pažoutová et al. 2002). Only C. fusiformis from India was purified from surface-sterilized sclerotia (Pažoutová et al., 2000). Herbarium Type specimens were deposited in the herbarium of the Mycological Department, National Museum in Prague (PRM). Ex-holotype and other representative strains were stored in liquid nitrogen and deposited in the Czech Collection of Clavicipitaceae (CCC; Institute 97 Table 2. Characteristics and origins of sphacelial specimens. Species Claviceps fusiformis Specimen Accession Host PRM 857332 Pennisetum glaucum PRM 857331 Pennisetum glaucum PRM 857333 Pennisetum glaucum Sphacelia eriochloae PRM 857335 Eriochloa sericea Sphacelia texensis PRM 857334 * Eriochloa sericea Cenchrus ciliaris PRM 857338 Cenchrus ciliaris * Pennisetum glaucum PRM 857336 Pennisetum glaucum PRM 857337 Pennisetum glaucum BRIP13544 Cenchrus ciliaris BRIP16635 Cenchrus ciliaris PRM 857342 Urochloa sp. * Urochloa trichopus PRM 857340 Urochloa mosambicensis PRM 857341 Urochloa oligotricha PRM 857339 Eragrostis sp. Claviceps hirtella Sphacelia lovelessii * Small specimens consumed by microscopy were not deposited SD – standard deviation 98 Location Shamva, Zimbabwe Matopos, Zimbabwe Gulbarga, Karnataka, India Agua Dulce, Texas Kingsville, TX Agua Dulce, Texas Kenedy county, Texas Weslaco, Texas Corpus Christi, Texas Weslaco, Texas Meandarra, Queensland, Australia Blackall, Queensland, Australia Matopos, Zimbabwe Matopos, Zimbabwe Matopos, Zimbabwe Matopos, Zimbabwe Matopos, Zimbabwe Year Collector Alkaloids per sphacelia (µg) 6 Conidial dimensions (µm) Length SD width SD 9.5 < 18.9 < 27.5 3.7 3.9 < 4.8 < 6.3 0.4 1999 N. W. Mc Laren 2000 D.E. Frederickson 16 10.2 < 17.9 < 21.6 2.0 2.8 < 4.8 < 8.5 0.5 2006 R.H. Angadi 30 15 < 20.9 < 26.5 2.7 3.2 < 4.9 < 6.5 0.8 1997 G. N. Odvody 24 6.7 < 10.6 < 14.9 1.5 2.6 < 3.6 < 4.6 0.4 2006 1997 G. N. Odvody G. N. Odvody 13 0 7.7 < 12.1 < 15.7 8.2 < 12.6 < 19.6 1.5 1.6 2.5 < 3.7 < 5.5 3.0 < 4.0 < 4.6 0.5 0.5 1997 O. Rodriguez 0 8.8 < 12.1 < 16.8 1.4 3.1 < 3.7 < 4.6 0.4 1998 G. N. Odvody 0 7.1 < 11.6 < 15.1 1.5 2.8 < 3.9 < 5.5 0.4 2003 G. N. Odvody 0 8.8 < 12.9 < 17.2 1.6 2.8 < 4.0 < 4.8 0.4 2003 G. N. Odvody 0 8.1 < 13.1 < 18.2 1.6 3.5 < 4.2 < 5.3 0.4 1982 V. French 0 9.6 < 14.1 < 24.8 1.9 3.4 < 4.3 < 5.4 0.4 0 9.1 < 12.5 < 16.6 1.5 2.2 < 3.5 < 4.5 0.4 1989 2000 D.E. Frederickson 0 15.6 < 19.9 < 25.2 2.2 2.9 < 5.2 < 5.7 0.5 2001 D.E. Frederickson 0 10.6 < 15.4 < 20.4 1.9 3.0 < 5.2 < 5.3 0.5 2001 D.E. Frederickson 0 10 < 14.3 < 20.7 1.7 3.6 < 4.5 < 6.1 0.6 2001 D.E. Frederickson 0 12.4 < 17.6 < 23.2 2.0 4.2 < 4.7 < 6.4 0.5 2001 D.E. Frederickson 0 12.5 < 18.1 < 24.8 2.1 4.1 < 5.6 < 6.3 0.9 Fungal Diversity Table 3. Origin and identity of isolates used for DNA sequencing and alkaloid production. Species Specimen of origin Isolate Host Location Year Collector Pennisetum sp. Pennisetum typhoides Pennisetum glaucum Africa French Central Africa Shamva, Zimbabwe 1960’s 1957 1999 N.W. Mc Laren Pennisetum glaucum Eriochloa sericea Gulbarga, Karnataka, India Kingsville, Texas 2005 2006 R.H. Angadi G. N. Odvody Pennisetum glaucum Pennisetum glaucum Weslaco, Texas Corpus Christi, Texas 2003 2003 G. N. Odvody G. N. Odvody EF052277 EF052278 EF052279 EF473873 EF473874 EF473878 Claviceps hirtella PRM 857332 " PRM 857333 PRM 857334 " " " PRM 857337 PRM 857336 " " " BRIP 43959 CCC 110 (F27) CCC 106 (47A) CCC 525 CCC 524 CCC 846 CCC 859 CCC863 CCC 868 CCC 872 CCC 774 CCC 776 CCC 778 CCC 777 CCC 858 CCC 786 Urochloa sp. Goondiwindi, Queensland, Australia 2004 R.G. Shivas, T.S. Marney EF052280 EF473872 Sphacelia lovelessii " " " " " " PRM 857340 CCC 792 CCC 787 CCC 788 CCC 789 CCC 790 CCC 791 CCC 642 PRM 857341 PRM 857339 CCC 646 CCC 647 CBS 125.63 Claviceps fusiformis Sphacelia eriochloae Sphacelia texensis Claviceps viridis Accession No. rDNA β-tubulin EF052275 EF473876 AJ133392 AM498382 AJ626727 EF473867 “ EF473866 EF052276 EF473877 EF473864 EF473875 EF473871 Urochloa mosambicensis Urochloa oligotricha Eragrostis sp. Oplismenus compositus Matopos, Zimbabwe 2001 D.E. Frederickson EF052282 EF473870 Matopos, Zimbabwe Matopos, Zimbabwe India 2001 2001 D.E. Frederickson D.E. Frederickson EF052281 AJ605996 AJ133404 EF473869 EF473868 EF473865 99 Fig. 1. Neighbor-joining tree obtained from phylogenetic analysis of a combination of rDNA and partial β-tubulin sequences with Claviceps viridis as an outgroup. Bootstrap confidence levels, based on 1000 replicates, are given on the appropriate branches. Bootstrap values for maximum parsimony tree (with the same topology) are given in brackets. Note that a recent C. fusiformis isolate from India (CCC 846) is more closely related to African isolates originating from around 1960 (CCC 106, 110) than to recent African isolates (CCC 524, 525). of Microbiology, Academy of Sciences of the Czech Republic, Prague). Media and cultivation Isolates were maintained on T2 agar slants by transfer every two months. For alkaloid production, seed cultures in sucroseasparagine medium (TI) were inoculated with 3 ml of conidial suspension from a slant culture and incubated for 10 days. Five ml of seed culture were transferred to fermentation culture (T2) and incubated for 20 - 29 days (Pažoutová et al., 1981) The cultivations proceeded on a rotary shaker at 24 oC in 250 ml Erlenmeyer flasks with 60 ml of the respective medium. Colony morphology was observed after 15 days of growth on T2 agar plate Microscopy Only honeydew specimens were used for observation and measurement of conidial size as conidia from cultures exhibit greater 100 variability in shape and size. Conidia were stained in 1% cotton blue in lactic acid and photographed and measured using an Olympus BX51 microscope equipped with a digital camera (CAMEDIA) and image-processing software (QuickPHOTO Camera 2.2.). At least 50 conidia from each sample were measured. Statistical analyses of spore size data were performed using Kyplot 2.0 beta 15 (Yoshioka, 2002). Alkaloid analysis Cultures were centrifuged and in the suitably diluted supernatant, the alkaloids were measured colorimetrically using Van Urk’s reagent (Pažoutová et al., 1981) with elymoclavine as a standard. For qualitative alkaloid analysis, the culture was centrifuged; supernatant was alkalized with NH4OH to pH 8-9 and twice extracted with the same volume of dichloromethane. The extracts were combined and dried using anhydrous Na2SO4, then Fungal Diversity evaporated to dryness and dissolved in 200 µl of methanol. Alkaloid content was analyzed by high performance liquid chromatography (HPLC) (Pažoutová et al., 2000). Thin layer chromatography (TLC) of alkaloids was performed on silica gel plates (Merck) exposed briefly to NH3 vapors, developed in chloroform:methanol (8:2), and detected using Ehrlich’s reagent spray. search of the initial tree with a search factor 3 and random addition of trees with 60 repetitions. The stability of clades was evaluated by a bootstrap test with 1000 replications. The dataset and analysis were archived in TreeBase under the submission ID number SN3321. DNA preparation and analyses DNA was purified from 4-7 days-old mycelium grown on T2 plates overlaid with cellophane using an UltraClean Microbial DNA Isolation Kit (Mo-Bio Laboratories, Solana Beach, California) according to the manufacturer’s manual. Nuclear rDNA containing internal transcribed spacers (ITS1 and ITS2), 5.8S and D1- D2 domains of the 28S region were amplified with primers ITS5 and NL4 (White et al., 1990). A region of the β-tubulin gene containing part of intron 1, introns 2, 3 and 4, exons 2, 3, 4 and the first 56 base pairs of exon 5 was obtained using primers T1 and T22 (O'Donnell and Cigelnik, 1997). The reaction conditions in a Mastercycler Gradient thermocycler (Eppendorf, Hamburg, Germany) were as follows: 1 cycle of 3 minutes at 95°C, 30 seconds at 55°C and 1 minute at 72°C; 30 cycles of 30 seconds at 95°C, 30 seconds at 55°C and 1 minute at 72°C; 1 cycle 30 seconds at 95°C, 30 seconds at 55°C and 10 minutes at 72°C. The reaction mixture consisted of PCR buffer, 1U of DynaZyme (both Finnzymes, Oy, Finland), 0.2 mM deoxynucleotides, 2 pmol of each primer, and 5-50 ng of DNA in 25 µl of total volume. Custom sequencing of DNA was done at Macrogen Inc. (Seoul, Korea). All infections sampled (Table 2) exhibitted the sphacelial stage of ergot development. Honeydew production (fresh drops or a dried crust on glumes) was the only sign of infection. Sphacelia were still hidden in the glumes and no mature sclerotia were found, except for the pearl millet specimen from India. Attempts to germinate these sclerotia into perithecial heads were unsuccessful. Phylogenetic analysis Two sequence alignments were constructed: the datasets for rDNA and β-tubulin. Sequences were aligned using MUSCLE (Edgar, 2004) and the alignments were edited with BioEdit (Hall, 1999). Phylogenetic analyses were performed using MEGA 3.1 (Kumar et al., 2004). A neighbor-joining tree was constructed using the Kimura-2 parameter model with complete deletion option. A maximum parsimony (MP) tree was computed using close neighbor interchange, mini-heuristic Results Phylogenetic results The dataset of rDNA sequences contained 1187 sites, yielding 22 variable positions with 12 singletons and only 10 sites informative for parsimony. The dataset of partial βtubulin sequences contained 1554 sites with 60 variable positions, 36 of which were informative for parsimony. All mutations in the coding regions of the β-tubulin gene were synonymous at the third codon position. Both datasets were combined and Claviceps viridis was added as the outgroup. The topology of the distance tree and consensus MP tree was identical (Fig. 1). Parsimony analysis found 15 equally parsimonious trees, which differed only in the mutual positions of C. fusiformis isolates from Africa and India. Two well-supported clades were found. The fusiformis-hirtella clade comprised four well-supported lineages with very small distances between sequences: the African and Indian C. fusiformis isolates from pearl millet, a Texas isolate from Eriochloa sericea, isolates from pearl millet from the same area and C. hirtella at the ancestral position. The second clade consisted of African isolates from Urochloa and Eragrostis. Alkaloid production Alkaloids in planta (Table 2) were only detected in pearl millet florets containing sphacelia and honeydew of C. fusiformis from 101 Fig. 2. Morphology of conidia from honeydew and colonies (reverse and obverse) on medium T2. a-c. Claviceps fusiformis. d-f. Sphacelia eriochloae. g-i. Sphacelia texensis. j-l. Claviceps hirtella. m-o. Sphacelia lovelessii. Bars = 20 µm. 102 Fungal Diversity Africa and India and in Eriochloa florets with S. eriochloae from Texas. All lineages defined by the phylogenetical analysis were also clearly distinguishable by the specific qualitative combination of alkaloids. In shaken cultures (Table 4), C. fusiformis isolates from Africa and India produced agroclavine and elymoclavine as major alkaloids, accompanied by chanoclavine. C. hirtella readily produced a whole range of clavines with agroclavine and festuclavine predominating, but ergometrine was also detected. Of the Texas isolates, the isolate from Eriochloa produced mainly agroclavine, elymoclavine and festuclavine and the cultures of isolates from pearl millet contained small amounts of agroclavine and festuclavine. African isolates from Urochloa and Eragrostis produced only traces of clavines on the detection threshold limit. Morphology Colony morphology was observed on T2 agar, specifically suited for Claviceps growth and secondary metabolite production (Fig. 2). On widely-used fungal media such as maltextract agar (MEA) and Czapek-Dox (CZD), culture growth is about half of that on T2 and the typical pigmentation and structure of the colonies is not expressed. Therefore, MEA and CZD are unsuitable for morphological characterization of ergot fungi. Colonies of C. fusiformis isolates from pearl millet (Africa, India) and those from Texas from Eriochloa and pearl millet were similar (Fig. 2) on T2 agar. Typically, all showed rapid growth (2-4.8 cm in 14d) with a diffuse and markedly radiating margin. Colony was mostly velutinous, consisting of highly sporulating conidiophores, giving a powdery appearance; raised with cerebriform wrinkles in the centre and plane toward the margin; obverse, off-white to grayish; reverse, similar or slightly brown, typically turning reddish brown with age in some strains of C. fusiformis (CCC 525, CCC 846); soluble pigments yellowish to reddish brown. African isolates from Urochloa and Eragrostis exhibited restricted growth ( < 1.5 cm) and absence of sporulation and pigment production. In these isolates, conidiation ability was lost during the first year of isolation, whereas cultures of all the other isolates retained conidiation ability, sometimes for decades. Conidia of C. fusiformis from Africa and India (Table 2, Fig. 2) were long (mean 18.9 µm), mostly straight and fusiform. Conidia from honeydew of Texas isolates from Eriochloa and pearl millet, and of C. hirtella, were smaller (12 µm) and more lunate. Conidia from African honeydew specimens from Urochloa and Eragrostis were the same length as C. fusiformis but slightly wider (5.2 µm; c.f. 4.8 µm C. fusiformis) and more lunate. Taxonomy All analyses reveal that the specimens from Africa and India, determined traditionally as C. fusiformis according to Loveless (1967), represent a taxonomically homogenous group with host specificity to Cenchrus and Pennisetum. Similarly, the Australian species C. hirtella has a distinct position. Specimens with fusiform conidia colonizing pearl millet and C. ciliaris in Texas were closely related to C. fusiformis, but showed distinct differential characters in DNA sequence, conidia size and alkaloid production; therefore this group is described here as new species. Because there was no teleomorph available (sphacelia were mostly infected in the course of development by Epicoccum andropogonis and did not mature) and the International Code of Botanical Nomenclature does not allow the application of a teleomorph name in the absence of the teleomorph (Greuter et al., 2000), the species will be described as Sphacelia texensis. Also a new name, Sphacelia eriochloae, is proposed here for the fungus from Eriochloa because of its unique combination of host specificity with alkaloid and DNA characteristics. African isolates from Urochloa and Eragrostis differed from all other specimens in all aspects and the new name S. lovelessii is designated here for this new species. Claviceps fusiformis Loveless, Transactions of the British Mycological Society 50: 17. (1967) (Fig. 2a,b,c ) Cultural characteristics: colonies on T2 medium, (14d, 24°C) 21-48 mm in diameter; diffuse and markedly radiating margin, colony mostly velutinous, consisting of highly sporulating conidiophores giving a powdery appearance, 103 Table 4 Alkaloid yield of Claviceps/Sphacelia isolates in shaken culture. Claviceps fusiformis Sphacelia eriochloae Sphacelia texensis Isolate CCC No. 524 525 106 110 846 859 863 868 872 774 776 777 Cultivation 21 21 21 21 21 21 21 21 21 29 29 29 (days)*) Total alkaloids 1520 1050 1265 510 800 1280 7930 3780 5210 65 91 111 (mg.L-1) Constituent alkaloids (%): Ergometrine Elymoclavine 21 26.7 13.2 68 10.5 22.1 21.4 24.3 22.3 Chanoclavine 1.3 2.9 1.4 24 0.8 2 6.5 2 6.1 Chanoclavine-1- 0.7 2.9 aldehyde DH-setoclavine iso-DHsetoclavine Agroclavine 75.7 68.5 80.3 8 70.6 41.4 34.4 30.1 40.8 17.1 17.5 34.7 Festuclavine 12.4 9.3 8.1 9.1 82.9 35.3 46.5 Pyroclavine 1.9 1.4 10.4 1.1 Lysergol, DH4.2 lysergol unknown 1.31 1.9 2.1 14 20 27 25 20 0 47.2 18.8 clavines *) Cultivations were terminated when viscosity due to glucan production rendered aeration ineffective ND - not determined, alkaloid content too low to permit a qualitative analysis 104 778 29 858 29 98 117 8.4 27.1 55.2 44.8 64.5 0 Claviceps hirtella 786 787 788 27 24 24 789 29 790 24 791 24 792 27 Sphacelia lovelessii 645 646 649 29 29 29 247 2 971 1354 143 3 586 326 219 37 6.1 12.7 1.9 2.6 4.5 1.9 3.3 4.7 1.6 3.4 6.6 3 3.2 6.6 2.5 8.7 7.9 3.3 7.2 3.6 2.6 100 6.7 1 5.7 1.3 3.1 1.1 6.7 1.2 4.9 1.1 5.9 1.2 4.2 1.2 24 42.3 4.2 1 19.2 57.9 5.7 1.3 24.2 59.6 1.3 1.1 26.1 48.7 3 1.2 13.2 65 2.4 1.1 5.8 63.5 2.5 1.2 12.5 66.4 1.2 1.2 36 31 ND 100 ND Fungal Diversity raised with typically cerebriform wrinkles in the centre and plane toward the margin, obverse off-white to grayish, reverse similar or light brown getting typically reddish brown in age in some strains, soluble diffuse pigment yellowish to reddish brown to vinaceous. Macroconidia: fusiform, straight, rarely lunate (10-28 × 3-9 µm, mean 19 × 5 µm). Teleomorph: not examined, see Loveless (1967) and Thakur et al.(1984). Alkaloids: production in the culture: spontaneous, high (500-2000 mg/L); main alkaloids: chanoclavine, agroclavine, elymoclavine. Habitat: in living florets of Pennisetum and Cenchrus spp. Known distribution: Africa, India. Material examined: ZIMBABWE, Matopos, from floret of Pennisetum glaucum 2001, D.E. Frederickson (PRM 857331); Shamwa, from floret of Pennisetum glaucum 2001, N. McLaren, (PRM 857332) (cultures CCC524, 525); INDIA, Karnataka, Gulbarga, from floret of Pennisetum glaucum 2005, R.H. Angadi, (PRM 857333) (culture CCC846). Notes: Our concept of C. fusiformis agrees with definition of Loveless (1967) and Thakur et al.(1984). The species infects typically species of Cenchrus and Pennisetum. Long fusiform conidia which are rarely lunate are also distinct. Claviceps hirtella Langdon, Proceedings of the Royal Society of Queensland 54: 27. (1942) (Fig. 2j, k, l) Cultural characteristics: colonies on T2 medium (14d, 24°C) from 25 mm to 38 mm in diameter, obverse off-white, reverse with shades of rose-grey. Two types of growth occurred even in colonies of the same isolate. Colonies growing more rapidly were more diffuse on margin, markedly radiating, velutinous and plane. Colonies with slower growth consisted of dense mycelial mat. Sporulation was absent in both cases. Macroconidia: lunate to fusiform (9-25 × 2-5.5 µm, mean 13 × 4 µm). Teleomorph: not examined, see Langdon (1942) Alkaloids: production in the culture: spontaneous, high (200-2500 mg/L); main alkaloids: ergometrine, chanoclavine, setoclavine, elymoclavine, agroclavine, festuclavine. Habitat: in living florets of Urochloa, Cenchrus, Paspalidium, Eriochloa, and Entolasia. Known distribution: Australia. Material examined: AUSTRALIA, Qld, Meandarra, in florets of Cenchrus ciliaris, 1982 (BRIP13544); AUSTRALIA, Qld, Blackall, in florets of Cenchrus ciliaris, 1989 (BRIP16635); AUSTRALIA, Qld, Goondiwindi, in florets of Urochloa sp., 2004 (BRIP 43959) (cultures CCC786-CCC792). Notes: The typical differential character of this species is the production of ergometrine. Sphacelia eriochloae Pažoutová & Odvody sp. nov. (Fig. 2d, e, f) Etymology: Referring to the host plant name. Hab. in ovariis et in inflorescentiis Eriochloae sericeis, Texas Species similis Claviceps fusiformis sed differt per suam combinationem characterum; macroconidiis hyaliniis, fusoideis vel lunatis (6.7-15.7 × 2.5-5.5 µm, mediet. 11.6 × 3.6 µm); regio ‘rDNA ITS’, ‘rDNA28S cum polymorphismis unicis sequentiae (GenBank EF473864). Teleomorphosis ignota. Cultural characteristics: colonies on T2 medium (14d, 24°C) 21-26 mm in diameter, like C. fusiformis. Macroconidia: lunate to fusiform (6.715.7 × 2.5-5.5 µm, mean. 11.6 × 3.6 µm) Teleomorph: not observed. Alkaloids: production in culture: spontaneous, very high (2000-8000 mg/L); main alkaloids: chanoclavine, festuclavine, elymoclavine, agroclavine. Habitat: in living florets of Eriochloa sericea Known distribution: Texas, Mexico. Material examined: TEXAS, Agua Dulce, in florets of Eriochloa sericea, 1997, G.N. Odvody (PRM 857335); TEXAS, Kingsville, in florets of Eriochloa sericea, 2006, G.N. Odvody (HOLOTYPE, PRM 857334) (ex type culture CCC859), Mycobank Accession No.: MB 510710 Molecular characters: sequence of the ITS and D1D2 regions of the 28S rDNA unique (EF473864). The sequence is given in the format [three-prime 18S]ITS1*5.8S*ITS2 [five-prime 28S]. [ATCATTA]CCGAGTTTTCAACTCCCAAAC CCACTGTGAACCTATACCAAAAACGTTGCCTCG GCGGGACATGCGCCCCGGACCGCCCCCCCCCCT CGCGGGGGAGGGCGCCGGATCCCACGGCCGCCC GCCGGGGGCCCCAAACTCTGTATTCCCATAGCG GCATGTCTGAGTGGATTTATCCAATAAATCA*AA ACTTTCAACAACGGATCTCTTGGTTCTGGCATCG ATGAAGAACGCAGCGAAATGCGATAAGTAATGT GAATTGCAGAATTCAGTGAATCATCGAATCTTT GAACGCACATTGCGCCCGCCAGTACTCTGGCGG 105 GCATGCCTGTTCGAGCGTCATT*TCAACCCTCAG GCCCCCGGGCCTGGTGTTGGGGACCGGCTCACG GGGGGGAGGCACAGCGCCCCCCCCCCCTGCCGC CCCCTAAATGGATCGGCGGCCACGCCGCGGCCT CCCCTGCGCAGTAACATACCACCTCGCAGGCGG CTGGCTCGGCGCGGCCACTGCCGTAAAACGCCC AACTTCTCCAGAG[TTGACCTCGAATCAGGTAGG AATACCCGCTGAACTTAAGCATATCAATAAGCG GAGGAAAAGAAACCAACAGGGATTGCCCCAGT AACGGCGAGTGAAGCGGCAACAGCTCAAATTTG AAATCTGGCCCCCCGGGGCCCGAGTTGTAATTT GCAGAGGATGCTTTTGGCGAGGCGCCTTCCGAG TTCCCTGGAACGGGACGCCATAGAGGGTGAGAG CCCCGTCTGGTCGGACGCCGAGCCTCTGTAAAG CTCCTTCGACGAGTCGAGTAGTTTGGGAATGCT GCTCTAAATGGGAGGTATATGTCTTCTAAAGCT AAATACCGGCCAGAGACCGATAGCGCACAAGT AGAGTGATCGAAAGATGAAAAGCACTTTGAAA AGAGGGTTAAACAGTACGTGAAATTGTTGAAAG GGAAGCGCCTGTGACCAGACTTGCGCCCGTCGG ATCACCCAGCGTTCTCGCTGGTGCACTCCGGCG GGCGCAGGCCAGCATCAGCTCGTCTCGGGGGAC AAAGGCGGCGGGAACGTGGCTCCTCCGGGAGTG TTATAGCCCGCCGTGCAATGCCCTGGGGCGGGC TGAGGACCGCGCGTAAGCATGGATGCTGGCGTA ATGGTCATCAGCGACCCGTCTTGAAACACGGAC CAA]. Notes: The species has smaller conidia than C. fusiformis, C. hirtella or S. lovelessii. The only ergot fungus with similar conidia occurring in Texas is Sphacelia texensis, from which it differs in host, DNA sequences, and high overall production of alkaloids. Sphacelia texensis Pažoutová & Odvody sp. nov. (Fig. 2g, h, i) Etymology: Named after the geographical origin Hab. in ovariis et in inflorescentiis Penniseti glauci et Cenchri ciliari, Texas. Claviceps fusiformis var. texensis differt per suam combinationem characterum; macroconidiis hyaliniis, fusoideis vel lunatis (7-19 × 2.8-5.5 µm, mediet.12.5 × 4 µm); regio ‘rDNA ITS’, ‘rDNA28S cum polymorphismis unicis sequentiae (GenBank EF052277, EF052278 et EF052279). Teleomorphosis ignota. Cultural characteristics: colonies on T2 medium (14d, 24°C) from 35 mm to 53 mm in diameter, margin diffuse and markedly radiating, colony granular, velutinous, consisting of numerous conidiophores, raised with cerebriform wrinkles in the centre and plane toward the margin, obverse off-white, reverse similar or with shades of brown, soluble pigment slightly yellow. Macroconidia: lunate to fusiform (7-19 × 2.8-5.5 µm, mean 12.5 × 4 µm). Teleomorph: Claviceps sp. (not seen, based on phylogenetic inferences). 106 Alkaloids: production in the culture: spontaneous, low (≤ 100 mg/L); main alkaloids: agroclavine, festuclavine. Habitat: in living florets of Pennisetum glaucum and Cenchrus ciliaris. Known distribution: Texas. Material examined: TEXAS, Agua Dulce, in florets of Cenchrus ciliaris, 1997; TEXAS, Kenedy county, in florets of Cenchrus ciliaris, 1997 (PRM 857338); TEXAS, Weslaco, in florets of Pennisetum glaucum, 1998, G.N. Odvody; TEXAS, Weslaco, in florets of Pennisetum glaucum, 2003, G.N. Odvody (PRM 857337); TEXAS, Corpus Christi, in florets of Pennisetum glaucum, 2003, G.N. Odvody (HOLOTYPE, PRM 857336) (culture CCC858), Mycobank Accession No.: MB 510709 Molecular characters: sequences of the ITS and D1D2 regions of the 28S rDNA unique (EF052277, EF052278 et EF052279). The consensus sequence is given in the format [three-prime 18S]ITS1*5.8S*ITS2 [five-prime 28S]. [ATCATTA]CCGAGTTTTCAACTCCCAAAC CCACTGTGAACCTATACCAAAAACGTTGCCTCG GCGGGACATGCGCCCCGGACCGCCCCCCCCTCG CGGGGGAGGGCGCCGGATCCCACGGCCGCCCGC CGGGGGCCCCAAACTCTGTATTCCCATAGCGGC ATGTCTGAGTGGATTTATCCAATGAATCA*AAAC TTTCAACAACGGATCTCTTGGTTCTGGCATCGAT GAAGAACGCAGCGAAATGCGATAAGTAATGTG AATTGCAGAATTCAGTGAATCATCGAATCTTTG AACGCACATTGCGCCCGCCAGTACTCTGGCGGG CATGCCTGTTCGAGCGTCATT*TCAACCCTCAGG CCCCCGGGCCTGGTGTTGGGGACCGGCTCACGG GGGGGAGGCACAGCGCCCTCCCCCCCTGCCGCC CCCTAAATGGATCGGCGGCCACGCCGCGGCCTC CCCTGCGCAGTAACATACCACCTCGCAGGCGGC TGGCTCGGCGCGGCCACTGCCGTAAAACGCCCA ACTTCTCCAGAG[TTGACCTCGAATCAGGTAGGA ATACCCGCTGAACTTAAGCATATCAATAAGCGG AGGAAAAGAAACCAACAGGGATTGCCCCAGTA ACGGCGAGTGAAGCGGCAACAGCTCAAATTTGA AATCTGGCCCCCCGGGGCCCGAGTTGTAATTTG CAGAGGATGCTTTTGGCGAGGCGCCTTCCGAGT TCCCTGGAACGGGACGCCATAGAGGGTGAGAGC CCCGTCTGGTCGGACGCCGAGCCTCTGTAAAGC TCCTTCGACGAGTCGAGTAGTTTGGGAATGCTG CTCTAAATGGGAGGTATATGTCTTCTAAAGCTA AATACCGGCCAGAGACCGATAGCGCACAAGTA GAGTGATCGAAAGATGAAAAGCACTTTGAAAA GAGGGTTAAACAGTACGTGAAATTGTTGAAAGG GAAGCGCCTGTGACCAGACTTGCGCCCGTCGGA TCACCCAGCGTTCTCGCTGGTGCACTCCGGCGG GCGCAGGCCAGCATCAGCTCGTCTCGGGGGACA AAGGCGGCGGGAACGTGGCTCCTCCGGGAGTGT TATAGCCCGCCGCGCAATGCCCTGGGGCGGGCT GAGGACCGCGCGTAAGCATGGATGCTGGCGTAA TGGTCATCAGCGACCCGTCTTGAAACACGGACC AA] Fungal Diversity Notes: Sphacelia texensis differs from Claviceps fusiformis and C. hirtella by combination of characters: smaller conidial dimensions and low alkaloid production in vitro and in planta. Sequence of the ITS, D1D2 regions of the 28S rDNA (EF052277, EF052278 and EF052279) and β–tubulin are unique and identical (EF473873, EF473874 and EF473878); next nearest known relatives being S. eriochloae and C. hirtella. Sphacelia lovelessii Pažoutová, M. Kola ík & Freder. sp. nov. (Fig. 2m, n, o) Etymology: Named after A.R. Loveless (British mycologist). Hab. in ovariis et in inflorescentiis Urochloae spp. et Eragrostidis sp., Zimbabwe. Sphacelia lovelessii differt ab aliis speciebus per suam combinationem characterum; macroconidiis hyaliniis, lunatis vel fusioideis (10-25 × 3-6.4 µm, mediet.17 × 5 µm); regio ‘rDNA ITS’, ‘rDNA28S cum polymorphismis unicis sequentiae (EF052282, EF052281, AJ605996). Teleomorphosis ignota. Cultural characteristics: colonies on T2 medium (14d, 24°C) from 15mm (CCC 642, 646) to 27 mm (CCC 647) in diameter, margin narrow or lobate, colony plane (CCC 647) or raised centrally and wrinkled (CCC 642, 646) consisting of dense myceliar mat, sporulation absent, obverse white, reverse off-white to slightly brown in the centre similar or with shades of brown, soluble pigment absent. Macroconidia: lunate (10-25 × 3-6.4 µm, mean 17 × 5 µm). Teleomorph: Claviceps sp. (based on phylogenetic inferences but unknown) Alkaloids: traces of chanoclavine or agroclavine. Habitat: in living florets of Urochloa spp.,and Eragrostis sp. Known distribution: Zimbabwe. Material examined: ZIMBABWE, Matopos, in florets of Urochloa sp., 2000, D.E. Frederickson; ZIMBABWE, Matopos (PRM 857342); in florets of Urochloa mosambicensis, 2001, D.E. Frederickson (PRM 857340) (culture CCC642); ZIMBABWE, Matopos, in florets of Urochloa oligotricha, 2001, D.E. Frederickson (culture CCC646); ZIMBABWE, Matopos, in florets of Urochloa trichopus, 2001, D.E. Frederickson; ZIMBABWE, Matopos, in florets of Eragrostis sp., 2001, D.E. Frederickson (HOLOTYPE, PRM 857340) (culture CCC648). Mycobank Accession No.: MB 510615 Molecular characters: sequences of the ITS and D1D2 regions of the 28S rDNA unique (EF052282, EF052281, AJ605996). The consensus sequence is given in the format [three-prime 18S]ITS1*5.8S*ITS2[five-prime 28S]. Nucleotides in bold indicate differences among the sequences: underlined – transitions/transversions; bracketed – not present in all sequences. [ATCATTA]CCGAGTTTTCAACTCCCAAAC CCACTGTGAACCCGTACCAAAAACGTTGCCTCG GCGGGAGATGCGCCCCGGACCGCCCCCCCCC(C C)TCGCGGGGGAGGGCGCCGGATCCCACGGCCG CCCGCCGGGGGCCCCAAACTCTGTATTCCCATA GCGGCATGTCTGAGTGGATTTATCCAATGAATC A*AAACTTTCAACAACGGATCTCTTGGTTCTGGC ATCGATGAAGAACGCAGCGAAATGCGATAAGT AATGTGAATTGCAGAATTCAGTGAATCATCGAA TCTTTGAACGCACATTGCGCCCGCCAGTACTCTG GCGGGCATGCCTGTTCGAGCGTCATT*TCAACCC TCAGGCCCCCGGGCCTGGTGTTGGGGACCGGCT CACGGGGGG(G)R(CA)ACAGCGCCCSCCCTGCCG CCCCCTAAATGGATCGGCGGCCACGCCGCGGCC TCCC(C)TGCGCAGTAACATACCACCTCGCAGGC GGCTGGCTCGGCGCGGCCACTGCCGYAAAACGC CCAACTTCTCAAGAG[TTGACCTCGAATCAGGTA GGAATACCCGCTGAACTTAAGCATATCAATAAG CGGAGGAAAAGAAACCAACAGGGATTGCCCCA GTAACGGCGAGTGAAGCGGCAACAGCTCAAATT TGAAATCTGGCCCCCCGGGGCCCGAGTTGTAAT TTGCAGAGGATGCTTTTGGCGAGGCGCCTTCCSA GTTCCCTGGAACGGGACGCCATAGAGGGTGAGA GCCCCGTCTGGTCGGACGCCGAGCCTCTGTAAA GCTCCTTCGACGAGTCGAGTAGTTTGGGAATGC TGCTCTAAATGGGAGGTATATGTCTTCTAAAGCT AAATACCGGCCAGAGACCGATAGCGCACAAGT AGAGTGATCGAAAGATGAAAAGCACTTTGAAA AGAGGGTTAAACAGTACGTGAAATTGTTGAAAG GGAAGCGCCTGTGACCAGACTTGCGCCCGCCGG ATCACCCAGCGTTCTCGCTGGTGCACTCCGGCG GGCACAGGCCAGCATCAGCTCGTCTCGGGGGAC AAAGGCGGCGGGAACGTGGCTCCTCCGGGAGTG TTATAGCCCGCCGTGCAATGCCCTGGGGCGGGC TGAGGACCGCGCGTACGCATGGATGCTGGCGTA ATGGTCATCAGCGACCCGTCTTGAAACACGGAC CAA]. Notes: Sphacelia lovelessii differs from Claviceps fusiformis by markedly lunate conidia. It differs from C. fusiformis, C. hirtella, S. eriochloae and S. texensis by absence of alkaloids in vitro and in planta and by lack of conidiation and slow and compact growth on agar medium T2. Discussion DNA and alkaloid analyses now confirm the hypothesis of an introduction of C. fusiformis to India from Africa, since the Indian isolate was very similar to African isolates, especially to those collected around 1960. Whereas C. fusiformis, appearing as a 107 clade terminal on the phylogram, is a known parasite of Cenchrus and Pennisetum spp., ancestral C. hirtella was collected from a broader range of native hosts. The first record of C. hirtella on Cenchrus ciliaris is from 1976 (BRIP 11355), although this grass was introduced to Australia around 1870. The occurrence of diverging populations from Pennisetum, Cenchrus and Eriochloa show that, in common with species complexes of other phytopathogenic fungi (Crous and Groenewald, 2005), evolution has been towards narrower host preferences. In southern India, honeydew specimens containing lunate conidia were observed on various host grasses many years before C. fusiformis was introduced. At that time, neither large-scale epiphytotics of disease on pearl millet nor alkaloid toxicoses of grazing animals were recorded. Langdon (1952) suspected that the Indian records prior to 1950 might refer to C. hirtella or a closely related fungus. However, the descriptions of conidia on the Indian grass specimens as long (Table 1) are more consistent with S. lovelessii, which also does not produce any alkaloids. Unfortunately, no Indian isolates or specimens from hosts other than pearl millet are available for comparison with the other fungi of the C. fusiformis species complex. Members of the C. fusiformis species complex also differ in their alkaloid biosynthesis gene cluster. In the ”standard” biosynthetic pathway, elymoclavine (C-17 hydroxyagroclavine) is an end product of the clavine pathway that may be further oxidized to Dlysergic acid and its amides e.g. ergometrine, containing a simple amino alcohol as the amide component (for review see Flieger et al., 1997). In C. fusiformis, Lorenz et al. (2007) found that the oxidation of clavines to lysergic acid is inhibited because the respective clavine oxidase gene, cloA, is inactive. Therefore, it seems that ergometrine-producing C. hirtella still possesses a functional gene that is missing in all other members of the fusiformis-hirtella clade. Due to only trace amounts of alkaloids being produced by S. lovelessii (Table 4), production of ergometrine in this species could be neither proved nor disproved. From the whole species complex, only C. fusiformis and S. eriochloae have the potential 108 to cause human and/or animal toxicoses as they are able to produce alkaloids in planta even in the sphacelial stage of development. The origin of ergot infections on pearl millet, Cenchrus, and Eriochloa in Mexico and Texas (San Martín et al., 1997; VelásquezValle et al., 1998) is rather unclear. In the survey by Alderman et al. (2004) no records for grasses with lunate or fusiform conidial infections were mentioned. The presented results clearly preclude C. fusiformis from Africa or India as the causative pathogen. The close genetic similarity of S. texensis to Paleotropic C. fusiformis and C. hirtella, however, suggests that S. texensis may be yet another introduction from the Paleotropics (Africa and Southeast Asia) to the Americas, possibly with grass seed. The presence of S. lovelessii in Zimbabwe represents another puzzle. Differences in conidial shape and size as well as in DNA markers among the isolates and specimens suggest considerable infraspecific variability. It is also the only Claviceps/Sphacelia species with markedly lunate macroconidia to be recorded in Africa – C. fusiformis tends to straight, fusiform conidia. The older studies of herbarium specimens and collections from southern and eastern Africa (Doidge, 1950; Langdon, 1952; Loveless, 1964a, b; Herd and Loveless, 1965; Loveless, 1965) were extensive and it is surprising that, at least until collections stopped in the late 1960’s, infections with the fusiform to lunate conidial shape occurring on a grass genus as common and widely distributed as Urochloa were overlooked. However, only Claviceps sulcata with allantoid conidia was reported (Loveless, 1964a; Loveless and Herd, 1964). Therefore, it may be that infection with S. lovelessii is rather rare. In contrast, the signs of ergot infection on the tiny florets of Eragrostis, another widely distributed grass genus, were very inconspicuous and may easily have gone unnoticed. A polygeneric host range has been documented not only in S. lovelessii (although here it is even switching between host tribes), but also in C. sorghi (Sorghum – Heteropogon; Pažoutová et al., 2002) and C. africana (Sorghum – Hyparrhenia; Pažoutová and Frederickson, 2005) which contradicts the original Fungal Diversity paradigm of “one Claviceps species – one host genus” or at least “a group of closely related host genera”. From the evolutionary point of view, the less rigid host specificity may enable the parasite to colonize alternative hosts after migration to another region as a first step towards adaptive radiation. Acknowledgements This work was supported by the Czech Institutional Research Concept No. AV0Z5020903 and by the International Sorghum/Millet Collaborative Research Program (INTSORMIL). References Adyanthaya, N.R. (1946). A note on the occurrence of Sphacelia on Cenchrus ciliaris. Current Science 15: 286-287. Alderman, S.C., Halse, R.R. and White, J.F. (2004). A reevaluation of the host range and geographical distribution of Claviceps species in the United States. Plant Disease 88: 63-81. Banks, G.T., Mantle, P.G. and Szczyrbak, C.A. (1974). Large-scale production of clavine alkaloids by Claviceps fusiformis. Journal of General Microbiology 82: 345-361. Bhat, R.V., Roy, D.N. and Tulpule, P.G. (1976). Nature of alkaloids of ergoty pearl millet or bajra and its comparison with alkaloids of ergoty rye and ergoty wheat. Toxicology and Applied Pharmacology 36: 11-17. Bhide, V.P. and Hegde, R.K. (1957). Ergot on bajri [Pennisetum typhoides (Burm.) Stapf and Hubbard] in Bombay state. Current Science 26: 116. Craig, J. and Hignight, K.W. (1991). Control of ergot in buffelgrass with triadimefon. Plant Disease 75: 627-629. Crous, P.W. and Groenewald, J.Z. (2005). Hosts, species and genotypes: opinions versus data. Presented as a Keynote Address at the 15th Biennial Conference of the Australasian Plant Pathology Society, Geelong, Australia, 26-29 September 2005. Australasian Plant Pathology 34: 463-470. Doidge, E.M. (1950). The South African fungi and lichens to the end of 1945. Bothalia 5: 1-1094. Edgar, R.C. (2004). MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research 32: 1792-1797. Flieger, M., Wurst, M. and Shelby, R. (1997). Ergot Alkaloids - Sources, Structures and Analytical Methods. Folia Microbiologica 42: 3-30. Frederickson, D.E. (1990). Ergot disease of Sorghum. PhD thesis, University of London, London. Frederickson, D.E., Mantle, P.G. and De Milliano, W.A.J. (1991). Claviceps africana sp. nov.; the distinctive ergot pathogen of sorghum in Africa. Mycological Research 95: 1101-1107. Greuter, W., McNeill, J., Barre, F.R., Burdet, H.M., Demoulin, V., Filgueiras, T.S., Nicolson, D.H., Silva, P.C., Skog, J.E., Trehane, P., Turland, N.J. and Hawksworth, D.L. (2000). International Code of Botanical Nomenclature (St Louis Code) Regnum Vegetabile138. Koeltz Scientific Books, Königstein. Hall, T.A. (1999). BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symposium Series 41: 95-98. Herd, G. and Loveless, A. (1965). Studies of Rhodesian ergots. III. Claviceps rhynchelytri sp. nov. and Claviceps digitariae Hansford. . Kirkia 5: 21-24. Janardhanan, K.K., Singh, H.N. and Husain, A. (1982). Studies on Claviceps parasitic on Panicum species in India. Folia Microbiologica 27: 121125. Kumar, A. and Arya, H.C. (1978). Estimation and identification of alkaloids produced by Claviceps fusiformis Loveless on some varieties of pearl millet. Current Science 47: 633-635. Kumar, S., Tamura, K. and Nei, M. (2004). MEGA3: Integrated software for molecular evolutionary genetics analysis and sequence alignment. Briefings in Bioinformatics 5: 150-163. Langdon, R.F.N. (1942). Ergots on native grasses in Queensland. Proceedings of the Royal Society Queensland 54: 23-32. Langdon, R.F.N. (1952). Studies on ergot. PhD. thesis, University of Queensland, Brisbane, Australia. Lorenz, N., Wilson, E.V., Machado, C., Schardl, C.L. and Tudzynski, P. (2007). Comparison of ergot alkaloid biosynthesis gene clusters in Claviceps species indicates loss of late pathway steps in evolution of C. fusiformis. Applied and Environmental Microbiology 73: 7185-7191. Loveless, A. (1967). Claviceps fusiformis sp. nov., the causal agent of agalactia of sows. Transactions of the British Mycological Society 50: 13-18. Loveless, A.R. (1964a). Use of the honeydew state in the identification of ergot species. Transactions of the British Mycological Society 47: 205-213. Loveless, A.R. (1964b). Studies of Rhodesian ergots 1. Claviceps maximensis Theis and Claviceps pusilla Cesati. Kirkia 4: 35-40. Loveless, A.R. and Herd, G.W. (1964). Studies of Rhodesian ergots. II. A redescription of Claviceps sulcata Langdon. Kirkia 4: 41-43. Loveless, A.R. (1965). Studies on Rhodesian ergots 4. Claviceps cynodontis Langdon. Kirkia 5: 25-29. O'Donnell, K. and Cigelnik, E. (1997). Two divergent intragenomic rDNA ITS2 types within a monophyletic lineage of the fungus Fusarium are nonorthologous. Molecular Phylogenetics and Evolution 7: 103-116. Pažoutová, S., Taisinger, J., Flieger, M., Sajdl, P., Bass, A. and eháček, Z. (1981). The relationship between intensity of oxidative metabolism and predominance of agroclavine or elymoclavine in submerged Claviceps purpurea cultures. Journal of Natural Products 44: 225-235. 109 Pažoutová, S. and Tudzynski, P. (1999). Claviceps sp. PRL 1980 (ATCC 26245), 59 and Pepty 695/ch-I: their true story. Mycological Research 103: 10441048. Pažoutová, S., Olšovská, J., Linka, M., Kolínská, R. and Flieger, M. (2000). Chemoraces and habitat specialization of Claviceps purpurea populations. Applied and Environmental Microbiology 66: 5419-5425. Pažoutová, S. (2001). The phylogeny and evolution of the genus Claviceps. Mycological Research 105: 275-283. Pažoutová, S. and Frederickson, D.E. (2005). Genetic diversity of Claviceps africana on sorghum and Hyparrhenia. Plant Pathology 54: 749-763. Ramakrishnan, T.S. (1937). Some observations on Sphacelia spp. occurring in Coimbatore. The Madras Agricultural Journal 25: 119-121. Ramakrishnan, T.S. (1947). The natural occurrence of ergot in south India - III. Proceedings of the Indian Academy of Sciences - B 26: 136-141. San Martín, F., Lavín, P., García, A. and García, G. (1997). Estados anamórficos de Claviceps africana y Claviceps fusiformis (Ascomycetes, Clavicipitaceae) asociados a diferentes pastos en Tamaulipás, México. Revista Mexicana de Micología 13: 52-57. Shinde, P.A. and Bhide, V.P. (1958). Ergot of bajri (Pennisetum typhoides) in Bombay state. Current Science 27: 499-500. Siddiqui, M.R. and Khan, I.D. (1973). Renaming Claviceps microcephala ergot fungus on Pennisetum typhoides in India as Claviceps fusiformis. Transactions of the Mycological Society of Japan 144: 195-198. Singh, H.N. and Husain, A. (1977). Saprophytic production of ergot alkaloids by bajra ergot (Claviceps fusiformis Loveless). Indian Journal of Experimental Biology 15: 585-586. Sundaram, N.V., Bhowmik, T.P. and Khan, I.D. (1969). A new host for Claviceps microcephala (Wallr.) Tul. . Indian Journal of Agricultural Sciences 39: 703-705. 110 Thakur, D.P. and Kanwar, Z.S. (1978). Ability of naturally incident Claviceps microcephala from Panicum antidotale to produce ergot symptoms in Pennisetum typhoides. Indian Journal of Agricultural Sciences 48: 540-542. Thakur, R.P., Rao, V.P. and Williams, R.J. (1984). The morphology and disease cycle of ergot caused by Claviceps fusiformis in pearl millet. Phytopathology 74: 201-205. Thirumalachar, M.J. (1945). Ergot and sphacelial stages on some wild grasses in Mysore. Current Science 14: 22. Thomas, K.M., Ramakrishnan, T.S. and Srinivasan, K.V. (1945a). The natural occurrence of ergot in south India - II. Proceedings of the Indian Academy of Sciences - B 22: 191-192. Thomas, K.M., Ramakrishnan, T.S. and Srinivasan, K.V. (1945b). The natural occurrence of ergot in south India. Proceedings of the Indian Academy of Sciences - B 21: 93-100. Tyler, V.E., Jr. (1958). Some factors influencing saprophytic production of clavine alkaloids by Claviceps purpurea. Journal of American Pharmaceutical Association Sri. Ed. 47: 787-792. Velásquez-Valle, R., San Martín, F., Odvody, G.N., Isakeit, T., Narro-Sánchez, J. and Mora-Nolasco, R. (1998). Reporte preliminar sobre especies del género Claviceps asociadas con pastos en algunos estados de México y Texas, EUA. Revista Mexicana de Fitopatología 16: 42-45. White, T.J., Bruns, T., Lee, S. and Taylor, J. (1990). Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: PCR Protocols: A Guide to Methods and Applications (eds. M.A. Innis, D.H. Gelfand, J.J. Sninsky and T.J. White). Academic Press, San Diego: 315322. Yoshioka, K. (2002). KyPlot - a user-oriented tool for statistical data analysis and visualization. Computational Statistics 17: 425-437. Fungal Diversity 111