Mycologia, 97(5), 2005, pp. 1023–1033. # 2005 by The Mycological Society of America, Lawrence, KS 66044-8897 Phylogenetic origins of two cleistothecial fungi, Orbicula parietina and Lasiobolidium orbiculoides, within the operculate discomycetes K. Hansen1 B.A. Perry D.H. Pfister eral times within apothecial and perithecial lineages of ascomycetes. Molecular phylogenetic analyses indicate one major group (composed of Ascosphaerales, Onygenales and Eurotiales) containing most cleistothecial ascomycetes but that other cleistothecial fungi fall within other ascomycete groups (Berbee and Taylor 1994, LoBuglio et al 1996). We investigate the placement of two cleistothecial fungi, Orbicula parietina and Lasiobolidium orbiculoides, that have been suggested to have pezizalean affinities. The genus Orbicula Cooke (1871) produces epigeous, globose ascomata, up to 2 mm in diam, without an opening (FIGS. 1–3). At first sight it resembles, and is often mistaken for, a slime mold (a miniature Lycogala) (Hughes 1951). Asci lack an apical opening mechanism and disintegrate at maturity. The ascomata become filled with a dry mass of ascospores that are liberated when the peridium is disturbed and broken by an external force (FIG. 2). Orbicula parietina, the only accepted species, has been repeatedly described as new and assigned to nine genera, four in the myxomycetes (Hughes 1951). Moreover, the family Orbiculaceae and order Orbiculales (Ascomycota) were erected for Orbicula (Locquin 1974), but invalidly published (ICBN Art. 36.1). Malloch and Cain (1971) thought that Orbicula was closely related to members of the Thelebolaceae and placed it, along with other cleistothecial fungi—Cleistothelebolus, Eoterfezia, Lasiobolidium, Microeurotium and Xeromyces—in the cleistothecial family Eoterfeziaceae, which they considered Pezizales. Jeng and Krug (1976) transferred the genera of Eoterfeziaceae including Orbicula to the tribe Theleboleae sensu Korf (1972) of the Pyronemataceae, following the notion that closely related genera with exposed hymenia and cleistothecial genera are better accommodated in one rather than separate families. Benny and Kimbrough (1980) maintained Orbicula in Eoterfeziaceae, while Dennis (1981) placed it in Eurotiaceae (Plectascales). Arx (1981) treated Orbicula in the Pezizales and Campbell et al (1991) likewise suspected affinities with the operculate discomycetes. Malloch (in Dissing and Schumacher 1994) suggested that both Orbicula and Lasiobolidium, with clearly pezizalean characteristics, might better be included in Pyronemataceae or in the Pezizales without assignment rather than in Eoterfeziaceae. Currently Orbicula is placed in the Pyronemataceae, but its disposition is indicated tentatively with a question mark (Eriksson 2005). Harvard University Herbaria, Cambridge, Massachusetts 02138 Abstract: Parsimony, maximum-likelihood and Bayesian analyses of SSU rDNA sequences of representative taxa of Pezizomycetes, Eurotiomycetes, Dothideomycetes, Leotiomycetes and Sordariomycetes, all strongly support the cleistothecial fungi Orbicula parietina and Lasiobolidium orbiculoides to be of pezizalean origin. Previous hypotheses of close affinities with cleistothecial or highly reduced fungi now placed in the Thelebolales, Eurotiales or Onygenales are rejected. Orbicula parietina and L. orbiculoides are deeply nested within Pyronemataceae (which subsumes the families Ascodesmidaceae, Glaziellaceae and Otideaceae). LSU rDNA sequences suggest that Orbicula is nested within the apothecia-forming genus Pseudombrophila (including Nannfeldtiella and Fimaria) and that L. orbiculoides is closely related. Ascodesmis and Lasiobolus, which have been suggested as closely related to Orbicula and Lasiobolidium, are identified as a sister lineage to the Pseudombrophila lineage. Cleistothecial forms that have lost the ascus operculum and ability to discharge spores actively have evolved at least once in the Pseudombrophila lineage. Some species of Pseudombrophila produce subglobular ascomata initials that are closed early in development and open only in the mid-mesohymenial phase. We hypothesize that, in the Pseudombrophila lineage, ascomata forms that never open are derived from ascomata that open late in development. The placement of O. parietina and L. orbiculoides within Pseudombrophila is supported by morphological characters, ecology and temperature optima for fruiting. Key words: ascoma evolution, molecular phylogenetics, Pezizales, Pseudombrophila, Pyronemataceae INTRODUCTION Morphological and molecular evidence show that cleistothecial fungi have evolved independently sevAccepted for publication 30 Aug 2005. 1 Corresponding author. E-mail: khansen@oeb.harvard.edu 1023 1024 MYCOLOGIA FIGS. 1–6. Orbicula parietina. 1. Cleistothecia on dung of dove (?) 32.5. 2. Close-up of cleistothecia, globose with a flattened base, glabrous over the upper surface and with hyphoid hairs originating from the base. In one ascomata the thin outer excipulum is broken open to show the contents of the yellowish spores 35. 3. Median, longitudinal section through a mature ascomata to show basal cushion on which asci and paraphyses are borne 315. (FIGS. 1–3: JHP-01.013/PH01-001, C). 4. Light microscopic slide (LM) of part of lower cleistothecial wall in surface view, with long, hyphoid hairs coming from the base of the cleistothecia 3100. 5. LM of hyphoid, flexuous, pale brownish, thick-walled and remotely septate ascomatal hairs 3200. 6. LM of outer excipulum cells seen in surface view; cells thick-walled, small and irregularly lobed (puzzle-like) with intercellular, amorphous, brownish pigment. Hairs originating from outer excipulum cells, with forked bases 3400. (FIGS. 4–6: C F-24441). Photos: 1–3 J.H. Petersen. 4–6 K. Hansen. Lasiobolidium orbiculoides was the second species described in Lasiobolidium (Malloch and Benny 1973). The specific epithet orbiculoides refers to its similarity to Orbicula; O. parietina and L. orbiculoides both have broadly oblate, uniseriate ascospores (Malloch and Benny 1973). Lasiobolidium orbiculoides differs from the type species of Lasiobolidium, L. spirale, by being fast-growing, producing ascomata with wavy to flexuous, septate appendages, an excipulum of one tissue type, cylindrical uniseriate asci and oblate ascospores (Moustafa and Ezz-Eldin 1989). Lasiobolidium was considered a cleistothecial counterpart of Lasiobolus (Thelebolaceae, Pezizales), but the genus was at first placed in the Eoterfeziaceae (Malloch and Cain 1971). Jeng and Krug (1976) transferred Lasiobolidium to the Theleboleae sensu Korf (1972) of the Pyronemataceae. Developmental studies of L. orbiculoides however, have shown little evidence of affinities with either Lasiobolus or Thelebolaceae, rather they support a relationship with other operculate discomycetes, especially with Ascodemis (Janex-Favre and Locquin-Linard 1979). The type genus of Thelebolaceae, Thelebolus, was found to be non-pezizalean based on ascoma development and ascus structure (e.g., Samuelson and Kimbrough 1978, Kimbrough 1981). Based on mo- lecular phylogenetic analyses (Momol et al 1996, Landvik et al 1997) the Thelebolaceae was found to be closely related to Erysiphales and Leotiales, and was moved to the Leotiomycetes (Eriksson and Winka 1997). Orbicula, Lasiobolus and Lasiobolidium were maintained in the Pezizales, in Otideaceae (Eriksson 1999) (5Pyronemataceae [Eriksson et al 2001, Eriksson 2005]). Further studies by Landvik et al (1998) confirmed the relationship of Thelebolaceae with members of the Leotiomycetes, and grouped Lasiobolus with Ascodesmis in the Pezizales. In the present study we performed phylogenetic analyses of the SSU rDNA to address the phylogenetic position of O. parietina and L. orbiculoides within the ascomycetes; and analyses of the LSU rDNA to address specific relationships of O. parietina and L. orbiculoides within the Pyronemataceae. MATERIALS AND METHODS Specimens.—Specimens sequenced and morphologically examined are listed (TABLE I). To test hypotheses regarding relationships of Orbicula and L. orbiculoides, the SSU rDNA sequences were analyzed with 61 sequences retrieved from GenBank from representative taxa of Pezizomycetes, Eurotiomycetes (Eurotiales and HANSEN ET AL: ORIGINS OF TWO CLEISTOTHECIAL FUNGI 1025 TABLE I. Source of material examined and sequenced by the authors. Numbers in parentheses indicate multiple collections of a single taxon Species Collection number (Herbarium), date and collector Ascodesmis nigricans Tiegh. CBS 389.68, Netherlands, Wageningen, 29.I.1986, G. Tichelaar Eleutherascus lectardii (Nicot) Arx CBS 626.71, France, Moselle, I.1968, P. Lectard Geopyxis carbonaria (Alb. & Schwein.) C F-49793 (C), Denmark, Jutland, 13.XI.1982, Sacc. T. Læssøe Geopyxis sp. KH.04.48 (FH, DBG), USA, Colorado, 13.IX.2004, K. Hansen, V. Evenson Glaziella aurantiaca (Berk. & Curt.) PR-5954 (FH), Puerto Rico, Loquillo Mts., Cooke 11.VI.1998, N.C. Clum, D.J. Lodge. Greletia reticulosperma Donadini, Part of isotype (herb. Roy Kristiansen), France, Riousset & G. Riousset 1984, G. Riousset Lasiobolidim orbiculoides Malloch & Benny CBS 344.73, USA, California, dung of deer, 10.VI.1953, G.L. Benny Lasiobolus cuniculi Velen. Rana 76.053 (C), Norway, Nordland, 9.IX,1976, H. Dissing Lasiobolus ciliatus (Berk.) Sacc. KS-94-005 (C), Denmark, Møn, 5.IV.1994, K. Hansen, S.K. Sandal Orbicula parietina (Schrad.) S. Hughes C F-24441 (C), Denmark, Zealand, on rush mat, 1988, H.F. Gøtzsche Otidea onotica (Pers.) Fuckel KH-98-107 (C), Denmark, Zealand, 21.VII.1998, K. Hansen Otidea umbrina (Pers.) Bres. KH.01.09 (C), Denmark, Bornholm, 30.IX.2001, C. Lange Paurocotylis pila Berk. Trappe 12583 (OSC), New Zealand, South Island, 24.IX.1993, M. Amaranthus Pseudombrophila guldeniae Svrcek (1) Kongsv. 85.10B (C), Norway, Oppdal, 23.VIII.1985. H. Dissing, S. Sivertsen Pseudombrophila guldeniae (2) s.n. (FH, part in C and TRH), Norway, Oppdal, 13.VI.1985, S. Sivertsen, I. Dissing, H. Dissing Pseudombrophila merdaria (Fr.) s.n. (FH), USA, ME, on manured soil, Brumm. (1) 30.VII.1994, D.H. Pfister Pseudombrophila merdaria (2) s.n. (FH), USA, VE, on composted silage at edge of hay field, VI.1979, M. Shemluck Pseudombrophila merdaria (3) s.n. (FH), USA, IA, no date, T.J. Farrell Pseudombrophila theioleuca Rolland C F-70057 (C), Denmark, on deer dung, 25.IX.1982, H. Knudsen Pulvinula constellatio (Berk. & Broome.) KH.03.64 (FH), Norway, Nordland, Boud. 22.VIII.2003, K. Hansen, C. Lange Pulvinula convexella (P. Karst.) Pfister KH.01.020 (C), Denmark, Fyn, 5.X.2001, K. Hansen Smardaea amethystina (W. Phillips) KH-97-132 (C), Denmark, Zealand, 1997, C. Svrcek Lange, K. Hansen Tarzetta catinus (Holmsk.) Korf & J.K. KS.94.10A (C), Denmark, Møn, 11.V.1994, K. Rogers Hansen, S.K. Sandal Tarzetta pusilla Harmaja KH.03.66 (FH), Norway, Nordland, 22.VIII.2003, K. Hansen, C. Lange a LSU SSU DQ168335a — DQ168334a DQ168336a DQ062997 — DQ062985 — — DQ062996 AY500532 — DQ062995 DQ063000 DQ168338a — DQ167411a — DQ062988 DQ062998 AF335121 — AY500540 — DQ168337a — DQ062993 DQ063001 DQ062994 — DQ062990 — DQ062991 — DQ062992 DQ062989 DQ062999 DQ062987 — DQ062986 — AF335176 — DQ062984 — DQ062983 — Sequences from manus in prep. Onygenales), Dothideomycetes (Pleosporales), Leotiomycetes (Helotiales, Erysiphales and Thelebolales) and Sordariomycetes (Sordariales and Hypocreales): Ascobolus carbonarius P. Karst., AY544720; Ascodesmis sphaerospora W. Obrist., U53372; Ascozonus woolhopensis Renny, AF010590; Barssia oregonensis Gilkey, U42657; Byssonectria terrestris (Alb. & Schwein. : Fr.) Pfister, Z30241; Caccobius minusculus Kimbr.; Chaetomium elatum Kunze, M83257; Chalazion helveticum Dissing, AF061716; Cheilymenia stercorea (Pers. : Fr.) Boud., U53375; Chorioactis 1026 MYCOLOGIA geaster (Peck) Kupfer, AF104340; Cookeina speciosa (Fr. : Fr.) Dennis (5C. sulcipes (Berk.) Kuntze), U53376; Desmazierella acicola Lib., AF104341; Discina macrospora Bubák, U42651; Elaphomyces maculatus Vittad., U45440; Erysiphe orontii Castagne, AB033483; Eurotium herbariorum (F.H. Wigg.) Link, AB008402; Geopyxis carbonaria (Alb. & Schwein. : Fr.) Sacc., AF104665; Gymnoascus reessii Baran., AB015774; Gyromitra montana Harmaja, U42652; Hymenoscyphus ericae (D.J. Read) Korf & Kernan, AY524847; Iodophanus carneus (Pers. : Fr.) Korf, U53380; Lamprospora kristiansenii Benkert, AF121075; Lasiobolus papillatus (Pers. : Fr.) Sacc., AF010588; Monascus ruber Tiegh., AB024048; Morchella elata Fr. : Fr., U42641; Nectria cinnabarina (Tode : Fr.) Fr., AB003949; Neottiella rutilans (Fr.) Dennis, AF061720; Neurospora crassa Shear & B.O. Dodge, X04971; Helvella lacunosa Afzel. : Fr., U53378; Leucangium carthusianum (Quél.) Paol., U42647; Onygena equina (Willd. : Fr.) Pers., U45442; Paurocotylis pila Berk., U53382; Peziza succosa Berk., U53383; Phyllactinia guttata (Wallr. : Fr.) Lév., AF021796; Plectania rhytidia (Berk.) Nannf. & Korf, AF104344; Pleospora herbarum (Pers. : Fr.) Rabenh., U05201; Pulvinula archeri (Berk.) Rifai, U62012; Pyronema domesticum (Sowerby : Fr.) Sacc., U53385; Reddellomyces donkii (Malençon) Trappe et al, U42660; Rhizina undulata Fr., U42664; Sarcosphaera coronaria (Jacq.) Boud., AY544712; Saccobolus sp., U53393; Sarcoscypha austriaca (Berk.) Boud., AF006318; Sarcosoma globosum (Schmidel : Fr.) Casp., U53386; Sclerotinia sclerotiorum (Lib.) de Bary, L37541; Scutellinia scutellata (L. : Fr.) Lambotte, U53387; Sordaria fimicola (Roberge ex Desm.) Ces. & De Not., AY545724; Sporormia lignicola W. Phillips & Plowr., U42478; Talaromyces flavus (Klöcker) Stolk & Samson, M83262; Tarzetta catinus (Holmsk. : Fr.) Korf & J.K. Rogers, U53389; Thecotheus holmskjoldii (E.C. Hansen) Chenant., AF010589; Thelebolus crustaceus (Fuckel) Kimbr., U53394; Thelebolus stercoreus Tode : Fr., U49936; Trichophaea hybrida (Sowerby.) T. Schumach., U53390; Tuber gibbosum Harkn., U42663; Wilcoxina mikolae (Chin. S. Yang & Wilcox) Chin. S. Yang & Korf, U62014; Xeromyces bisporus L.R. Fraser, AB024049; Zopfia rosatii (Segretain & Destombes) D. Hawksw. & C. Booth, L76623. Neolecta Speg. and Taphrina Fr. were used to root the SSU trees (Neolecta vitellina [Bres.] Korf & J.K. Rogers, Z27393; Neolecta irregularis [Peck] Korf & J.K. Rogers, Z47721; Taphrina pruni Tul., AJ495828). Representatives of Saccharomycetes were also included (Kluyveromyces lactis (Boidin, Abadie, J.L. Jacob & Pignal) Van der Walt, AY790534; Saccharomyces cerevisiae E.C. Hansen, V01335; Pichia sp., AY227899). Ingroup taxa in the LSU data set, in addition to the Pseudombrophila lineage, were selected based on preliminary results of a large-scale molecular study of the Pyronemataceae (in prep). In that study, based on analyses of LSU rDNA sequences, the type species of Lasiobolidium, L. spirale, is placed in a different lineage of Pyronemataceae distant from L. orbiculoides (data not shown). Therefore, L. spirale is not included in the present study. The genus Pseudombrophila sensu Brummelen (1995) includes two sections, Pseudombrophila and Nannfeldtiella. The type species of both sections, P. merdaria and P. guldeniae, were included. Smardaea and Greletia were used to root the LSU trees, based on the results of our large-scale Pyronemataceae study (in prep.), which resolved these taxa in a clade basal to the ingroup of this investigation. The effect of additional, alternative outgroups was explored (Pyronema and Byssonectria). Molecular methods and Analyses.—Laboratory techniques generally followed procedures outlined in Hansen et al (2002). SSU rDNA was amplified using the PCR with primer pair SL1 (Landvik et al 1997) and NS8 (White et al 1990). In addition to the primers used for PCR, internal primers SL122 and SL344 (Landvik et al 1997), and NS2, NS4 and NS6 (White et al 1990) were used for sequencing. The 59 end of the LSU-rDNA was amplified using the primer pairs LROR and LR5 (Moncalvo et al 2002). These primers and two internal primers, LR3 and LR3R (Moncalvo et al 2002), were used for sequencing. Electrophoresis and data collecting were done on an ABI PRISMH 3100 or 3730 DNA sequencer (Perkin-Elmer/ABI). To verify the LSU rDNA sequence of Orbicula, DNA was extracted and sequenced twice, on two different occasions, from coll. no. F-24441 (C). Sequences were edited using Sequencher 3.0 (Gene Codes, Ann Arbor, Michigan). Sequences are deposited in GenBank (TABLE I). Nucleotide sequences were aligned by hand using the software program Se-Al v. 2.0a11 (Rambaut 2002). Alignments are available from TreeBase (http://www.treebase.org/treebase/) as accessions M2365 (SSU) and M2366 (LSU). Analyses were performed using PAUP* 4.0b 10 for Unix (Swofford 2002) and MrBayes 3.0b4 (Huelsenbeck and Ronquist 2001) on a G5 Macintosh computer. Maximum parsimony (MP), Bayesian and Maximum likelihood (ML) analyses were performed as in Hansen et al (2005), except MP analyses of the LSU data used branch-andbound searches, ML model parameter values were estimated on the LSU data set, and Bayesian MCMC were run for 2 000 000 generations. To select the model of nucleotide substitution with the least number of parameters best fitting the SSU data set, hierarchical likelihood ratio tests were performed as implemented in the program MrModeltest 2.2 (Nylander 2004). In Bayesian analyses of the SSU and LSU data, the first 2000 trees and 1000 trees were deleted respectively as the ‘‘burn-in’’ period of the chain. In addition to parsimony bootstrap proportions (BP) and Bayesian posterior probabilities (PP), ML bootstrap proportions (ML-BP) were generated using 100 bootstrap replicates of ‘‘fast’’ stepwise sequence addition for the SSU data, and using 500 bootstrap replicates of heuristic searches, with 10 random addition sequences and TBR branch swapping for the LSU data. The model parameters estimated for the ML analyses were entered manually into PAUP for the ML-BP searches. Topologically constrained MP and ML analyses of the LSU data set were used to evaluate if the cleistothecial form originated once or twice from apothecia-forming taxa HANSEN ET AL: ORIGINS OF TWO CLEISTOTHECIAL FUNGI with loss of active spore discharge. MacClade 4.05 (Maddison and Maddison 2002) was used to construct a constraint tree with O. parietina and L. orbiculoides forced to be monophyletic. Parsimony and ML analyses were performed under the constraint, using the same settings as specified above. For the ML analyses the estimated model parameters were used. The KishinoHasegawa test (Kishino and Hasegawa 1989) was used to compare the trees under the constrained and unconstrained topologies in PAUP. RESULTS Morphological features.—The material of O. parietina used for DNA extraction (C F-24441) agrees with the description by Hughes (1951) and Udagawa and Furuya (1972). This collection is fully mature; asci have completely disintegrated and the dry ascoma is filled with loose, powdery, yellow spores. We want to highlight the following characters in O. parietina. Dried mature ascomata show no signs of breaking open (FIG. 1), but a light touch ruptures the thin and brittle outer excipular wall (FIG. 2). Ascomata are glabrous and loosely covered at the base with pale brown hyphal hairs, that are up to 120 mm long or possibly longer, up to 5 mm wide, flexuous, thick-walled (1.2 mm) and remotely septate (FIGS. 4 and 5). The hairs originate from the base of the ascoma, are rounded at the tips, abundant and often form a loose ‘‘mat’’ at the base of the ascoma, almost like a subiculum. The cells of the outer excipulum are, in surface view, brown, thick-walled, small and irregularly lobed in outline (puzzle-like) (FIG. 6). Phylogenetic position of Orbicula and L. orbiculoides among the Ascomycetes; the SSU phylogeny.—-The SSU rDNA data set included 1721 characters with 398 being parsimony informative. Parsimony analysis resulted in 16 MPTs (1667 steps, CI 5 0.496, RI 5 0.702). ML analysis resulted in one optimal tree (2lnL 5 11 857.5723, FIG. 7) under the GTR + I + G model of sequence evolution selected by MrModeltest with base frequencies A 5 0.2593, C 5 0.2042, G 5 0.2645, T 5 0.2720, and the substitution model: [A–C] 5 1.2438, [A–G] 5 2.3987, [A–T] 5 1.5819, [C–G] 5 0.8876, [C–T] 5 4.9308, [G–T] 5 1.0000, with a proportion of invariable sites I 5 0.4510, and a gamma distributed shape parameter 5 0.5648. The Pezizomycetes form a moderately supported monophyletic group in the MP and Bayesian analyses (BP 77%, PP 100%), as a sister group to all other Euascomycetes sampled (including Thelebolales) (PP 97%, FIG. 7). Although relationships between classes are resolved with only weak support (BP , 50%), the 1027 Dothideomycetes, Eurotiomycetes and Sordariomycetes are highly supported (all 100%). Leotiomycetes is resolved as monophyletic in all analyses but with low support. The orders, Thelebolales and Erysiphales are strongly supported as monophyletic (BP 92–100%, ML-BP 86–100%, PP 100%) within the Leotiomycetes. Orbicula parietina and L. orbiculoides form a highly supported group with two species of the apothecial genus Pseudombrophila in all analyses (BP 96%, MLBP 92%, PP 100%, FIG. 7). The Pseudombrophila lineage is deeply nested within a highly supported lineage of members of Pyronemataceae s.l., Sarcoscyphaceae and Sarcosomataceae (BP 92%, ML-BP 78%, PP 100%) (lineage I) within the Pezizales. A few potential sister lineages are supported e.g., the Ascodesmis-Lasiobolus lineage (BP 84%, ML-BP 75%, PP 100%) and the Geopyxis-Tarzetta lineage (PP 100%). The Pyronemataceae as sampled here, including Ascodesmidaceae and Glaziellaceae, are monophyletic and supported by Bayesian analyses (PP 97%, FIG. 7). The families Discinaceae, Helvellaceae, Morchellaceae, Rhizinaceae and Tuberaceae are highly supported (BP 90–100%, PP 98–100%) and form a weakly supported sister group (II) to the lineage I (FIG. 7). Lineages I and II form a moderately supported group by MP and Bayesian analyses (BP 73%, PP 100%). Ascobolaceae and Pezizaceae are strongly supported (each BP and PP 100%, ML-BP 98–100%) and form a weakly supported lineage (III, BP 65%) that is sister to the lineages I and II. Relationships of the Pseudombrophila lineage within Pyronemataceae; the LSU phylogeny.—Branch and Bound searches resulted in 2 MPTs (501 steps, CI 5 0.653, RI 5 0.749) produced from 915 total characters, of which 182 are parsimony informative. ML analysis resulted in one optimal tree (2lnL 5 3797.8522, FIG. 8) under the GTR + I + G model with base frequencies A 5 0.2519, C 5 0.2051, G 5 0.3044, T 5 0.2386, and the substitution model: [A–C] 5 0.6839, [A–G] 5 3.0327, [A–T] 5 1.9889, [C–G] 5 0.6644, [C–T] 5 6.2107, [G–T] 5 1.0000, with a proportion of invariable sites I 5 0.5745, and a gamma distributed shape parameter 5 0.7125. The trees recovered by the different analyses of the LSU data did not posses any conflict. Trees obtained in analyses with Pyronema and Byssonectria as alternative outgroups (not shown) were identical to trees obtained with Smardaea and Greletia as outgroup. Confirming the SSU data, O. parietina and L. orbiculoides form a highly supported group with Pseudombrophila theioleuca, P. merdaria and P. guldeniae, in all analyses (BP 90%, ML-BP 77%, PP 100%, FIG. 8). Orbicula groups with P. theioleuca with 1028 MYCOLOGIA high support (BP 96%, ML-BP 93%, PP 100%), otherwise relationships within the Pseudombrophila lineage are not supported. A clade of three lineages, the Geopyxis-Tazzetta, Ascodesmis-Lasiobolus and Pulvinula lineages, is resolved as the sister group to the Pseudombrophila lineage in all analyses, and highly supported by MP bootstrap (BP 86%). The GeopyxisTazzetta, Ascodesmis-Lasiobolus and Pulvinula lineages are all highly supported (BP 92–100%, ML-BP 97– 100%, PP 100%), but relationships between the lineages are unresolved in the strict consensus tree of the MPTs and by Bayesian analyses, and with only HANSEN ET AL: ORIGINS OF TWO CLEISTOTHECIAL FUNGI 1029 FIG. 8. Phylogenetic relationships of Orbicula and Lasiobolidium orbiculoides among members of the Pyronemataceae (taxa selection are based on a large-scale molecular study of the Pyronemataceae (in prep.) inferred from LSU rDNA sequences. The tree with the highest likelihood (2lnL 5 3797.8522) obtained from maximum likelihood analyses. Branch length corresponds to genetic distance (expected nucleotide substitutions per site). Numbers above branches are posterior probabilities (PP $ 95%), obtained from the 50% majority rule consensus tree of the 19 000 trees sampled from a Bayesian MCMC analysis. Numbers below branches that are before the backslash are ML bootstrap proportions (ML-BP . 70%) and numbers after the backslash are MP bootstrap proportions (BP . 70%). low ML-BP support (,50%, FIG. 8). Otidea forms a separate lineage, sister to the highly supported clade of the Pseudombrophila, Geopyxis-Tarzetta, AscodesmisLasiobolus and Pulvinula lineages (99–100%). The most parsimonious interpretation of the molecular LSU phylogeny suggests that the cleistothecial form originated twice in the Pseudombrophila lineage (FIG. 8). Nevertheless, constraint MP and ML analyses forcing O. parietina and L. orbiculoides into a monophyletic group could not be rejected using the Kishino-Hasegawa test (P , 0.05). Forced monophyly of the cleistothecial taxa did not yield trees that were significantly less likely or longer than the uncon- strained trees (MLT: 2lnL 5 3806.9008, difference in 2LnL 5 9.0486; MPT: 505 steps, 4 steps longer than the unconstrained MPTs). DISCUSSION Evolutionary relationships.—Molecular phylogenetic analyses confirm the evolutionary origins of O. parietina and L. orbiculoides within Pyronemataceae (Pezizales) as suggested by Malloch (in Dissing and Schumacher 1994). Other hypotheses, such as close affinities with cleistothecial or highly reduced fungi now placed in the Thelebolales, r FIG. 7. Phylogenetic placement of Orbicula and Lasiobolidium orbiculoides within the ascomycetes inferred from SSU rDNA sequences. The tree with the highest likelihood (2lnL 5 11 857.5723) obtained from maximum likelihood analyses. Branch length corresponds to genetic distance (expected nucleotide substitutions per site). Numbers above branches are posterior probabilities (PP $ 95%), obtained from the 50% majority rule consensus tree of the 18 000 trees sampled from a Bayesian MCMC analysis. Numbers below branches that are before the backslash are ML ‘‘fast’’ bootstrap proportions (ML-BP . 70%) and numbers after the backslash are MP bootstrap proportions (BP . 70%). Selected classifications are indicated on the tree for discussion. The lineages I, II and III represents the pezizalean families: Pyronemataceae (including Ascodesmidaceae and Glaziellaceae), Sarcoscyphaceae and Sarcosomataceae (I); Discinaceae, Helvellaceae, Morchellaceae, Rhizinaceae and Tuberaceae (II); and Ascobolaceae and Pezizaceae (III). 1030 MYCOLOGIA FIGS. 9–10. Apothecia of Pseudombrophila ripensis (E.C. Hansen) Brumm. developing from large sclerotia, on old dung of cow (JHP-95.104, C). 9. Apothecia at first sub-globular and closed, here deeply cup-shaped to urnulate; the excipular roof over the hymenium has opened and left the margins raised and ragged 31. 10. Apothecia fully expanded, flattened to curvedlobate, showing dentate to irregular laciniate prominent margin from the disruption of the excipular roof, and receptacles covered with woolly hairs 31. Photos: J.H. Petersen. Eurotiales or Onygenales are rejected (FIG. 7). The position within the Pezizales is supported by morphology; the genus Orbicula and L. orbiculoides possess some clearly pezizalean characteristics such as uniseriate, narrowly clavate to cylindrical asci and unicellular, large, hyaline, thin-walled ascospores (Hughes 1951, Udagawa and Furuya 1972, Malloch and Benny 1973). For the first time, a close relationship between species of Pseudombrophila and O. parietina is suggested. Furthermore, a close relationship between O. parietina and L. orbiculoides, as indicated by Malloch and Benny (1973), is shown. Evolution of cleistothecial forms.—The cleistothecial form, with loss of active spore discharge, has evolved at least once in the Pseudombrophila lineage (FIG. 8). Species of Pseudombrophila produce discto cup-shaped apothecia with an exposed hymenium at maturity (FIGS. 9, 10) and forcibly discharging, operculate asci. Epigeous, open apothecia of various shapes with forcible discharge are the most common ascoma form in the Pezizales and presumably the ancestral state. Cain (1956a, b, 1961) and Malloch (1979, 1981) were among the first to suggest that certain cleistothecial ascomycetes were derived from apothecial and perithecial forms, a concept now widely accepted. Cain (1956a) believed that the cleistothecial fungi represented a large number of unrelated and highly evolved lineages, arguing that these fungi were adapted to passive spore dispersal in very specific ecological niches. Malloch (1979, 1981) placed several cleistothecial taxa in the Pezizales, of which only Cleistothelebolus, Warcupia and Orbi- cula are still retained in the order. Recent molecular phylogenetic analyses, especially within perithecial lineages, have strengthened the evidence that the cleistothecial ascoma with concomitant loss of active spore discharge has arisen independently from ostiolate forms multiple times (e.g., Berbee and Taylor 1992, Rehner and Samuels 1995, Suh and Blackwell 1999). An analogous situation within the Pezizales is the repeated evolution of hypogeous or semi-hypogeous, closed ascomata, with loss of active spore discharge. Truffle and truffle-like forms evolved independently at least 10 times in six different families (e.g. Trappe 1979, O’Donnell et al 1997, Landvik et al 1997, Percudani et al 1999, Hansen et al 2001 and unpublished). Molecular data supports the view (e.g. Trappe 1979) that the truffles are all derived from epigeous apothecial ancestors, through selection forces related to animal mycophagy, reduction in water loss (Thiers 1984, Bruns et al 1989), or selection in mycorrhizal fungi for deposition into the soil spore-bank (Miller et al 1994). The grouping of O. parietina and L. orbiculoides, taxa that produce completely closed ascomata, with Pseudombrophila, is not as surprising as it first might seem. All species of Pseudombrophila produce subglobular ascomatal primordia and in P. hepatica, P. leporum, P. cervaria and P. theioleuca these are closed at early stages. The excipular roof over the hymenium opens in the mid-mesohymenial phase when the asci are ripening (Brummelen 1995). In most species of Pseudombrophila section Pseudombrophila the evidence of the tearing of the excipular roof is seen in mature ascomata, by a ragged rim of the excipular margin (FIGS. 9, 10). We hypothesize that in the HANSEN ET AL: ORIGINS OF TWO CLEISTOTHECIAL FUNGI Pseudombrophila lineage, ascomata forms that never open (FIG. 1) are derived from ascomata that open late in development (FIGS. 9, 10). Once the opening of the ascomata is lost, relaxation of selection for forcible spore discharge may permit loss of some or all of the accompanying morphological traits, such as the loss of the operculum and/or loss of a distinct hymenial layer (Malloch 1981). In a transition from open apothecia to cleistothecial forms, Orbicula may be considered a morphological intermediate and L. orbiculoides a more derived form. In O. parietina asci arise in a hymenium at the base of the ascoma accompanied by paraphyses, whereas in L. orbiculoides asci are not arranged in a parallel layer but rather arise in small clusters at several points within the ascoma and paraphyses are lacking (Malloch and Benny 1973). Morphology and ecology.—The excipular hairs found in Orbicula (FIGS. 4, 5) are of the same type as found in some species of Pseudombrophila, most notably in species previously placed in Fimaria, e.g. P. theioleuca, P. hepatica and P. dentata (Brummelen 1962, Jeng and Krug 1977 with P. dentata as Fimaria trochospora Jeng & Krug, Pfister 1984). All hairs in Pseudombrophila are pale brown, hyphoid and originate from the outermost layer of the excipulum from base to margin (Brummelen 1995). Ascomatal hairs of L. orbiculoides also originate from the outermost excipular cells, are flexuous to wavy, remotely septate, thick-walled, very long (2 to 3 mm; Malloch and Benny 1973), and scattered all over the ascoma. In Pseudombrophila species the outer excipulum is differentiated from the underlying medullary excipulum (Brummelen 1995) and is comparable to the excipulum of O. parietina and L. orbiculoides. It is composed of thick-walled cells, subglobose to irregularly lobed in outline in surface view (consisting of closely compacted oblong or isodiametric cells) (FIG. 6). Orbicula parietina has a more differentiated pezizalean excipulum than L. orbiculoides, which has a simple one-celled excipulum (Malloch and Benny 1973). In O. parietina the outermost cells of the excipulum are thick-walled, small and brown (FIG. 6), compared to the inner cells, which are thin-walled, large and hyaline (Hughes 1951). In all species of Pseudombrophila and in O. parietina the pigment is intercellular and amorphous (Brummelen 1995, and FIG. 6). The similarities in ecological requirements and substrate also support the close relationships between species of Pseudombrophila, O. parietina and L. orbiculoides. All are likely saprobes, fruiting on dung or well-rotted material, and species of Pseudombro- 1031 phila and O. parietina fruit in cold conditions. The optimal temperature for developing fruit-bodies of Pseudombrophila is between 10 and 15 C, and thus they are found more frequently in spring and in mild winters (Brummelen 1995). Arx (1981) listed O. parietina as psychrophilous and Campbell et al (1991) reported Orbicula fruiting in early spring on Pseudocercosporella-inoculated oat kernels left on sand over winter, and considered it to be psychrophilic. Species of Pseudombrophila are strictly coprophilous (FIGS. 9 and 10); occur on soil or vegetable debris contaminated with dung, urine or urea; on decaying stems and leaves of plants; or on rotting materials (Brummelen 1995). Orbicula parietina is most commonly reported on a variety of rotting substrates, including damp paper, cardboard, straw, willow baskets and leaves, or directly on dung (Hughes 1951, Udagawa and Furuya 1972, FIGS. 1–3). Hughes (1951) found the best medium for fruiting of Orbicula to be tapwater agar with ground weathered rabbit pellets. Lasiobolidium orbiculoides occurs on dung but also has been reported from soil (in Yaguchi et al 1996). Conclusions and taxonomic directions.—Based on LSU rDNA sequence data, morphology and ecology we consider Orbicula and Pseudombrophila sensu Brummelen (1995) to be congeneric. The genus Pseudombrophila sensu Brummelen (1995) includes the formerly recognized genera Fimaria and Nannfeldtiella. Further sampling of species of Pseudombrophila for molecular phylogenetic study may reveal that these genera represent separate lineages that could deserve recognition. Orbicula groups with P. theioleuca, a species placed in Fimaria by Brummelen (1962). The type species of Fimaria, P. hepatica, was not sampled. Ascomata that open late in development and similarities in excipular hair morphology supports the placement of O. parietina in Fimaria. In the Pseudombrophila lineage Orbicula (Cooke 1871) is the earliest generic name and thus has priority. If this lineage is recognized as a single genus, we will propose to conserve Pseudombrophila against Orbicula, because Pseudombrophila includes a larger number of species (28 total, Brummelen 1995). Lasiobolidium orbiculoides also may be considered congeneric with Pseudombrophila. However, based on LSU rDNA sequences L. orbiculoides is not unambiguously nested within Pseudombrophila (FIG. 8) and therefore no combination will be made. To further understand phylogenetic relationships in the Pseudombrophila lineage, and to determine the number of times the cleistothecial form has arisen, a larger sampling of taxa within the lineage is needed. Detailed developmental 1032 MYCOLOGIA studies of Pseudombrophila species and L. orbiculoides have been done (Brummelen 1995, JanexFavre and Locquin-Linard 1979), but comparable studies with O. parietina are lacking. Studies of ascomatal development in O. parietina would provide data that could further clarify the evolution of closed forms in the apothecia-forming operculate lineage. ACKNOWLEDGMENTS We wish to thank the curator at C for arranging loan of material, Jens H. Petersen for use of his photographs (FIGS. 1–3, 9 and 10), Gustavo Romero for assembling the photographic plates, and Thomas Læssøe and Katherine F. LoBuglio for comments on the manuscript. This research was financially supported by an NSF grant to KH and DHP (DEB-0315940). LITERATURE CITED Arx JA von. 1981. The genera of fungi sporulating in pure culture. 3rd ed. Vaduz: J Cramer. 424 p. Benny GL, Kimbrough JW. 1980. A synopsis of the orders and families of Plectomycetes with keys to genera. Mycotaxon 12:1–91. Berbee ML, Taylor JW. 1992. Convergence in ascospore discharge mechanism among Pyrenomycete fungi based on 18S ribosomal RNA gene sequence. Mol Phylogenet Evol 1:59–71. ———, ———. 1994. 18S ribosomal DNA sequence data and dating, classifying and ranking the fungi. In: Hawksworth DL. ed. Ascomycete Systematics: Problems and Perspectives in the nineties. NATO ASI Series 269. New York: Plenum Press. 453 p. Brummelen J van. 1962. Studies on Discomycetes-II. On four species of Fimaria. Persoonia 2:321–330. ———. 1995. A world-monograph of the genus Pseudombrophila (Pezizales, Ascomycotina). Libri Botanici 14. München: IHW-Verlag. 117 p. Bruns TD, Fogel R, White TJ, Palmer JD. 1989. Accelerated evolution of a false-truffle from a mushroom ancestor. Nature 339:140–142. Cain RF. 1956a. Studies on Coprophilous Ascomycetes II. Phaeotrichum, a new cleistocarpous genus in a new family, and its relationships. Canad J Bot 34:675–684. ———. 1956b. Studies on Coprophilous Ascomycetes IV. Tripterospora, a new cleistocarpous genus in a new family. Canad J Bot 34:699–710. ———. 1961. Studies of Coprophilous Ascomycetes VII. Preussia. Canad J Bot 39:1633–1649. Campbell C, Rodgers JD, Murray TD. 1991. A psychrophilic Orbicula associated with oat kernels. Mycologist 5:113–114. Cooke MC. 1871. Handbook of British fungi 2:926. Dennis RWG. 1981. British Ascomycetes. 4th ed. Vaduz: Cramer. 585 p. Dissing H, Schumacher T. 1994. Pezizales. In: Hawksworth DL. ed. Ascomycete systematics: problems and perspec- tives in the nineties. NATO ASI Series 269. New York: Plenum Press. 453 p. Eriksson OE, ed. 1999. Outline of Ascomycota—1999. Myconet 3:1–88. ———, ed. 2005. Outline of Ascomycota 2005. Myconet. http://www.umu.se/myconet/Myconet.html ———, Baral H-O, Currah RS, Hansen K, Kurtzman CP, Rambold G, Laessøe T, eds. 2001. Outline of Ascomycota—2001. Myconet 7:1–88. ———, Winka K. 1997. Supraordinal taxa of Ascomycota. Myconet 1:1–16. Hansen K, Læssøe T, Pfister DH. 2001. Phylogenetics of the Pezizaceae, with an emphasis on Peziza. ———, ———, ———. 2002. Phylogenetic diversity in the core group of Peziza inferred from ITS sequences and morphology. Mycol Res 106:879–902. ———, LoBuglio KF, Pfister DH. 2005. Evolutionary relationships of the cup-fungus genus Peziza and Pezizaceae inferred from multiple nuclear genes: PRB2, b-tubulin, and LSU rDNA. Mol Phylogenet Evol 36:1–23. Huelsenbeck JP, Ronquist F. 2001. MRBAYES: Bayesian inference of phylogeny. Bioinformatics 17:754–755. Hughes SJ. 1951. Studies on micro-fungi. VIII Orbicula and Lilliputia. Mycological Papers No. 42: 1–27. Janex-Favre MC, Locquin-Linard M. 1979. Le développement et la structure des ascocarpes du Lasiobolidium orbiculoides Malloch et Benny (Ascomycete périsporié et plectascé). Revue De Mycologie 43:373–391. Jeng RS, Krug JC. 1976. Coprotiella, a new cleistothecial genus of the Pyronemataceae with ascospores possessing de Bary bubbles. Mycotaxon 4:545–550. ———, Krug JC. 1977. New records and new species of coprophilous Pezizales from Argentina and Venezuela. Cand J Bot 55:2987–3000. Kimbrough JW. 1981. Cytology, ultrastructure, and taxonomy of Thelebolus (Ascomycetes). Mycologia 73:1–27. Kishino H, Hasegawa M. 1989. Evaluation of the maximum likelihood estimate of the evolutionary tree topologies from DNA sequence data, and the branching order in Hominoidea. J Mol Evol 29:170–179. Korf RP. 1972. Synoptic key to the genera of the Pezizales. Mycologia 64:937–994. Landvik S, Egger K, Schumacher T. 1997. Towards a subordinal classification of the Pezizales (Ascomycota): phylogenetic analyses of SSU rDNA sequences. Nordic Journal of Botany 17:403–418. ———, Kristiansen R, Schumacher T. 1998. Phylogenetic and structural studies in the Thelebolaceae (Ascomycota). Mycoscience 39:49–56. LoBuglio KF, Berbee ML, Taylor JW. 1996. Phylogenetic origins of the asexual mycorrhizal symbiont Cenococcum geophilum Fr. and other mycorrhizal fungi among the ascomycetes. Locquin MV. 1974. Synopsis generalis fungorum. In: De taxia fungorum I. Syllabus. Paris. Maddison DR, Maddison WP. 2002. MacClade: Analysis of phylogeny and character evolution. Version 4. Sinauer Associates, Sunderland, Massachusetts. HANSEN ET AL: ORIGINS OF TWO CLEISTOTHECIAL FUNGI Malloch D. 1979. Plectomycetes and their anamorphs. In: Kendrick [W]B. The whole fungus 1:153–165. ———. 1981. The Plectomycete Centrum. In: Reynolds DR. ed. Ascomycete systematics: the Luttrellian concept. New York: Springer Verlag. 241 p. ———, Benny GL. 1973. California Ascomycetes: Four new species and a new record. Mycologia 65:648–660. ———, Cain RF. 1971. Four new genera of cleistothecial Ascomycetes with hyaline ascospores. Canad J Bot 49:847–854. Miller SL, Torres P, McClean TM. 1994. Persistence of basidiospores and sclerotia of ectomycorrhizal fungi and Morchella in soil. Mycologia 86:89–95. Momol EA, Kimbrough JW, Eriksson OE. 1996. Phylogenetic relationships of Thelebolus indicated by 18S rDNA sequence analyses. Systema Ascomycetum 14:91–100. Moncalvo J-M, Lutzoni FM, Rehner SA, Johnson J, Vilgalys R. 2000. Phylogenetic relationships of agaric fungi based on nuclear large subunit ribosomal DNA sequences. Syst Biol 49:278–305. Moustafa AF, Ezz-Eldin EK. 1989. Lasiobolidium aegyptiacum, a new ascomycete from Egyptian soils. Mycol Res 92:376–378. Nylander JAA. 2004. MrModeltest 2.2. Program distributed by the author. Evolutionary Biology Centre, Uppsala University. O’Donnell K, Cigelnik E, Weber NS, Trappe JM. 1997. Phylogenetic relationships among ascomycetous truffles and the true and false morels inferred from 18S and 28S ribosomal DNA sequence analysis. Mycologia 89:48–65. Percudani R, Trevisi A, Zambonelli A, Ottonello S. 1999. Molecular phylogeny of truffles (Pezizales: Terfeziaceae, Tuberaceae) derived from nuclear rDNA sequence analysis. Mol Phylogenet Evol 13:169–180. 1033 Pfister DH. 1984. On Fimaria dentata, a new combination, with a review of synonyms and comments on Fimaria (Pezizales) Mycologia 76:843–852. Rambaut A. 2002. Se-Al. Sequence Alignment Editor. Version 2.0 alpha 11. University of Oxford, Oxford. http://evolve.zoo.ox.ac.uk/Se-Al/Se-Al.html Rehner SA, Samuels GJ. 1995. Molecular systematics of the Hypocreales: a teleomorph gene phylogeny and the status of their anamorphs. Can J Bot 73:816–823. Samuelson DA, Kimbrough JW. 1978. Asci of the Pezizales. IV. The apical apparatus of Thelebolus. Botanical Gazette 139:346–361. Suh S-O, Blackwell M. 1999. Molecular phylogeny of the cleistothecial fungi placed in Cephalothecaceae and Pseudeurotiaceae. Mycologia 91:836–848. Swofford DL. 2002. PAUP*. Phylogenetic analysis using parsimony (*and other methods). Version 4. Sunderland, Massachusetts: Sinauer Associates. Thiers HD. 1984. The secotioid syndrome. Mycologia 76:1–8. Trappe JM. 1979. The orders, families, and genera of hypogeous Ascomycotina (truffles and their relatives). Mycotaxon 9:297–340. Udagawa S, Furuya K. 1972. Notes on some Japanese Ascomycetes X. Trans Mycol Soc Japan 13:49–56. White TJ, Bruns TD, Lee SB, Taylor JW. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis N, Gelfand D, Sninsky J, White TJ, eds. PCR protocols: A guide to methods and applications. New York: Academic Press. p 315– 322. Yaguchi T, Someya A, Udagawa S. 1996. A new species of Lasiobolidium from Californian soil. Mycoscience 37:255–259.