Unravelling the phylogenetic relationships of lichenised fungi in Dothideomyceta.

We present a revised phylogeny of lichenised Dothideomyceta (Arthoniomycetes and Dothideomycetes) based on a combined data set of nuclear large subunit (nuLSU) and mitochondrial small subunit (mtSSU) rDNA data. Dothideomyceta is supported as monophyletic with monophyletic classes Arthoniomycetes and Dothideomycetes; the latter, however, lacking support in this study. The phylogeny of lichenised Arthoniomycetes supports the current division into three families: Chrysothrichaceae (Chrysothrix), Arthoniaceae (Arthonia s. l., Cryptothecia, Herpothallon), and Roccellaceae (Chiodecton, Combea, Dendrographa, Dichosporidium, Enterographa, Erythrodecton, Lecanactis, Opegrapha, Roccella, Roccellographa, Schismatomma, Simonyella). The widespread and common Arthonia caesia is strongly supported as a (non-pigmented) member of Chrysothrix. Monoblastiaceae, Strigulaceae, and Trypetheliaceae are recovered as unrelated, monophyletic clades within Dothideomycetes. Also, the genera Arthopyrenia (Arthopyreniaceae) and Cystocoleus and Racodium (Capnodiales) are confirmed as Dothideomycetes but unrelated to each other. Mycomicrothelia is shown to be unrelated to Arthopyrenia s.str., but is supported as a monophyletic clade sister to Trypetheliaceae, which is supported by hamathecium characters. The generic concept in several groups is in need of revision, as indicated by non-monophyly of genera, such as Arthonia, Astrothelium, Cryptothecia, Cryptothelium, Enterographa, Opegrapha, and Trypethelium in our analyses.

INTRODUCTION

Mutualism is one of the three main modes of nutrition within Ascomycota, besides saprotrophism and parasitism. A large number of mutualistic ascomycetes form symbiotic relationships with algae and/or cyanobacteria, so-called lichens. Of the 64 000 species currently accepted in Ascomycota (Kirk ), about almost 30 % (17 600) are lichen-forming fungi (Feuerer & Hawksworth 2007, Kirk ). Lichenised fungi differ from all other fungi in the formation of complex, persistent vegetative thalli, which makes them a prime subject for evolutionary studies.

It was long believed that lichens evolved several times independently within Ascomycota (and Basidiomycota), an idea supported by the first molecular study testing this hypothesis (Gargas ). Lutzoni et al. (2001, 2004) were unable to conclusively determine whether there were multiple gains of lichenisation or whether an initial lichenisation event occurred deep within Ascomycota, however, Lutzoni et al. (2001) found some Eurotiomycetes to be secondarily de-lichenised. This is particularly intriguing as Eurotiomycetes includes economically important fungi in the genera Aspergillus and Penicillium that feature a complex secondary chemistry similar to that found in lichens produced by homologous polyketide synthase genes (Grube & Blaha 2003, Kroken , Schmitt , Schmitt & Lumbsch 2009).

Since then, the phylogeny and classification of Ascomycota has further advanced (Lindemuth , Lumbsch et al. 2001, 2002a, b, 2004, Grube , Lücking , Lutzoni , Persoh , Wedin , del Prado , Miadlikoswka et al. 2006, Schmitt , Spatafora , Hibbett , Hofstetter , Lumbsch & Huhndorf 2007a, Schoch et al. 2006, 2009a, b, c). Our current understanding suggests that there were several lichenisation events but also some major delichenisation events during the evolution of Ascomycota (Gargas , Lutzoni , Liu & Hall 2004, Gueidan , Schoch ). The largest clade of lichenised fungi, Lecanoromycetes, with 14 000 accepted species, appears to be the result of a single lichenisation event with at least one major delichenisation event in Ostropales and several delichenisation events throughout the class (Lumbsch , Persoh , Wedin , Miadlikoswka et al. 2006, Hofstetter , Schoch , Baloch et al. in prep.). A similar pattern is suggested within the second largest lichenised clade, Arthoniomycetes, with about 1 500 species (Tehler 1995, Myllys , Sundin 2000, Tehler & Irestedt 2007, Ertz et al. 2008). This class was recently shown to include the mazaediate genus Tylophoron (Lumbsch ), previously considered to be related to pyrenocarpous lichens (Aptroot ). Arthoniomycetes is composed primarily of lichenised fungi producing apothecia or apothecioid ascomata with partially ascolocular development and bitunicate asci (Henssen & Jahns 1974, Eriksson & Winka 1997). The base of this clade was reconstructed as lichenised (Schoch ) and it is presumed that non-lichenised and lichenicolous species within the class represent reversions to the unlichenised state. One family that has not yet been confirmed within Arthoniomycetes using molecular data is Chrysothrichaceae, a small family of two genera (Byssocaulon, Chrysothrix) and little over 20 species (Kirk ). The third primarily lichenised class is Lichinomycetes (350 species).

The remaining lichenised fungi are primarily restricted to Dothideomycetes and Eurotiomycetes (subclass Chaetothyriomycetidae). Gueidan et al. (2008) demonstrated that lichenisation may have evolved at least twice within Eurotiomycetes (once at base of Verrucariales and once at base of Pyrenulales), though, this is uncertain as the ancestral state of the common ancestor to Pyrenulales, Verrucariales and Chaetothyriales, is not unambiguously resolved (Gueidan , Schoch ). Within both Verrucariales and Pyrenulales, there appears to be at least one loss of lichenisation each. Dothideomycetes and Arthoniomycetes together form the rankless clade Dothideomyceta, a name introduced by Schoch et al. (2009a, b). The ancestral state of Dothideomyceta and Dothideomycetes nodes are not resolved with confidence (Gueidan , Schoch et al. 2009a, b). In this paper we do not aim to resolve this issue but rather attempt to clarify, confirm or reject the placement of lichenised lineages within Dothideomyceta, specifically Dothideomycetes.

The following families have been confirmed or are believed to belong in either Chaeothyriomycetidae or Dothideomycetes: Verrucariaceae (930 species), Pyrenulaceae (280 species), Celotheliaceae (eight species), Microtheliopsidaceae (three species), and Pyrenothrichaceae (three species) in Chaetothyriomycetidae (Herrera-Campos , del Prado , Lücking 2008), and Trypetheliaceae (200 species), Monoblastiaceae (130 species), Strigulaceae (120 species), and Arthopyreniaceae (120 species) in Dothideomycetes (Lutzoni , del Prado , Lumbsch & Huhndorf 2007b). Most of these families have traditionally been placed within Pyrenulales (Poelt 1973, Henssen & Jahns 1974, Hafellner 1986, Kirk et al. 2001, Eriksson , Cannon & Kirk 2007), and much of the confusion regarding previous classifications of these pyrenocarpous lichens stems from the fact that Pyrenulales were at some point considered synonymous with the ascolocular Melanommatales (currently regarded synonymous with Pleosporales; Barr 1980, Harris 1984, 1990, 1991, 1995), whereas other workers considered Pyrenulales to be ascohymenial (Henssen & Jahns 1974). The fact that Trypetheliaceae have no close relative within Dothideomycetes was reflected in the establishment of a separate order, Trypetheliales (Aptroot ).

In addition to the aforementioned families, there are several genera of uncertain position, such as Cystocoleus and Racodium, both of which belong in Capnodiales/Dothideomycetes (Muggia et al. 2007), as well as Julella, Mycoporum, Collemopsidium (Pyrenocollema), and others, of unconfirmed affinities (Harris 1995). Yet other lineages, such as the recently discovered Eremithallus (Lücking ) or the genera Thelocarpon and Vezdaea (Reeb , Lumbsch ) appear to fall outside the currently accepted classes known to contain lichen-forming fungi. The current phylogeny of Chaetothyriomycetidae suggests that the two large lichen-forming families in this subclass may have emerged from distinct lichenisation events, however, this could not be resolved with confidence (see node 18 in fig. 1 and table 1 of Gueidan , Schoch ). It thus appears that Dothideomycetes, the largest class of Ascomycota with an estimated number of 19 000 species (Kirk ), a class that has largely been neglected when assessing the phylogeny of lichenised fungi, might be the only class within Ascomycota containing several lineages that evolved through independent lichenisation. In addition to Trypetheliaceae, at least two other families, which exhibit substantial radiation accompanied with morphological variation at the generic and species level (Monoblastiaceae and Strigulaceae) have been suggested to belong to Dothideomycetes. The only sequenced species of Strigula has been suggested to belong to Eurotiomycetes (Schmitt ); however, re-examination of the specimen used in this study showed that it belonged in Verrucariaceae. Therefore the phylogenetic position of Strigulaceae remains unresolved. In addition, Anisomeridium polypori (Monoblastiaceae) was suggested to belong to Dothideomycetes (James ).

In this paper, we are using nuclear large subunit (nuLSU) and mitochondrial small subunit (mtSSU) rDNA data, to construct a phylogeny of lichenised fungi with bitunicate asci, focusing on Dothideomyceta. We also present novel data that require adjustments in the systematic classification of taxa within both classes. A further objective was to begin to examine generic concepts within the family Trypetheliaceae, which is comprised of 11 genera (Lumbsch & Huhndorf 2007b) and approximately 200 species (Harris 1984, Aptroot 1991b, del Prado ).

MATERIAL AND METHODS

Taxon sampling

Representatives of lichenised Dothideomyceta taxa were obtained through recent field work in the U.S.A., Central and South America, Europe, India, Thailand, and Fiji. Newly generated sequences were supplemented with other lichenised and non-lichenised Dothideomyceta from GenBank plus additional taxa in Pezizomycetes, Leotiomycetes, Sordariomycetes, Eurotiomycetes, and Lecanoromycetes, chiefly from a previous alignment published by Schoch et al. (2009a). In total, we analysed 162 operational taxonomic units (OTUs) representing 152 species and 111 genera. All OTUs included in the analyses, along with GenBank accession numbers and collection information for newly sequenced samples, are listed in Table 1 - see online Supplementary Information.

Table 1.

Taxa included in this study with GenBank accession numbers and collection information. Numbers following taxon names are DNA identification numbers used in this study.

Taxon	Collection	Accession Number
		nuLSU	mtSSU
Acrocordia subglobosa (HTL940)	Palice s.n., Poland (F)		GU327681
Amphisphaeria umbrina		FJ176863	FJ713609
Anisomeridium ubianum (94)	Lumbsch 19845j, Fiji (F)	GU327709	GU327682
Aptrootia terricola			DQ328995
Arthonia caesia		FJ469668	FJ469671
Arthonia didyma		EU704083	EU704047
Arthonia dispersa		AY571381	AY571383
Arthonia radiate			EU704048
Arthonia ruana (79B)	Zimmerman 1117, Germany (F)		GU327683
Arthonia rubrocincta (129)	Nelsen 4010, U.S.A. (F)		GU327684
Arthopyrenia salicis		AY538339	AY538345
		AY607730	AY607742
Ascobolus crenulatus		AY544678	FJ713607
Astrothelium cinnamomeum		AY584652	AY584632
Astrothelium confusum (98)	Nelsen 4004a, Peru (F)	GU327710	GU327685
Bacidia schweinitzii		DQ782911	DQ972998
Bathelium degenerans			DQ328987
			DQ328988
Bimuria novae-zelandiae		AY016356	FJ190605
Bionectria ochroleuca		AY489716	FJ713619
Botryosphaeria dothidea		DQ678051	FJ190612
Botryosphaeria stevensii		DQ678064
Botryosphaeria tsugae		DQ767655
Botryotinia fuckeliana		AY544651	AY544732
Caliciopsis orientalis		DQ470987	FJ190654
Caliciopsis pinea		DQ678097	FJ190653
Camarops ustulinoides		DQ470941	FJ190588
Capnodium coffeae		DQ247800	FJ190609
Capronia pilosella		DQ823099	FJ225725
Cercospora beticola		DQ678091	FJ190647
Cheilymenia stercorea		AY544661	AY544733
Chiodecton natalense		EU704085	EU704051
Chlorociboria aeruginosa		AY544669	AY544734
Chrysothrix flavovirens (L466)	Perlmutter 786, U.S.A. (NCU)	GU327711	GU327686
Chrysothrix xanthina (126)	Nelsen 4005, U.S.A. (F)	GU327712	GU327687
Cladosporium cladosporioides		DQ678057	FJ190628
Cochliobolus heterostrophus		AY544645	AY544737
Cochliobolus sativus		DQ678045	FJ190589
Columnosphaeria fagi		DQ470956	FJ713608
Combea mollusca		AY571382	AY571384
Coniothyrium palmarum		DQ767653	FJ190638
Cordyceps capitata		AY489721	FJ713628
Cryptothecia assimilis (86B)	Lumbsch 19815l, Fiji (F)		GU327688
Cryptothecia candida			EU704052
Cryptothelium amazonum (47)	Nelsen 4000a, Peru (F)	GU327713	GU327689
Cryptothelium cecidiogenum			DQ328991
Cryptothelium sepultum (63C)	Nelsen 4001a, Peru (F)	GU327714	GU327690
Cudoniella cf. clavus		DQ470944	FJ713604
Cystocoleus ebeneus		EU048578	EU048584
		EU048579	EU048585
		EU048580	EU048586
			EU048587
Delitschia winteri		DQ678077	FJ190644
Dendrographa alectoroides (100)	Lumbsch 19914g, U.S.A. (F)	GU327715	GU327691
Dendrographa leucophaea f. minor		AF279382	AY548811
Dendryphiella arenaria		DQ470971	FJ190617
Dermatocarpon miniatum		AY584644	AY584616
Diaporthe eres		AF408350	FJ190607
Dichosporidium boschianum (89B)	Lumbsch 19815a, Fiji (F)	GU327716	GU327692
Dirina catalinariae		EF081387
Dothidea insculpta		DQ247802	FJ190602
Dothidea sambuci		AY544681	AY544739
Dothiora cannabinae		DQ470984	FJ190636
Eleutherascus lectardii		DQ470966	FJ190606
Elsinoe centrolobi		DQ678094	FJ190651
Elsinoe phaseoli		DQ678095	FJ190652
Elsinoe veneta		DQ767658	FJ190650
Endocarpon pallidulum		DQ823097	FJ225674
Enterographa anguinella		EU704086	EU704054
Enterographa crassa		EU704088	EU704056
Erythrodecton granulatum		EU704090	EU704058
Eupenicillium javanicum		EF413621	FJ225778
Exophiala salmonis		EF413609	FJ225745
Flavobathelium epiphyllum (67)	Lücking s.n. Panama (F)	GU327717
Glomerella cingulata		AF543786	FJ190626
Glyphium elatum		AF346420	AF346425
Gnomonia gnomon		AF408361	FJ190615
Guignardia gaulteriae		DQ678089	FJ190646
Herpothallon rubrocinctum (128)	Nelsen 4006, U.S.A. (F)		GU327693
Herpotrichia diffusa		DQ678071	DQ384076
Hypocrea lutea		AF543791	FJ713620
Hysteropatella cf. elliptica		DQ767657	FJ190649
Kirschsteiniothelia aethiops		AY016361	FJ190604
		DQ678046	FJ190590
Lachnum virgineum		AY544646	AY544745
Laurera megasperma		FJ267702
Lecanactis abietina		AY548812	AY548813
Lecanactis sp.		EU704091	EU704059
Lecanora hybocarpa		DQ782910	DQ912273
Macrophomina phaseolina		DQ678088	FJ190645
Megalotremis verrucosa (104)	Lücking 26316, Colombia (F)	GU327718	GU327694
Monilinia laxa		AY544670	AY544748
Mycomicrothelia hemispherica (102)	Lücking 28641, Nicaragua (F)	GU327719	GU327695
Mycomicrothelia miculiformis (101B)	Lücking 28637, Nicaragua (F)	GU327720	GU327696
Mycomicrothelia obovata (95)	Nelsen 4007a, Peru (F)	GU327721	GU327697
Mycosphaerella fijiensis		DQ678098	FJ190656
Mycosphaerella punctiformis		DQ470968	FJ190611
Myriangium duriaei		DQ678059	AY571389
Nectria cinnabarina		U00748	FJ713622
Opegrapha celtidicola		EU704094	EU704066
Opegrapha filicina		EU704095	EU704067
Opegrapha lithyrga		EU704096	EU704068
Opegrapha varia		EU704103	EU704075
Ophionectria trichospora		AF543790	FJ713626
Peltigera degenii		AY584657	AY584628
Penicillium freii		AY640958	AY584712
Pertusaria dactylina		DQ782907	DQ972973
Phaeotrichum benjaminii		AY004340	AY538349
Phoma herbarum		DQ678066	FJ190640
Phyllobathelium anomalum (242)	Lücking s.n., Panama (F)	GU327722	GU327698
Phyllobathelium firmum (HTL3175)	Lücking s.n., Panama (F)	GU327723
Pleospora herbarum var. herbarum		DQ247804	FJ190610
Preussia terricola		AY544686	AY544754
Pseudopyrenula subgregaria (106)	Lücking 24079, Thailand (F)	GU327724	GU327699
Pseudopyrenula subnudata			DQ328997
Pyrenophora phaeocomes		DQ499596	FJ190591
Pyrenophora tritici-repentis		AY544672	FJ713605
Pyrenula pseudobufonia		AY640962	AY584720
Pyrgillus javanicus		DQ823103	FJ225774
Pyxine subcinerea		DQ883802	DQ912292
Racodium rupestre		EU048583	EU048588
		EU048581
		EU048582	EU048589
Ramichloridium anceps		DQ823102	FJ225752
Roccella canariensis		AY779328
Roccella fuciformis		AY584654	EU704082
Roccella montagnei (109)	Lumbsch 19700a, India (F)	GU327725	GU327700
Roccella tuberculata		AY779328
Roccellographa cretacea		DQ883696	FJ772240
Schismatomma decolorans		AY548815	AY548816
Schismatomma pericleum		AF279408	AY571390
Scorias spongiosa		DQ678075	FJ190643
Scutellinia scutellata		DQ247806	FJ190587
Simonyella variegate			AY584631
Sphinctrina turbinate		EF413632	FJ713611
Spiromastix warcupii		DQ782909	FJ225794
Sporormiella minima		DQ678056	FJ190624
Staurothele frustulenta		DQ823098	FJ225702
Strigula nemathora (72)	Lücking s.n., Costa Rica (F)		GU327701
Strigula schizospora (73)	Lücking s.n., Costa Rica (F)		GU327702
Stylodothis puccinioides		AY004342	AF346428
Sydowia polyspora		DQ678058	FJ190631
Syncesia farinacea		EF081452
Trematosphaeria heterospora		AY016369	AF346429
Trematosphaeria pertusa		DQ678072	FJ190641
Trimmatostroma abietis		DQ678092	FJ190648
Trypetheliopsis kalbii (243)	Lücking s.n., Panama (F)		GU327703
Trypethelium eluteriae			DQ328989
Trypethelium eluteriae (111)	Lumbsch 19701a, India (F)	GU327726	GU327704
Trypethelium marcidum			DQ329007
Trypethelium marcidum (132)	Nelsen 4008, U.S.A. (F)	GU327727	GU327705
Trypethelium nitidiusculum (139)	Nelsen 4002a, U.S.A. (F)	GU327728	GU327706
Trypethelium papulosum (97)	Nelsen 4009a, Peru (F)	GU327729	GU327707
Trypethelium platystomum			DQ329009
Trypethelium tropicum (25)	Nelsen 4003, Thailand (F)	GU327730	GU327708
Tubeufia cerea		DQ470982	FJ190634
Tylophoron crassiusculum		EU670258
Tylophoron moderatum		EU670256
Tyrannosorus pinicola		DQ470974	FJ190620
Vibrissea truncorum		FJ176874	FJ190635
Westerdykella cylindrical		AY004343	AF346430
Xylaria hypoxylon		AY544648	AY544760

Molecular methods

The Sigma REDExtract-N-Amp Plant PCR Kit (St. Louis, Missouri, U.S.A.) was used to isolate DNA, following the manufacturer's instructions, except only 10 μL of extraction buffer and 10 μL dilution buffer were used, following Avis et al. (2003). Dilutions of these extractions (rather than the stock DNA solution) were found to work best for PCR (C. Andrew, pers. comm. 2009), and a 20× DNA dilution was then used in subsequent PCR reactions.

Samples were PCR amplified and/or sequenced using the mrSSU1, mrSSU2, mrSSU2r and mrSSU3r primers (Zoller ) for the mitochondrial small subunit (mtSSU) and the AL2R (Mangold ), LR3R, LR3, LR5, LR6, LR7 (Vilgalys & Hester 1990) primers for the nuclear ribosomal large subunit rDNA (nuLSU). The 10 μL PCR reactions consisted of 5 μM of each PCR primer, 3 mM of each dNTP, 2 μL of 10 mg/mL 100x BSA (New England BioLabs, Ipswich, Massachusetts, U.S.A.), 1.5 μL 10× PCR buffer (Roche Applied Science, Indianapolis, Indiana, U.S.A.), 0.5 μL Taq, approximately 2 μL diluted DNA, and 2 μL water. The PCR cycling conditions were as follows: 95 °C for 5 min, followed by 35 cycles of 95 °C for 1 min, a locus-specific annealing temperature for 1 min, and 72 °C for 1 min, followed by a single 72 °C final extension for 7 min. An annealing temperature of 53 °C was used for mtSSU, while 57 °C was used for nrLSU.

Samples were visualised on a 1 % ethidium bromide-stained agarose gel under UV light and bands were gel extracted, heated at 70 °C for 5 min, cooled to 45 °C for 10 min, treated with 1 μL GELase (Epicentre Biotechnologies, Madison, WI, U.S.A.) and incubated at 45 °C for at least 24 h. The 10 μL cycle sequencing reactions consisted of 1–1.5 μL of Big Dye v. 3.1 (Perkin-Elmer Applied Biosystems, Foster City, California, U.S.A.), 2.5–3 μL of Big Dye buffer, 6 μM primer, 0.75–2 μL Gelased PCR product and water. The cycle sequencing conditions were as follows: 96 °C for 1 min, followed by 25 cycles of 96 °C for 10 s, 50 °C for 5 s and 60 °C for 4 min. Samples were precipitated and sequenced in an Applied Biosystems 3730 DNA Analyser (Foster City, California, U.S.A.), and sequences assembled in Sequencher 4.9 (Gene Codes Corporation, Ann Arbor, Michigan, U.S.A.).

Phylogenetic analysis

The alignment of Schoch et al. (2009a) was used as a starting point, from which a large number of sequences were removed. Newly generated sequences were added and manually aligned (nuLSU), or were separately aligned, added to the Schoch et al. (2009a) alignment, and manually adjusted (mtSSU). In addition to a representative set of dothideomycetous fungi, members of several Ascomycota classes were retained and Pezizomycetes taxa were used as the outgroup. The entire set of sequences generated in the present study plus those from GenBank were aligned in Se-Al v. 2.0a11 (Rambaut 1996) and BioEdit 7.0.9 (Hall 1999). An iterative procedure was used for the nuLSU in which ambiguous regions were aligned with Muscle 3.6 (Edgar 2004) through Mesquite 2.71 (Maddison & Maddison 2009); the alignment was again manually refined and other portions realigned with Muscle. After a final manual refinement, ambiguous regions and introns were removed and the alignment was deposited in TreeBase.

Alignments for each gene were concatenated in Mesquite 2.71 (Maddison & Maddison 2009) and analysed under the maximum likelihood (ML) optimality criterion in RAxML 7.0.4 (Stamatakis 2006). The data set was partitioned by locus and the GTRMIXI model with twenty-five rate parameter categories (default) was used for each partition. In addition, support was estimated by performing 1000 bootstrap replicates, and clades with bootstrap support of 70 % or greater were considered strongly supported. Additionally, the data sets were analyzed in GARLI 0.96 (Zwickl 2006) using the GTR-gamma-invariant model which is similar to the model used in RAxML.

RESULTS

The final alignment consisted of 1 915 unambiguously aligned characters (1 199: nuLSU; 716: mtSSU). Both ML analyses recovered the major class-level ingroup nodes (Fig. 1) corresponding to other recent studies (Leotiomycetes, Sordariomycetes, Eurotiomycetes, Lecanoromycetes, Arthoniomycetes, Dothideomycetes). Arthoniomycetes and Dothideomycetes form a strongly supported sister-group relationship, corresponding to Dothideomyceta. Individual gene phylogenies suggested some incongruence between loci (unpubl. data), however, the topology in the combined analysis is in agreement with previously reported phylogenies and we did not exclude taxa.

Fig.1.

The ML tree from RAxML maximum likelihood analysis with bootstrap percentages equal to or greater than 70 are plotted above or below branches. Lichenised taxa are in green, while non-lichenised taxa are in black.

The phylogeny of Arthoniomycetes (Arthoniales) largely confirmed previous analyses, with Chrysothrichaceae forming an additional family within this clade (Fig. 1). Arthoniaceae s. l. and Roccellaceae s. l. are both monophyletic and well separated. However, several smaller lineages that eventually could be reinstated at the family level show strong support: Arthoniaceae s. str., Cryptotheciaceae (Cryptothecia-Herpothallon), the Tylophoron clade, Roccellaceae s. str., Opegraphaceae s. str., and possibly Chiodectonaceae (as Chiodecton sphaerale is closely related to Erythrodecton and Dichosporidium whereas the sequenced C. natalense is apparently not a Chiodecton s. str.). Surprisingly, Arthonia caesia clustered with Chrysothrichaceae and not Arthoniaceae. Herpothallon rubrocinctum is nested within Cryptothecia s. l.

Six distinct, lichenised lineages were confirmed as belonging to Dothideomycetes (Fig. 1): the order Trypetheliales, the families Arthopyreniaceae, Monoblastiaceae, and Strigulaceae, and the genera Cystocoleus and Racodium. The latter two (Cystocoleus and Racodium) are members of the order Capnodiales, whereas Arthopyreniaceae, represented by the species Arthopyrenia salicis, was confirmed as clustering within Pleosporales. However, Arthopyreniaceae as currently defined, including the genera Julella (not sequenced) and Mycomicrothelia, is not monophyletic, as the sequenced species of Mycomicrothelia appeared outside Pleosporales and form a sister-group to Trypetheliaceae.

Strigulaceae is represented by five samples of the three genera Flavobathelium, Phyllobathelium, and Strigula, which formed a supported monophyletic clade sister to Kirschsteiniothelia aethiops, but without support. Monoblastiaceae was strongly supported and included four genera with one species each in this analysis: Acrocordia subglobosa, Anisomeridium ubianum, Megalotremis verrucosa, and Trypetheliopsis (syn. Musaespora) kalbii. Initially we also included a GenBank sequence of Anisomeridium polypori in the data set, but the nuLSU sequence was recovered in Eurotiomycetes and the taxon was excluded from the final analysis. It is possible that this sequence is derived from a contaminant or that it was confused with a similar species in an unrelated lineage.

Trypetheliaceae was strongly supported as monophyletic, being sister to the genus Mycomicrothelia. There was no support for the traditional separation into the perithecial and ascospore core genera Astrothelium, Laurera, and Trypethelium, as species of these genera were found scattered over the Trypetheliaceae clade.

DISCUSSION

This is the first molecular phylogenetic study that includes presumably all major lichenised lineages within Dothideomyceta. This rankless taxon was informally introduced by Schoch et al. (2009a, b) for the clade including Arthoniomycetes and Dothideomycetes. The sister group of Dothideomyceta is not yet resolved but Ruibal et al. (2009; this volume) demonstrated an unnamed lineage of melanised rock-inhabiting fungi to be basal to Arthoniomycetes (not included in our sampling).

Arthoniomycetes is the second largest class of primarily lichenised Ascomycota and exhibits considerable morphoanatomical variation (Fig. 2). The molecular phylogeny presented here confirms the current classification of lichenised Arthoniomycetes in three families: Arthoniaceae, Chrysothrichaceae, and Roccellaceae (Tehler 1995, Grube 1998, Tehler & Irestedt 2007). The morphological concept used to classify the single order included few large genera, with Arthonia and Opegrapha having the highest number of species (500 and 300, respectively). The infrageneric relationships of these species were repeatedly discussed and there was common agreement that these genera were not monophyletic and include morphologically distinct groups. Similarly the relationships of other genera with fewer species or of monospecific genera in the family Roccellaceae was unclear. Along with previous data (Tehler 1995, Myllys , Tehler & Irestedt 2007) and recent results by Ertz et al. (2009), the present tree is a further step to resolve these questions based on molecular data.

Fig. 2.

Select lichenised Arthoniomycetes. A. Chrysothrix xanthina; B. C. septemseptata; C. Arthonia caesia; D. A. cyanea; E. A. pulcherrima; F. A. rubrocincta; G. Cryptothecia candida; H. Herpothallon rubrocinctum; I. Tylophoron crassiusculum (teleomorph); J. T. crassiusculum (anamorph); K. Opegrapha filicina; L. O. astraea; M. Enterographa anguinella; N. Syncesia glyphysoides; O. S. byssina; P. Lecanactis epileuca; Q. Chiodecton sphaerale; R–S. Erythrodecton granulatum; T. Dichosporidium boschianum; U. D. nigrocinctum (ascomata); V. D. nigrocinctum (isidia); W. Mazosia rotula; X. Roccella spec. Photo credits: R. Lücking.

Little can be said regarding generic concepts of most genera, as the taxon sampling is still far too incomplete for this group, but it appears that some of the traditional concepts based on fruit body structure are not supported, which suggests some degree of parallel evolution. An example is the Chiodecton-Enterographa complex: while the sequenced Chiodecton natalense appears to be unrelated to the morphologically and anatomically similar Dichosporidium and Erythrodecton (Thor 1990), Enterographa and the similar Schismatomma (Sparrius 2004) were found in three different clades related to either Chiodecton natalense (Schismatomma), Dichosporidium (Enterographa crassa), and Opegrapha (Enterographa anguinella), respectively. This is in agreement with Ertz et al. (2009), who showed that Enterographa is not monophyletic and groups either with the core Opegrapha clade (here represented by O. lithyrgica), or with Chiodecton-like species (Dichosporidium and Erythrodecton). Consequently, Ertz et al. (2009) tranferred Enterographa anguinella to Opegrapha. Not surprisingly, neither Arthonia nor Opegrapha are monophyletic. Ertz et al. (2009) showed convincingly that despite different ascomatal structure, Opegrapha atra and O. calcarea (with distinct excipulum) are closely related to Arthonia radiata (lacking an excipulum), which is confirmed by similarities of ascus structure and pigment type. Subsequently, Ertz et al. (2009) suggested these two Opegrapha species be recognised as belonging to Arthonia. Opegrapha varia and O. celtidicola form another monophyletic lineage together with Simonyella variegata. Most likely this branch also includes other Opegrapha species, according to the results of Ertz et al. (2009). Opegrapha s. str. forms a further lineage including O. lithyrgica, which is closely related to the type species O. vulgata (Ertz ), the foliicolous O. filicina, as well as Combea mollusca and Roccellographa cretacea.

Herpothallon rubrocinctum is now confirmed as an ascomycete in Arthoniomycetes. This seems trivial as the species also morphologically shows clear affinities with Cryptothecia (Aptroot ), but the position of this taxon was questioned long ago and was even considered a basidiomycete (see discussion in Withrow & Ahmadjian 1983, Aptroot ). Our analysis shows Herpothallon nested within Cryptothecia, supporting the previous hypothesis that byssoid-isidiate species within this complex are indeed members of Cryptothecia rather than forming a separate genus, as proposed by Aptroot et al. (2008). However, a larger taxon sampling is needed to resolve the Cryptothecia-Herpothallon complex, especially considering that there are other genera such as Stirtonia involved and even further new genera have been segregated recently (Aptroot , Frisch & Thor 2010). The fruticose Roccella species form a clearly monophyletic branch together with several crustose species representing various genera; this assemblage of core Roccellaceae has already been recognised previously (Tehler 1995, Myllys , Tehler & Irestedt 2007). The placement of Tylophoron, a genus that has passive spore dispersal and was previously assigned to Caliciales, is here confirmed as a member of Arthoniaceae s. l., in agreement with Lumbsch et al. (2009a).

The strongly supported placement of Arthonia caesia within Chrysothrix is unexpected; however, fertile species of Chrysothrix are very similar to Arthonia in ascoma morphology and anatomy, and particularly A. caesia and allies can be easily perceived as non-pigmented species of Chrysothrix in apothecial anatomy and morphology and thallus structure (including the chlorococcoid photobiont). Similar Arthonia species include A. cupressina, which is closely related to A. caesia. Further studies are needed to elucidate which additional Arthonia taxa need to be placed in Chrysothrix. The latter genus was variously placed in its own family Chrysothrichaceae mainly due to the presence of pulvinic acids as secondary metabolites but also in Arthoniaceae due to similarities in ascus characters (Grube 1998). The present data strongly support Chrysothrichaceae as a separate family, especially as it is sister to all remaining Arthoniales and not to Arthoniaceae. It is therefore necessary to transfer Arthonia caesia (which lacks pulvinic acids) and related species to this family. The other Arthonia species sampled group form a fairly well supported monophyletic group, which includes a species formerly assigned to Arthothelium, i.e. Arthonia ruana, because of its muriform ascospores; however, it has been known for some time that most species with muriform ascospores are more closely related to Arthonia than to the type of Arthothelium, A. spectabile (Tehler 1990, Sundin & Tehler 1998, Cáceres 2007, Grube 2007), which has not yet been sequenced. Notably, Arthonia didyma and A. rubrocincta, two species with reddish pigments, form a weakly supported group. If future efforts confirm this grouping, the name Coniocarpon could be used for this clade (Cáceres 2007).

In contrast to Arthoniomycetes, the overwhelming majority of Dothideomycetes species are non-lichenised. In addition to Arthopyreniaceae, Trypetheliaceae and Cystocoleus and Racodium (Muggia et al. 2007), this study confirms the placement of Monoblastiaceae and Strigulaceae within Dothideomycetes. Although our support for the Dothideomycetes node is weak, the included non-lichenised taxa are well supported within this class in other studies (Schoch et al. 2006, 2009a, b); in addition, placement within Dothideomyceta is strongly supported. Both, Monoblastiaceae and Strigulaceae are comparatively large with over 100 accepted species each and show substantial morphological and ecological radiation (Fig. 3); both are chiefly tropical. The mostly corticolous Monoblastiaceae range from barely lichenised forms with exposed perithecia (many species of Anisomeridium) to taxa with well-developed, corticate thalli (Anisomeridium p.p., Megalotremis, Trypetheliopsis). Ascospores vary from small to large and thick-walled but are always simple or transversely septate only (Harris 1995). Substantial variation is found in the conidiomata, and many species, particularly in the genera Caprettia, Megalotremis, and Trypetheliopsis (= Musaespora) have developed unique pycnidia that in part are similar to campylidia or hyphophores found in certain Lecanoromycetes (Aptroot & Sipman 1993, Lücking , Aptroot , Lücking 2008). Secondary substances are few, including lichexanthone and anthraquinones. All species of Monoblastiaceae in which conidiomata are known share a particular synapomorphy: the conidia are always embedded in a strongly coherent, gelatinous matrix. Thus, besides the uniform hamathecium and ascus anatomy, there is substantial phenotypic evidence for monophyly of this family, now confirmed by molecular data.

Fig. 3.

Select lichenised Dothideomycetes; A. Arthopyrenia cinchonae; B. Mycomicrothelia modesta; C. Anisomeridium subprostans; D. Anisomeridium spec. (pycnidia); E. A. foliicola (pycnidia); F. Caprettia amazonensis (pycnidia); G. Megalotremis cauliflora (pycnidia); H. Trypetheliopsis (= Musaespora) coccinea (campylidia); I. Strigula viridiseda; J. S. laureriformis (pycnidia); K. S. smaragdula; L. Flavobathelium epiphyllum; M. Phyllobathelium firmum; N. P. leguminosae (pycnidia); O. Pseudopyrenula subnudata; P. Trypethelium tropicum; Q. T. platystomum; R. Bathelium degenerans; S. Laurera purpurina; T. Astrothelium cinnamomeum; U. A. eustomum; V. Trypethelium nitidiusculum; W. Laurera megasperma; X. Campylothelium spec. Photo credits: R. Lücking.

Strigulaceae share many characteristics with Monoblastiaceae, specifically the ascus type and the mostly 1- or 3-septate ascospores, although some species have muriform ascospores (Harris 1995, Aptroot , Lücking 2008). Species in this family are found on a variety of substrata, including rocks, bark, and living leaves. Poorly developed thalli are found in corticolous species with barely lichenised thalli and exposed perithecia (Strigula p.p.), whereas the genera Flavobathelium, Phyllobathelium, and Phyllocratera include taxa with well-developed, corticate thalli. Also in this family, the most characteristic synapomorphy are the conidia, which feature terminal gelatinous appendices (Harris 1995, Lücking 2008). Unfortunately, our taxon sampling of this family is poor but sufficient to confirm its monophyly and its placement in Dothideomycetes. This is the first molecule-based support for the inclusion of Phyllobatheliaceae within Strigulaceae, a concept first presented by Harris (1995).

The largest lichenised family within Dothideomycetes, Trypetheliaceae, contains members that are typically lichen-forming and tropical to subtropical in distribution, with some taxa extending into temperate regions (Aptroot 1991, Harris 1995, Brodo , Aptroot ). The species are almost exclusively corticolous, forming a crustose, endo- or epiperidermal thallus with algae belonging to Trentepohliaceae; however, Anisomeridium is often found lignicolous and Aptrootia grows on bryophytes. Detailed studies in Costa Rica suggest Trypetheliaceae to occur primarily on trunks and branches of trees in exposed habitats of lowland to lower montane (200–1000 m) rain and dry forests and savannas with rather distinct dry season (Aptroot , Rivas-Plata ). Trypetheliaceae species are quite variable in perithecial morphology (Fig. 3) but have a rather uniform hamathecium composed of thin, anastomosing pseudoparaphyses embedded in a stiff gelatinous matrix. The most characteristic synapomorphy are the usually hyaline ascospores with internal wall thickenings that cause more or less diamond-shaped septa, but these wall thickenings are often reduced or absent in species with multiseptate or muriform ascospores (Harris 1984, 1990, 1995, Aptroot 1991b, Aptroot ). The secondary chemistry is equally simple, with lichexanthone and pigments as most common substances, i.e. polyketide derived aromatic compounds produced through the acetyl-polymalonyl pathway (Elix & Stocker-Wörgötter 2008). However, the number of species with substances present is much higher in Trypetheliaceae than any other lineage within Dothideomycetes: more than 70 species are known to produce secondary substances in this family. The core genera Astrothelium, Campylothelium, Cryptothelium, Laurera, and Trypethelium, are separated primarily on the basis of perithecial arrangement and ostiolar orientation (solitary vs. aggregate, apical vs. excentric) and ascospore septation (transverse vs. muriform; Harris 1990, 1995, del Prado ). Because of the schematic classification, Harris (1995) suggested that these genera may be polyphyletic, and del Prado et al. (2006) subsequently illustrated the non-monophyly of Trypethelium. Aptroot et al. (2008) echoed Harris's (1995) sentiment and stated that generic concepts in Trypetheliaceae are in need of revision.

Surprisingly, Mycomicrothelia was recovered as sister to Trypetheliaceae. Mycomicrothelia has traditionally been considered a sister genus to Arthopyrenia with brown ascospores (Harris 1995). However, the hamathecium at least of the sequenced species is identical to that found in Trypetheliaceae, whereas Arthopyrenia has thicker and less branched and anastomosing pseudoparaphyses. Moreover, the ascospores are of a different type, often with internal wall thickenings. It remains to be tested whether Arthopyrenia and Mycomicrothelia in their current circumscriptions are monophyletic genera or whether at least some species currently assigned to these genera perhaps represent further lichenised lineages within Dothideomycetes. Whether Mycomicrothelia should be included within Trypetheliaceae or receive its own family rank is open to question. Mycomicrothelia has primarily thin-walled, dark brown ascospores, whereas in Trypetheliaceae they are primarily thick-walled with diamond-shaped lumina and hyaline (brown only in Aptrootia and Architrypethelium). Understanding the phylogenetic position of Polymeridium, which also has thin-walled ascospores, will hopefully help clarify this.

In spite of the many characters in parallel with Monoblastiaceae and Strigulaceae, also the Trypetheliaceae plus Mycomicrothelia (Trypetheliales) are quite unique genetically and there is no evidence that the three families would be related to each other or with Arthopyreniaceae. This supports the notion of several shifts in lichenisation within the Dothideomycetes (Aptroot 1991a, 1998). However, the often barely lichenised thalli in certain species of Anisomeridium, Arthopyrenia, Julella, Mycomicrothelia, Mycoporum, Pseudopyrenula, and Strigula (Aptroot 1991a, Aptroot 1998, Harris 1995) suggest that these species can possibly switch between being (almost) non-lichenised to distinctly lichenised, a situation also found in the unrelated genus Stictis within Lecanoromycetes (Wedin ).

The present study clarifies the systematic position of further pyrenocarpous lichenised lineages within the Ascomycota and shows that previous concepts in part diverged widely from our present understanding but also came suprisingly close even without molecular evidence (Table 2). This study emphasises that pyrenocarpous lichens with bitunicate asci are not only not monophyletic, but belong to at least two different classes (Dothideomycetes and Eurotiomycetes) and several different orders and families; the data at hand also suggest that these represent several independent lineages of lichenisation. Although we consider this study a contribution to clarify the systematic position of pyrenocarpous lichens and the evolution of lichenisation within Dothideomycetes, much remains to be done, considering that at present only a fraction of the presumably 600 species of lichens belonging in this class have been studied using DNA sequences. In particular, clarifying the generic and species concepts within Monoblastiaceae, Strigulaceae, and Trypetheliaceae, speciose families that are important elements of crustose lichen communities especially in the tropics, will be a major challenge in the near future.

Table 2.

Systematic placement of selected pyrenocarpous lichens according to different concepts.

Genus	Zahlbruckner 1926	Barr 1987	Harris 1995	current
Celothelium	Pyrenocarpeae	Loculoascomycetes	Loculoascomycetes	Eurotiomycetes
	(as Leptorhaphis)	Pleosporales	Melanommatales	Pyrenulales
	Pyrenulaceae	Pleosporaceae	Thelenellaceae	Celotheliaceae
Lithothelium	Pyrenocarpeae	Loculoascomycetes	Loculoascomycetes	Eurotiomycetes
	Astrotheliaceae	Melanommatales	Melanommatales	Pyrenulales
Pyrenula	Pyrenocarpeae	Pyrenulaceae	Pyrenulaceae	Pyrenulaceae
	Pyrenulaceae
Arthopyrenia	Pyrenocarpeae	Loculoascomycetes	Loculoascomycetes	Dothideomycetes
	Pyrenulaceae	Pleosporales	Pleosporales	Pleosporales
		Arthopyreniaceae	Pleosporaceae	Arthopyreniaceae
Acrocordia	Pyrenocarpeae	Loculoascomycetes	Loculoascomycetes	Dothideomycetes
Anisomeridium	(as Arthopyrenia)	Melanommatales	Melanommatales	incertae sedis
	Pyrenulaceae	Acrocordiaceae	Monoblastiaceae	Monoblastiaceae
Phyllobathelium	Pyrenocarpeae	Loculoascomycetes	Loculoascomycetes	Dothideomycetes
Strigula	Strigulaceae	Chaetothyriales	Melanommatales	incertae sedis
		Strigulaceae	Strigulaceae	Strigulaceae
Astrothelium	Pyrenocarpeae	Loculoascomycetes	Loculoascomycetes	Dothideomycetes
	Astrotheliaceae	Melanommatales	Melanommatales	Trypetheliales
Campylothelium	Pyrenocarpeae	Trypetheliaceae	Trypetheliaceae	Trypetheliaceae
	Paratheliaceae
Laurera	Pyrenocarpeae
	Trypetheliaceae
Pseudopyrenula	Pyrenocarpeae
	Pyrenulaceae
Trypethelium	Pyrenocarpeae
	Trypetheliaceae
Mycomicrothelia	Pyrenocarpeae	Loculoascomycetes	Loculoascomycetes	Dothideomycetes
	(as Microthelia)	Pleosporales	Pleosporales	Trypetheliales
	Strigulaceae	Arthopyreniaceae	Arthopyreniaceae	Trypetheliaceae?
Porina	Pyrenocarpeae		Hymenoascomycetes	Lecanoromycetes
	Pyrenulaceae		Trichotheliales	Ostropales
Trichothelium	Pyrenocarpeae		Trichotheliaceae	Porinaceae
	Strigulaceae	—