Volume 105, Issue 11 p. 1869-1887
Research Article
Free Access

Stable isotope analyses reveal previously unknown trophic mode diversity in the Hymenochaetales

Hailee B. Korotkin

Hailee B. Korotkin

Department of Ecology and Evolutionary Biology, University of Tennessee, 1416 Circle Drive, Knoxville, Tennessee, 37996 USA

Search for more papers by this author
Rachel A. Swenie

Rachel A. Swenie

Department of Ecology and Evolutionary Biology, University of Tennessee, 1416 Circle Drive, Knoxville, Tennessee, 37996 USA

Search for more papers by this author
Otto Miettinen

Otto Miettinen

Botanical Museum, Finnish Museum of Natural History, University of Helsinki, PO Box 7, FI-00014 Finland

Search for more papers by this author
Jessica M. Budke

Jessica M. Budke

Department of Ecology and Evolutionary Biology, University of Tennessee, 1416 Circle Drive, Knoxville, Tennessee, 37996 USA

Search for more papers by this author
Ko-Hsuan Chen

Ko-Hsuan Chen

Department of Biology, Duke University, Box 90338, Durham, North Carolina, 27708 USA

Search for more papers by this author
François Lutzoni

François Lutzoni

Department of Biology, Duke University, Box 90338, Durham, North Carolina, 27708 USA

Search for more papers by this author
Matthew E. Smith

Matthew E. Smith

Institute of Food and Agricultural Sciences, Plant Pathology, University of Florida, 2550 Hull Road, Gainesville, Florida, 32611 USA

Search for more papers by this author
P. Brandon Matheny

Corresponding Author

P. Brandon Matheny

Department of Ecology and Evolutionary Biology, University of Tennessee, 1416 Circle Drive, Knoxville, Tennessee, 37996 USA

Author for correspondence (e-mail: [email protected])Search for more papers by this author
First published: 26 October 2018
Citations: 17

Abstract

Premise of the Study

The Hymenochaetales are dominated by lignicolous saprotrophic fungi involved in wood decay. However, the group also includes bryophilous and terricolous taxa, but their modes of nutrition are not clear. Here, we investigate patterns of carbon and nitrogen utilization in numerous non-lignicolous Hymenochaetales and provide a phylogenetic context in which these non-canonical ecological guilds arose.

Methods

We combined stable isotope analyses of δ13C and δ15N and phylogenetic analyses to explore assignment and evolution of nutritional modes. Clustering procedures and statistical tests were performed to assign trophic modes to Hymenochaetales and test for differences between varying ecologies. Genomes of Hymenochaetales were mined for presence of enzymes involved in plant cell wall and lignin degradation and sucrolytic activity.

Key Results

Three different trophic clusters were detected – biotrophic, saprotrophic, and a second biotrophic cluster including many bryophilous Hymenochaetales and mosses. Non-lignicolous Hymenochaetales are generally biotrophic. All lignicolous Hymenochaetales clustered as saprotrophic and most terricolous Hymenochaetales clustered as ectomycorrhizal. Overall, at least 15 species of Hymenochaetales are inferred as biotrophic. Bryophilous species of Rickenella can degrade plant cell walls and lignin, and cleave sucrose to glucose consistent with a parasitic or endophytic life style.

Conclusions

Most non-lignicolous Hymenochaetales are biotrophic. Stable isotope values of many bryophilous Hymenochaetales cluster as ectomycorrhizal or in a biotrophic cluster indicative of parasitism or an endophytic life style. Overall, trophic mode diversity in the Hymenochaetales is greater than anticipated, and non-lignicolous ecological traits and biotrophic modes of nutrition are evolutionarily derived features.

The basidiomycete order Hymenochaetales Oberw. contains some 900 species of mostly polyporoid and corticioid fungi classified in 75 genera worldwide (Hibbett et al., 2014). Most are lignicolous white-rot saprotrophs that decompose wood (Wagner and Fischer, 2002; Binder et al., 2005; Larsson et al., 2006; Tedersoo et al., 2007). The order also includes a few species with diverse mycorrhizal abilities (Nouhra et al., 2013; Tedersoo and Smith, 2013; Kolařík and Vohník, 2018), plant pathogens (Larsson et al., 2006), and several agarics or mushroom-forming fungi that typically occur on or with bryophytes—particularly mosses or liverworts—or that occur on soil (Redhead et al., 2002) (Fig. 1). These non-lignicolous fungi, particularly the bryophilous agarics (Racovitza, 1959; Davey and Currah, 2006), were previously treated as Agaricales Underw. due to similarities in basidiome morphology (Redhead et al., 2002) but were recovered in the Hymenochaetales by molecular phylogenetic analyses (Moncalvo et al., 2000, 2002; Redhead et al., 2002). Later, the group was referred to as the Rickenella clade (Larsson et al., 2006) and classified in the family Rickenellaceae Vizzini (Vizzini, 2010; Nakasone and Burdsall, 2012), the name of which, unfortunately, is illegitimate due to inclusion of the type of the earlier described family Repetobasidiaceae Jülich (Jülich, 1981). In addition, the diversity of its constituents is unsettled, the monophyly of the group has been questioned (Larsson, 2007; Miettinen and Larsson, 2010), and their nutritional modes are largely unclear (Felix, 1988; Kost, 1988; Bresinsky and Schötz, 2006).

Details are in the caption following the image
Ecological and trophic diversity of Hymenochaetales. (A) Hymenochaete cinnamomea (lignicolous saprotroph). (B) Trichaptum abietinum (lignicolous saprotroph). (C) Contumyces rosellus (terricolous and possible biotroph). (D) Coltricia perennis (terricolous ECM biotroph). (E) Rickenella swartzii (bryophilous biotroph). (F) Rickenella minuta (terricolous ECM biotroph). (A–E) Photos M. G. Wood. (F) Photo P. B. Matheny. Scale bars = 10 mm.

The phyogenetic placement of the Rickenella clade in the Hymenochaetales raises several questions about its ecology and evolution: (1) is the non-lignicolous association derived or ancestral in the Hymenochaetales? (2) What are the trophic modes of bryophilous and terricolous Hymenochaetales? That is, are they saprotrophs that decay dead organic matter for carbon nutrition, or are they biotrophs that acquire carbon nutrition from a live symbiont? Or; (3) can they do both? If biotrophs, do they engage in a nutrient-exchange mutualism with bryophytes, or are they parasites of bryophytes?

Analysis of stable isotope signatures of nitrogen (δ15N) and carbon (δ13C) has been used to infer trophic modes of fungi (Hobbie et al., 2001; Trudell et al., 2004; Mayor et al., 2009; Seitzman et al., 2011; Hobbie and Högberg, 2012; Tedersoo et al., 2012; Birkebak et al., 2013; Trappe et al., 2015; Sánchez-García and Matheny, 2017). However, analysis of stable isotope data has not been used to explore the trophic status of bryophilous and terricolous Hymenochaetales, and very few Hymenochaetales overall have even been evaluated, probably due to the fact that most Hymenochaetales are lignicolous saprotrophs (Wagner and Fischer, 2002). Stable isotope data are useful to discriminate between Hymenochaetales that acquire their nutrition from a live associate or from dead organic matter, thus providing a powerful means to discern the degree of trophic diversity in the order without direct observations of trophic interactions in the field or laboratory.

Analysis of fungal nutritional modes can also be augmented by an array of approaches that include anatomical studies, pure syntheses or in vitro experiments, molecular ecology (e.g., barcoding ectomycorrhizal (ECM) root tips), and phylogenetic relatedness (Tedersoo et al., 2010). In addition, the presence or absence of unique suites of genes that encode enzymes involved in nutrient acquisition can also be used to characterize trophic modes in fungi (Parrent et al., 2009; Wolfe et al., 2012). To this end, the U.S. Department of Energy Joint Genome Institute has sequenced the whole genomes of two species of bryophilous Hymenochaetales, Rickenella fibula (Bull.) Raithelh. and R. mellea (Singer & Clémençon) Lamoure, and five genomes of lignicolous Hymenochaetales as of this writing.

Here we carry out a cluster analysis of stable C and N stable isotope data from nearly 1000 samples of Agaricomycetes Doweld, phylogenetic analyses of a multigene dataset, ancestral state reconstruction analyses, and genome searches for key enzymes used to degrade plant cell walls and lignin and sucrolytic activity, in order to determine whether bryophilous, lignicolous, and terricolous Hymenochaetales are saprotrophic or biotrophic, whether these states are derived or ancestral, and whether these ecological guilds share similar or dissimilar modes of nutrition.

MATERIALS AND METHODS

Stable isotope analyses

A total of 108 specimens, including 92 samples of Hymenochaetales (Table 1), were analyzed at the University of New Hampshire Isotope Lab (www. isotope.unh.edu) on an Elementar Americas Pyrocube (Elementar Americas Inc., Mt. Laurel, New Jersey, USA) elemental analyzer combined with a GeoVision isotope ratio mass spectrometer (GeoVision Inc., Taipei, Taiwan). Samples were prepared by grinding 2–3 mg of dried basidiome tissue. Several known biotrophic ECM (i.e., Amanita Pers., Cortinarius (Pers.) Gray, Inocybe (Fr.) Fr.), saprotrophic (i.e., Galerina marginata (Batsch: Fr.) Kühner, Trametes Fr., Trichaptum Murrill), plant pathogen (i.e., Rigidoporus populinus (Schumach.: Fr.) Pouzar), and autotrophic controls (i.e., Dicranum scoparium Hedw.) were also used. The analyses produced stable isotope signatures of carbon (δ13C) and nitrogen (δ15N) and elemental percent (%C, %N) (only δ13C and δ15N are shown in Table 1). Abundance values of 13C were measured relative to the Vienna Pee Dee Belemnite (Mayor et al., 2009) and 15N abundance values relative to atmospheric N.

Table 1. Metadata for 108 collections analyzed by stable isotope mass spectrometry, δ15N and δ13C stable isotope values, and trophic cluster assigned by mclust analysis. liv. = living gametophyte tissue; sen. = senescent gametophyte tissue. Herbarium abbreviations follow Thiers [continuously updated]
Taxon Herbarium accession no. Specimen-voucher no. Location Long./ Lat. Ecology Date of collection δ15N (‰) δ13C (‰)
Alloclavaria purpurea TENN071489 HRL0146 Quebec

48.7081

−67.3889

Terricolous On soil 2008-08-26 4.92 −24.28
Alloclavaria purpurea H6047394 Korhonenen 10305 Finland 66.5039 25.7294 Terricolous On soil 1991-08-18 3.10 −22.18
Alloclavaria purpurea H6047434 Korhonen 10411 Finland 65.9646 29.1887 Terricolous On soil 1991-08-26 1.41 −24.86
Alloclavaria purpurea H6034567

Niskanen

01-053

Finland

62.5000

22.7500

Terricolous On soil with Picea, Betula 2001-08-20 3.61 −24.46
Alloclavaria purpurea H6034547

Niskanen

01-044

Finland 64.8675 27.6752 Terricolous On soil 2001-08-17 4.01 −26.51
Alloclavaria purpurea H6047663 Ahokas s.n. Finland 60.9367 26.3996 No data 2005-09-04 2.80 −23.26
Amanita multisquamosa TENN070497 MGW1516 Tennessee

35.7525

−83.27305

Terricolous Soil 2015-07-11 4.11 −27.04
Bjerkandera adusta TENN059170 MF2 Tennessee

35.70778

−83.3800

Lignicolous On Wood 2001-05-23 2.53 −23.59
Blasiphalia pseudogrisella H6059323 Hoijer 3034 Finland

63.7530

26.9889

Bryophilous With Blasia 2001-08-14 1.07 −23.54
Blasiphalia pseudogrisella H7031951 Hoijer 4393 Estonia 59.3591 27.4211 Bryophilous With Blasia 2006-05-02 3.41 −26.18
Blasiphalia pseudogrisella H6059312 Hoijer 4118 Finland

62.2833

21.3833

Bryophilous With Blasia 2005-09-20 1.68 −23.76
Blasiphalia pseudogrisella H6019812 Hoijer 4539 Finland

69.0500

20.8000

Bryophilous With Blasia 2007-08-21 2.04 −24.66
Cantharellopsis prescotii TENN071487 HRL2135 Quebec

50.25

−63.6

Terricolous On soil 2015-09-10 0.42 −27.79
Cantharellopsis prescotii H6035464

Kytovuori

08-0808

Finland

60.2512

24.0675

Terricolous In damp depression, rich mixed forest 2008-09-03 −0.70 −32.82
Cantharellopsis prescotii H6050719

Kytovuori

93-850

Finland

60.2512

24.0675

Terricolous In grass-herb Picea forest 1993-09-09 0.39 −26.66
Cantharellopsis prescotii H6059300 Ohenoja s.n. Finland

64.8833

28.9166

Terricolous In moss with Picea 1992-08-24 −0.58 −28.85
Cantharellopsis prescotii (as Gerronema albidum) H6050710

Kytovuori

851422

Finland 60.3011 22.3022 Terricolous In rich Picea forest 1985-10-05 −1.86 −28.79
Cantharellopsis prescotii (as Gerronema albidum) H6059277

Kytovuori

90-231

Finland 64.8675 27.6752 Terricolous In rich Picea forest 1990-08-14 −1.56 −29.12
Coltricia cinnamomea H6059308 Niemela 6911 Finland

61.0900

25.0000

Terricolous On soil 2000-09-22 6.64 −26.91
Coltricia cinnamomea H6059331 Niemela 8199 Finland

61.8000

29.0170

Terricolous On soil 2005-09-15 5.65 −26.79
Coltricia montagnei TENN066402 MR070111-02 Tennessee

35.7388

−83.4222

Terricolous On soil 2011-07-11 8.19 −25.24
Coltricia montagnei TENN065217 SAT1017611 North Carolina

35.6111

−83.1250

Terricolous On soil 2010-06-25 8.28 −25.21
Coltricia perennis H6029159 Salo 10282 Finland

61.9126

29.2814

Terricolous On soil 2004-05-01 4.45 −24.94
Coltricia perennis H6002974 Salo 11024 Finland 60.3932 25.6653 Terricolous On soil 2007-04-05 11.86 −24.62
Coltricia perennis H6013608 Miettinen 18513 Finland

61.8000

29.0170

Terricolous On soil 2014-09-04 8.49 −25.66
Coltricia perennis H6059309 Haikonen 22781 Finland 61.9833 26.2666 Terricolous On soil 2003-09-10 8.56 −24.76
Contumyces rosellus TENN071494 MGW1462 California

38.2918

−122.4580

Terricolous On soil 2015-01-19 4.82 −27.62
Contumyces vesuvianus TUR203608 Italy 40.4350 18.0422 Bryophilous Moss 2015-11-30 −1.75 −27.83
Cortinarius corrugatus TENN070645 PBM4040 Tennessee

35.6725

−83.4930

Terricolous Soil 2015-07-16 12.04 −26.65
Cotylidia diaphana TENN071490 HRL1509 Quebec

45.4948

−73.8907

Terricolous On soil 2013-07-27 2.42 −23.20
Cotylidia pannosa TENN071488 HRL0287 Quebec

48.1841

−68.9843

Terricolous On soil 2009-08-30 8.32 −24.06
Cotylidia pannosa H6059288 Haikonen 25824 Finland 61.1721 25.5471 Terricolous On bank with mosses 2007-10-10 −1.87 −24.04
Cotylidia undulata TENN071491 HRL0625 Quebec

46.5220

−72.3213

Terricolous Moss site 2010-10-02 −4.41 −24.29
Cotylidia undulata H6059276 Haikonen 26475 Finland 60.7178 24.4417

Terricolous

In Funaria old fire site

2008-09-08 −3.97 −24.34
Cotylidia undulata H6059299 Saarenoksa 21494 Finland 60.2236 25.0678

Terricolous

Among mosses and ECM veg.

1994-09-20 −2.04 −23.29
Dicranum scoparium liv. TENN-B-071478 HBK013 Tennessee

35.9606

−83.9207

Terricolous On soil 2014-09-11 −2.07 −32.00
Dicranum scoparium sen. TENN-B-071478 HBK013 Tennessee

35.9606

-83.9207

Terricolous On soil 2014-09-11 −1.57 −31.18
Dicranum scoparium liv. TENN-B-0102885 DUKE bf20-3 North Carolina

35.9940

−78.8986

Terricolous On soil 2015-04-15 −4.67 −30.59
Dicranum scoparium sen. TENN-B-0102885 DUKE bf20-3 North Carolina

35.9940

−78.8986

Terricolous On soil 2015-04-15 −3.81 −30.80
Dicranum scoparium liv. TENN-B-071477 RAS051 Tennessee

36.2423

−83.9305

Terricolous On soil 2015-11-09 −4.39 −31.00
Dicranum scoparium sen. TENN-B-071477 RAS051 Tennessee

36.2423

−83.9305

Terricolous On soil 2015-11-09 −3.89 −30.93
Dicranum scoparium liv. TENN-B-071473 HBK016 Tennessee

35.6016

−83.9246

Terricolous On soil 2015-11-14 −1.20 −28.82
Dicranum scoparium sen. TENN-B-071473 HBK016 Tennessee

35.6016

−83.9246

Terricolous On soil 2015-11-14 −3.96 −27.90
Dicranum scoparium TENN-B-071474 HBK014 Tennessee

35.6016

−83.9246

Terricolous On soil 2015-11-14 −3.90 −32.42
Dicranum scoparium sen. TENN-B-071474 HBK014 Tennessee

35.6016

−83.9246

Terricolous On soil 2015-11-14 −3.10 −32.18
Galerina marginata TENN071079 FPD26 Tennessee

35.9338

−83.8774

Lignicolous On wood 2016-02-13 0.42 −23.33
Inocybe subochracea TENN062488 PBM2659 Tennessee

35.6019

−83.8136

Terricolous On soil 2013-09-13 −3.06 −24.84
Loreleia marchantiae TUR83652 Finland No data

Bryophilous

Liverwort

No data 1.51 −26.07
Loreleia marchantiae TUR203090 Lahti 24/14 Finland 61.3230 25.7322 Bryophilous Liverwort 2014-07-20 1.89 −25.58
Loreleia postii H6055478 Harmaja s.n. Finland

60.9300

27.6000

Bryophilous Soil, old fire place 1976-08-29 −0.66 −23.41
Loreleia postii H6059335 Saarenoksa 48384 Finland 60.2236 25.0678

Bryophilous

In moss-covered old fire place

1984-10-21 −2.56 −25.71
Muscinupta laevis H6059292

Haikonen

19745

Finland 60.9827 25.6612

Bryophylous

Polytrichum

1999-10-10 −0.90 −25.81
Muscinupta laevis H6059303

Saarenoksa

43385

Finland 60.2129 25.0779

Bryophylous

Polytrichum

1985-09-14 −3.55 −27.24
Muscinupta laevis H6003362 Salo 9493 Finland

60.2371

25.1371

Bryophylous

Polytrichum

2003-05-14 1.15 −25.93
Odonticium romellii H6059330 Murdoch 11 Finland 61.6987 23.7896 Lignicolous On wood 2006-09-30 −2.90 −22.32
Onnia tomentosa H6048516 Niemela 9079 Finland 61.7695 23.0658

Lignicolous

Ground

2013-08-30 −1.43 −21.37
Onnia tomentosa H6048491 Niemela 9081 Finland

61.7695

23.0658

Lignicolous

Ground

2013-08-30 −4.28 −22.66
Phellinus nigricans H6012648 Miettinen 14031 Finland 64.6791 28.4840 Lignicolous On wood 2010-04-12 0.44 −23.44
Phellinus nigricans H6048524 Niemela 9065 Finland 61.5496 23.5961 Lignicolous On wood 2013-07-22 −4.48 −23.56
Rickenella fibula TENN066160 SAT1117302 Tennessee

35.7055

−83.3833

BryophilousOn moss 2011-06-22 −3.25 −26.04
Rickenella fibula TENN060941 TFB13109 Tennessee

35.6111

−83.8166

BryophilousOn moss 2006-05-24 −3.79 −28.44
Rickenella fibula TENN071481 JMB101914-06 Washington

47.6080

−122.3351

BryophilousOn moss 2014-10-19 1.87 −28.23
Rickenella fibula TENN071478 HBK013 Tennessee

35.6016

−83.9246

Bryophilous Dicranum 2014-09-11 2.08 −25.42
Rickenella fibula TENN071480 BPL872 Tennessee

36.1028

−84.7200

Bryophilous Dicranum 2015-10-03 −4.96 −28.71
Rickenella fibula TENN071477 RAS051 Tennessee

36.2423

−83.9305

Bryophilous Dicranum 2015-09-15 −4.52 −27.37
Rickenellafibula” (R. mellea) TENN071476 HBK012 Washington

47.6080

−122.3351

Bryophilous Dicranum 2014-10-10 0.73 −28.54
Rickenella fibula TENN071482 HBK015 Washington

47.6080

−122.3351

Bryophilous Dicranum 2014-10-11 −2.92 −23.40
Rickenella fibula TENN071474 HBK014 Tennessee

35.6016

−83.9246

Bryophilous Dicranum 2015-11-14 −3.25 −29.67
Rickenella fibula TENN071473 HBK016 Tennessee

35.6016

−83.9246

Bryophilous Dicranum 2015-11-14 −2.70 −25.54
Rickenella fibula TENN071479 JMB101914-07 Washington

47.6080

−122.3351

Bryophilous On moss 2014-10-19 −2.09 −29.22
Rickenella fibula TENN060305 TFB12057 Tennessee

35.7075

−83.3825

Bryophilous On moss 2006-09-04 −2.09 −28.67
Rickenella fibula TENN066877 PBM2506 Mass.

42.2627

−71.8019

Bryophilous On moss 2003-10-31 −0.16 −24.87
Rickenella fibula TENN066876 PBM2503 Mass.

42.2627

−71.8019

Bryophilous On moss 2003-10-25 0.32 −25.95
Rickenella fibula TENN066245 MGW992 Tennessee

35.7583

−83.2125

Bryophilous Dicranum 2011-06-25 −4.87 −28.63
Rickenella fibula H6059291 Ohenoja s.n. Finland 62.1332 22.5581 Bryophilous On moss 2006-09-22 −2.49 −25.81
Rickenella fibula H6034921

Kytovuori

93-040

Finland 62.2409 23.7702 Bryophilous On moss 1993-08-04 −2.82 −28.36
Rickenella fibula H6034922

Kytovuori

94-482

Finland 60.2251 25.0201 Bryophilous On moss 1994-04-14 −0.25 −25.26
Rickenella fibula H6019327 Salo 1882 Finland 67.6507 24.9158 Bryophilous On moss 1995-10-15 −1.59 −24.89
Rickenella fibula H6059302

Kytovuori

96-071

Finland 60.4748 25.0971 Bryophilous On moss 1996-07-12 −0.13 −25.09
Rickenella fibula CORD MES950 Chile

−40.1738

−73.4735

Bryophilous On moss 2015-05-01 −1.21 −22.95
Rickenella fibula TENN061243 TFB13157 North Carolina

35.5016

−83.4008

Bryophilous On moss 2006-07-19 −2.93 −30.03
Rickenella fibula TENN061272 TFB13163 Quebec

52.9527

−73.5488

Bryophilous On moss 2006-07-29 −5.04 −28.89
Rickenella minuta TENN055098 TFB8528 Argentina −42.7244 -71.7530 Terricolous On soil 1996-05-12 6.45 −25.43
Rickenella minuta TENN055094 TFB8524 Argentina

−42.7244

−71.7530

Terricolous On soil 1996-05-06 2.37 −25.28
Rickenella minuta CORD MES1535 Chile

−40.6785

−72.1120

Terricolous On soil 2016-05-03 −3.77 −25.98
Rickenella minuta TENN071469 MES1781 Chile

−40.6549

−72.1121

Terricolous On soil 2016-05-07 0.72 −25.36
Rickenella minuta TENN071467 MES1950 Argentina

−41.0000

−71.500

Terricolous On soil 2016-05-13 0.78 −24.83
Rickenella minuta CORD MES2168 Argentina

−41.1499

−71.2999

Terricolous On soil 2016-05-18 2.36 −24.46
Rickenella minuta TENN071472 MES1558 Chile

−40.5785

−73.1335

Terricolous On soil 2016-05-03 −1.65 −26.44
Rickinella minuta TENN071468 MES1892 Argentina

−41.1499

−71.2999

Terricolous On soil 2016-05-13 5.74 −26.45
Rickinella minuta CORD MES1891 Argentina

−41.1499

−71.2999

Terricolous On soil 2016-05-13 8.14 −25.30
Rickinella minuta TENN071471 MES2110 Argentina

−41.0000

−71.5000

Terricolous On soil 2016-05-16 3.28 −25.12
Ricknella minuta CORD MES1965 Argentina

−41.1499

−71.2999

Terricolous On soil 2016-05-13 2.96 −26.35
Rickenella minuta CORD MES1656 Chile

−40.6785

−72.1120

Terricolous On soil 2016-05-04 0.43 −26.14
Rickenella minuta TENN071466 MES1054 Chile

−40.6785

−72.1119

Terricolous On soil 2015-05-05 −2.13 −25.11
Rickenella minuta TENN071470 MES1259 Argentina

−41.0006

−71.4999

Terricolous On soil 2015-05-14 1.41 −26.64
Rickenella setipes H6059279

Kytovuori

82-234

Finland

60.1900

25.0400

Bryophilous

Soil, garden meadow

1982-09-01 −0.60 −27.22
Rickenella setipes H6059290

Saarenoksa

06978

Finland 60.2091 24.9647 Bryophilous On moss 1978-07-18 −1.84 −26.38
Rickenella swartzii TENN071475 HBK017 Washington

46.9474

−122.3091

Bryophilous On moss 2014-10-11 −3.22 −29.59
Rickenella swartzii TENN071493 MGW1075 California

39.2683

−123.7871

Bryophilous On moss 2011-11-18 0.73 −24.23
Rickenella swartzii TENN071492 MGW1341 California

39.2682

−123.7871

Bryophilous On moss 2013-11-16 −0.18 −26.68
Rickenella swartzii TENN071484 HRL1399 California

39.3649

−123.8144

Bryophilous On moss 2012-12-17 0.49 −26.30
Rigidoporus populinus H6059304 Airaksinen s.n. Finland

60.3768

25.2690

Lignicolous On wood 1993-04-11 −3.64 −23.39
Rigidoporus populinus H6052662 Kotiranta 27027 Finland 60.0459 24.0046 Lignicolous On wood 2004-04-23 −0.09 −22.11
Sphagnomphalia brevibasidita (=Gerronema cinctum) H6059278 Askola 1633 Finland 60.5256 24.7621 Bryophilous In moist spring site 1985-06-27 −1.51 −23.48
Trametes ochracea TENN060148 TFB12211 Russia

53.3655

49.8927

Lignicolous On wood 2004-08-21 −3.50 −22.42
Trichaptum fuscoviolaceum H6059286

Miettinen

7555,3

Finland 61.3509 25.2781 Lignicolous On wood 2003-09-04 −4.15 −22.12
Trichaptum fuscoviolaceum H6059317 Haikonen 25849 Finland 61.2102 26.0458 Lignicolous On wood 2007-10-16 −2.14 −23.11

In order to assign trophic states to Hymenochaetales samples, we assembled a global data set of 957 taxa (Appendix S1; see Supplemental Data with this article). δ15N and δ13C data were available for 849 of these taxa from Mayor et al. (2009), Birkebak et al. (2013), and Sánchez-García and Matheny (2017). Stable isotope data for the remaining 108 taxa were produced during this study. Previously available data contained only five Hymenochaetales samples (i.e., two from Inonotus tomentosus (Fr.) Teng and one each from Onnia vallata (Berk.) Y.C. Dai & Niemelä, Oxyporus cuneatus (Murrill) Aoshima, and Trichaptum biforme (Fr.) Ryvarden). Of the δ15N and δ13C measurements from our 108 samples, 92 originated from the Hymenochaetales (thus 97 total in the global dataset), 49 of which were scored as bryophilous following Redhead et al. (2002), 14 as lignicolous, and 34 as terricolous based on specimen-voucher metadata or the literature (Hansen and Knudsen, 1997; Bernicchia and Gorjón, 2010; Ryvarden, 2010; Table 1). Of the 849 taxa in the global dataset, these were assigned known trophic modes following Mayor et al. (2009), Seitzman et al. (2011), Birkebak et al. (2013), and Sánchez-García and Matheny (2017) – saprotrophic (SAP), ectomycorrhizal (ECM), neither saprotrophic nor ectomycorrhizal (NS-NE), autotrophic (AUTO), or, in the case of most our Hymenochaetales samples, as unknown. Of our samples, trophic states in the Hymenochaetales were scored for Coltricia Gray (ECM; Danielson, 1984; Tedersoo et al., 2007), the five samples from Mayor et al. (2009) (SAP), and nine lignicolous samples in the genera Onnia P. Karst., Phellinus Quél., Rigidoporus Murrill, and Trichaptum (SAP; Larsson et al., 2006; Wu et al., 2017).

We used the mclust package in R version 3.3.1 to formulate discrete clusters of functional trophic groups (R Development Core Team, 2014; Scrucca et al., 2016). This package models the stable isotope values as a Gaussian finite mixture while permitting differences in covariance structures, sample sizes of clusters, and number of clusters (components). Different models are compared using Bayesian information criterion (BIC) values. The package has an advantage over other clustering procedures, viz., k-means clustering (Steinley, 2006), where the number of clusters (k) need not be pre-assigned.

After assignment to clusters, the distribution of bryophilous, lignicolous, and terricolous Hymenochaetales was tabulated according to mclust results from the best-fit model. Tests for equal distribution of ecological guild by trophic cluster were performed using Fisher's exact test (Fisher, 1922). Analyses of variance (ANOVA) and Tukey-Kramer post-hoc tests (Tukey, 1949) were performed to test null hypotheses of no differences between stable N and C isotope values of trophic modes (SAP, ECM, NS-NE, AUTO) of taxa in the same trophic cluster. The clustering procedures and statistical tests were conducted in R. Because corrections or normalizations for the Suess effect (Tans et al., 1979; McCarroll and Loader, 2004) do not significantly improve analysis of the stable isotope data (Mayor et al., 2009; Trappe et al., 2015), no adjustments were made to the results presented here.

Collections and taxon sampling for phylogenetic analyses

Samples of Hymenochaetales were selected for phylogenetic analyses from the University of Tennessee (TENN), the University of Helsinki (H), University of Turku (TUR), Universidad Nacional de Córdoba (CORD), the University of Florida (FLAS), and independent collectors (Appendix S2; herbarium abbreviations follow Thiers [continuously updated; http://sweetgum.nybg.org/science/ih/]. For phylogenetic analyses of the order, we downloaded available sequences of nuclear 28S rRNA (28S), 18S rRNA (18S), and RNA polymerase II second largest subunit (rpb2) from GenBank.

Fungal DNA extractions, PCR, and sequencing

Genomic DNA was extracted from 10–30 mg of dried basidiome tissue from fungal samples. These were ground with liquid nitrogen and a pinch of sterile sand using a porcelain mortar and pestle. DNA was then extracted using an E.Z.N.A. HP Fungal DNA kit (Omega Bio-Tek, Norcross, Georgia, USA). Elutions of genomic DNA were then diluted with sterile water into two 1:10 serial dilutions. PCR, PCR purification, and sequencing were performed following protocols outlined in Judge et al. (2010). PCR products were viewed on 1% agarose gels prepared with SYBR safe (Thermo Fisher Scientific, Rockwood, Tennessee, USA) and a UV trans-illuminator. The following primer pairs were used for PCR and sequencing: (1) LR0R-LR7 or LR0R-LR5 (28S; Vilgalys and Hester, 1990); (2) PNS1-NS41 (18S; Hibbett, 1996); and (3) rpb2-b6F-rpb2-b7.1R (rpb2; Matheny, 2005). In addition, we amplified and sequenced nuc rDNA ITS1-5.8S-ITS (ITS) to aid in species confirmation using primers ITS1F-ITS4 (White et al., 1990; Gardes and Bruns, 1993). Raw sequences were edited and assembled using Sequencher 5.0.1 (Gene Codes Corp., Ann Arbor, Michigan, USA). All new DNA sequences produced during this study were deposited at GenBank (Appendix S2).

Phylogenetic analyses

Nucleotide sequences of 18S, 28S, and rpb2 were assembled and aligned in ClustalX (Larkin et al., 2007), manually adjusted in AliView (Larsson, 2014) or MacClade (Maddison and Maddison, 2005), and concatenated into a supermatrix in SeaView version 4.5.2 (Gouy et al., 2010) after inspection for strongly supported interlocus conflict (ML bootstraps >70% for conflicting clades). The supermatrix included taxa across the Hymenochaetales and represents a three-locus alignment of concatenated 28S, 18S, and rpb2 gene regions. For members of the Rickenella clade sensu Larsson et al. (2006), samples were included in the Hymenochaetales dataset if represented by at least one locus, usually 28S. DNA sequences were not available for Kurtia argillacea (Bres.) Karasiński, an ericoid mycorrhizal member of the Rickenella clade (Kolařík and Vohník, 2018). Gene regions absent for taxa were coded as missing data. To this data set we added sequences of outgroups from the Polyporales Gäum – Phlebia brevispora Nakasone, Bjerkandera adusta (Willd.: Fr.) P. Karst., Polyporus brumalis (Pers.: Fr.) Fr., and Punctularia strigosozonata (Schwein.) P.H.B. Talbot; Phallomycetidae Hosaka et al. – Ramaria rubella (Schaeff.) R.H. Petersen; and Auriculariales J. Schröt. – Auricularia subglabra Looney et al. Outgroup choice was based on Hibbett et al. (2014). Auricularia subglabra was used to root resulting phylogenetic trees. Ambiguously aligned sites were excluded prior to phylogenetic analyses.

The supermatrix was analyzed using maximum likelihood (ML) and Bayesian inference (BI) phylogenetic methods. Single and linked locus data sets were analyzed with ML only. RAxML version 8 (Stamatakis, 2014) was used to conduct phylogenetic analyses under the ML criterion with 1000 bootstrap replicates, and MrBayes version 3.2.6 (Ronquist et al., 2012) was used to estimate posterior probabilities (PP) from a sample of trees drawn from a posterior distribution. In the RAxML analyses, a GTRGAMMA model of evolution was assigned to gene partitions as recommended by the user manual. Data were partitioned using the best partitioning scheme from PartitionFinder version 2.1.1 (Lanfear et al., 2017) with the 28S+18S loci modeled in combination but separate from the three rpb2 codon positions resulting in four partitions total. A GTR model with gamma distributed rate heterogeneity was also used in BI analyses. The Hymenochaetales dataset was run for 15 million generations, and trees were sampled every 1000 generations. Convergence diagnostics and run length were determined based on recommendations in the MrBayes user manual. The first 25% of the trees in the posterior distribution were removed prior to sump and sumt commands and PP calculations.

Ancestral state reconstructions (ASR) were conducted in Mesquite version 2.74 (Maddison and Maddison, 2010) using the maximum parsimony criterion on the best ML tree. The strength of ASR assignments was assessed by examination of a given internode and its frequency in 1000 posterior trees sampled from the BI analysis. Trophic states were scored for tips in the supermatrix dataset based on the mclust analysis of the best-fit model (Appendix S1). Datasets and resulting tree files were submitted to TreeBASE (submission 21259; www.treebase.org) and are available from the corresponding author (PBM).

Presence/absence of genes involved in plant tissue degradation

We used enzyme names as key words to search against genome annotations based on InterPro (Jones et al., 2014) and Pfam (Finn et al., 2016) to assess the presence of enzymes involved in plant tissue degradation across seven available genomes of Hymenochaetales at MycoCosm (http://jgi.doe.gov/fungi) on 17 June 2017. The seven genomes searched were: (1) Rickenella fibula (JGI 333301 v1.0; (2) R. mellea (JGI 334780 v1.0); (3) Fomitiporia mediterranea M. Fisch. (JGI 56107 v1.0); (4) Onnia scaura (Lloyd) Imazeki (JGI 245618 v1.0); (5) Porodaedalea niemelaei M. Fisch. (JGI 333975 v1.0); (6) Schizopora paradoxa (Schrad.: Fr.) Donk (JGI 239088 v1.0); and (7) Trichaptum abietinum (Pers.: Fr.) Ryvarden (JGI 210203 v1.0) (Grigoriev et al., 2011, 2014). The functional annotations were conducted according to the DOE-JGI Microbial Genomic Annotation Pipeline (Huntemann et al., 2015). The protein sequences were searched against the Pfam database using HMMER version 3.0 (Eddy, 2011). The gathering threshold (–cut_ga) was chosen when using the pfam_scan.pl script. InterProScan was conducted with default settings.

Search words included the following: arabinosidase, cellulase, cellobiohydrolase, chitinase, galactosidase, glucanase, invertase, mannosidase, polygalacturonase, and xylanase, which are mainly enzymes of various glycoside hydrolase families involved in degradation of plant cell walls common in biotrophic parasites and saprotrophic fungi (Talbot et al., 2008; Zhao et al., 2013; Kohler et al., 2015). In addition we searched for the enzymes laccase, ligninase, and peroxidase involved in lignin degradation (Read et al., 2004; Talbot et al., 2008; Zhao et al., 2013; Kohler et al., 2015) under the assumption that the Rickenella genomes should lack such enzymes because bryophytes do not produce wood. For enzymes involved in lignin degradation and sucrolytic activity in the two Rickenella genomes, we searched the Conserved Domain Database (CDD: https://www.ncbi.nlm.nih.gov/Structure/cdd/cdd.shtml) with amino acid sequences of each enzyme to identify active sites across these enzymes to ensure protein homology.

RESULTS

Stable isotope data and cluster analyses

Ratios of stable isotopes are recorded as δ15N and δ13C values in Table 1, including 108 new samples, 92 of which represent Hymenochaetales. The remaining new samples are from control saprotrophic, ectomycorrhizal, and autotrophic taxa (the latter from the moss Dicranum scoparium, an associate of Rickenella fibula). In summary, 97 Hymenochaetales samples are included in Table 1.

The program mclust produced BIC values for three models: (1) mclust VVV,3 (ellipsoidal, varying volume, shape, and orientation, model with three components; BIC −9631.316 – the best-fit model); (2) mclust VEV,4 (ellipsoidal, equal shape, model with four components; BIC −9631.394 – second best-fit model); and (3) mclust VVI (diagonal, varying volume and shape; model with four components; BIC −9633.893 – third best-fit model).

According to the best-fit model, the 957 stable isotope samples cluster into three discrete groups or components in the mclust analyses (Fig. 2A). A biotrophic cluster of 637 samples (cluster 1) is dominated by ECM taxa (78%) and indicated by circles (Table 2). A saprotrophic cluster of 248 samples (cluster 2) is dominated by SAP taxa (88%) and indicated by squares. A third cluster was formed from 72 samples, including 10 Dicranum autotrophic samples and 62 samples of Agaricomycetes of varying trophic states (SAP, ECM, unknown) and is indicated by triangles. No NS-NE trophic samples grouped in clusters 2 and 3. Nearly half of the unknown samples of bryophilous Hymenochaetales grouped in cluster 3. This cluster contained none of our lignicolous samples and only one terricolous sample. Bryophilous Hymenochaetales are represented by black-filled circles, squares, and triangles, thus appearing in all three clusters, whereas non-bryophilous Hymenochaetales are indicated by orange-filled circles or squares (and one triangle) and appear almost exclusively in clusters 1 and 2.

Details are in the caption following the image
Mclust analysis of 957 stable isotope samples from Appendix S1. (A) Mclust classification according to the best-fit model (VVV, 3 components). Circles = trophic cluster 1; squares = trophic cluster 2; triangles = cluster 3. Spheres indicate covariance for each cluster. 78% of cluster 1 (circles) is composed of known ECM fungi and all NS-NE fungi. 88% of cluster 2 (squares) includes known SAP fungi; cluster 3 (triangles) includes known autotrophs and a small number of ECM and SAP fungi and nearly half of bryophilous Hymenochaetales. The latter are represented by black-filed circles, squares, and triangles, and appear in all three clusters. Cluster 3 (triangles) contains no lignicolous samples and only one terricolous sample. Terricolous Hymenochaetales are indicated by orange-filled circles (cluster 1) and one orange-filled triangle (cluster 3). Lignicolous Hymenochaetales are indicated by orange-filled squares. (B) Mclust classification according to the second best-fit model (VEV, 4 components). Circles = trophic cluster 1 (mainly NS-NE and ECM taxa); squares = trophic cluster 2 (SAP); triangles = trophic cluster 3 (other ECM); crosses = trophic cluster 4 (biotrophic other). Spheres indicate covariance for each cluster. (C) Mclust classification according to the third best-fit model (VIV, 4 components). Circles = trophic cluster 1 (mainly NS-NE and ECM taxa); squares = trophic cluster 2 (SAP); triangles = trophic cluster 3 (other ECM); crosses = trophic cluster 4 (biotrophic other). Spheres indicate covariance for each cluster.
Table 2. Three trophic clusters recovered by mclust analysis of stable isotope values provided in Appendix S1. The total number of taxa by cluster is indicated along with the number of Hymenochaetales samples that are bryophilous, lignicolous, and terricolous. For each cluster the percentage (%) and total number of taxa that are considered saprotrophic (SAP), ectomycorrhizal (ECM), neither saprotrophic nor ectomycorrhizal (NS-NE), autotrophic (AUTO), and unknown (UKN) are also shown
Cluster Trophic status Total taxa Bryophilous Hymeno-chaetales Lignicolous Hymeno-chaetales Terricolous Hymeno-chaetales % / Total Trophic States from Appendix S1
SAP ECM NS-NE AUTO UKN
1 Biotrophic 637 23 0 29

9%

58

78%

495

7%

45

0%

0

6%

39

2 Saprotrophic 248 5 14 4

88%

218

7%

17

0%

0

0%

0

3%

13

3 Biotrophic 72 21 0 1

36%

26

18%

13

0%

0

14%

10

32%

23

The second and third best-fit models produced a fourth component into which taxa typically with high δ15N ratios (Camarophyllopsis Herink, Clavaria Vaill. ex L., Clavulinopsis Overeen, Ramariopsis (Donk) Corner – all NS-NE taxa of Seitzman et al. (2011) and Birkebak et al. (2013)) clustered with mainly known ECM taxa also with high δ15N ratios (Appendix S1; Fig. 2B–C). The only Hymenochaetales samples to group into this cluster were four samples of the known ECM genus Coltricia.

Hypothesis testing

Table 2 shows the number of samples of bryophilous, lignicolous, and terricolous Hymenochaetales distributed across the three different trophic clusters (clusters 1, 2, and 3) inferred by the best-fit model (Fig. 2A). This table also includes the number of samples and their percentages by known and unknown trophic assignment from Appendix 1 for each cluster according to the best-fit model. ANOVA results and subsequent Tukey-Kramer post-hoc tests strongly support N and C stable isotope mean differences between each of the three clusters (Table 3). If bryophilous Hymenochaetales have more or less equal ratios across clusters (the null hypothesis), then we can conclude that bryophilous Hymenochaetales are more functionally diverse than expected in that bryophilous Hymenochaetales exhibit different trophic modes. Indeed, this is what is observed from Fisher's exact test where one-third of the samples are expected to group in the biotrophic cluster, and the remaining two-thirds in the other two clusters (saprotrophic and unknown cluster 3) (P = 0.22) in support of the null hypothesis. That is, bryophilous Hymenochaetales do not particularly favor one trophic cluster over another.

Table 3. Comparison of trophic cluster assignments by mclust to test differences in sample means using ANOVA and Tukey-Kramer post-hoc tests. Stable isotope data from Mayor et al. (2009), Birkebak et al. (2013), Sánchez-García and Matheny (2017), and this study. An asterisk (*) indicates a statistically significant difference at the 0.05 level
Test Trophic comparison δ15N (‰) δ13C (‰)
ANOVA All clusters

F2, 954 = 301.1,

P <0.00001*

F2, 954 = 301.1,

P <0.00001*

Tukey-Kramer post-hoc cluster 1 vs. cluster 3 P <0.00001* P <0.001*
cluster 1 vs. cluster 2 P <0.00001* P <0.00001*
cluster 2 vs. cluster 3 P = 0.03* P <0.00001*

At this stage, we then tested whether bryophilous taxa in cluster 1 (ECM) have stable isotope values similar to those of known saprotrophic and NS-NE taxa, and whether bryophilous taxa in cluster 3 (biotrophic other) share similar isotope values with known ECM fungi, saprotrophs, and autotrophs (Dicranum scoparium) (Table 4). Bryophilous samples in cluster 1 share similar C ratio signatures as known ECM and NS-NE taxa in the global dataset. Their N ratio signatures, however, are significantly different compared to ECM and NS-NE taxa but similar to saprotrophs. These results support the hypothesis that bryophilous Hymenochaetales in cluster 1 receive their C source from photosynthetic partners, as do ECM and NS-NE taxa, but do not engage in nutrient exchange of N with their bryophyte associates similar to saprotrophic fungi.

Table 4. Comparison of stable isotope results from Appendix S1 to test differences in sample means between trophic clusters and trophic states using ANOVA and Tukey-Kramer post-hoc tests. Stable isotope data from Mayor et al. (2009), Birkebak et al. (2013), Sánchez-García and Matheny (2017), and this study. ECM = ectomycorrhizal; NS-NE = neither saprotrophic nor ectomycorrhizal (both ECM and NS-NE are biotrophic); SAP = saprotrophic; Moss samples are from Dicranum scoparium. An asterisk (*) refers to a statistically significant difference at the 0.05 level
Test Trophic comparisona δ 15N (‰) δ 13C (‰)
ANOVA Bryophilous Hymenochaetales in cluster 1; ECM non-Hymenochaetales; SAP non-Hymenochaetales; NS-NE

F3, 863 = 278.2,

P <0.00001*

F 3, 863 = 194.6,

P <0.00001*

Tukey-Kramer post-hoc Bryophilous Hymenochaetales in cluster 1 vs. ECM non-Hymenochaetales P <0.00001* P = 0.95
Bryophilous Hymenochaetales in cluster 1 vs. SAP non-Hymenochaetales P = 0.89 P <0.00001*
Bryophilous Hymenochaetales in cluster 1 vs. NS-NE non-Hymenochaetales P <0.00001* P = 1
ANOVA Bryophilous Hymenochaetales in cluster 3; ECM non-Hymenochaetales; SAP non-Hymenochaetales; moss

F 3, 833 = 180.1,

P <0.00001*

F 3, 833 = 311.6,

P <0.00001*

Tukey-Kramer post-hoc Bryophilous Hymenochaetales in cluster 3 vs. ECM non-Hymenochaetales P <0.00001* P <0.00001*
Bryophilous Hymenochaetales in cluster 3 vs. SAP non-Hymenochaetales P = 0.047* P <0.00001*
Bryophilous Hymenochaetales in cluster 3 vs. moss P = 0.88 P = 0.001*
ANOVA Terricolous Hymenochaetales in cluster 1; ECM non-Hymenochaetales; SAP non-Hymenochaetales; NS-NE

F 3, 874 = 268.7,

P <0.00001*

F 3, 874 = 194.7,

P <0.00001*

Tukey-Kramer post-hoc Terricolous Hymenochaetales in cluster 1 vs. ECM non-Hymenochaetales P = 0.22 P = 0.95
Terricolous Hymenochaetales in cluster 1 vs SAP non-Hymenochaetales P <0.00001* P <0.00001*
Terricolous Hymenochaetales in cluster 1 vs NS-NE non-Hymenochaetales P <0.00001* P = 0.90
ANOVA Lignicolous Hymenochaetales in cluster 2; ECM non-Hymenochaetales; SAP non-Hymenochaetales

F 2, 816 = 243.9,

P <0.00001*

F 2, 816 = 292.3,

P <0.00001*

Tukey-Kramer post-hoc Lignicolous Hymenochaetales in cluster 2 vs. ECM non-Hymenochaetales P <0.00001* P <0.00001*
Lignicolous Hymenochaetales in cluster 2 vs. SAP non-Hymenochaetales P = 0.03* P = 0.76

Note

  • a Trophic states were not tested in Tukey-Kramer tests if absent from a given cluster. For example, mosses (autotrophs) are absent from clusters 1 and 2, and thus not tested for comparison.

Bryophilous Hymenochaetales in cluster 3, however, have significantly different C and N values than those of ECM and saprotrophic fungi (Table 4). Indeed, their N values overall are no different from N values of the autotrophs in this cluster. These results support the hypothesis that the bryophilous fungi in cluster 3 are not ECM (and thus not N exchange mutualists), NS-SE (none are in this cluster), or saprotrophic. Instead, these fungi are deriving their N similar to that of autotrophs. Given their overall very low δ13C ratios (Appendix S1) it seems reasonable to assume their source of C is from live autotrophs. We thus conclude that the bryophilous unknown samples in cluster 3 are best interpreted as parasites or endophytes.

We also tested whether lignicolous Hymenochaetales are distributed equally across trophic clusters under the expectation that one-third of these should group in the saprotrophic cluster, because they produce basidiomes on wood, and two-thirds in clusters 1 and 3. All 14 lignicolous samples grouped in the saprotrophic cluster 2 (Fig. 2; Table 2). Doing Fisher's exact test, we can reject the null hypothesis (P < 0.001). Therefore, lignicolous Hymenochaetales are saprotrophic significantly more so than expected by chance. Do lignicolous Hymenochaetales in cluster 2 share similar stable isotope values to taxa of other known trophic assignments (ECM, saprotrophic)? This hypothesis is partially supported in that lignicolous Hymenochaetales share similar C values with known saprotrophs (both acquire C from dead organic material) but have deviating N values (Table 4; see also Fig. 2, orange-filled squares). This raises further questions whether these differences in N ratios might be due to small sample size, geographic location, substrate, or other factors.

Finally, we tested whether terricolous Hymenochaetales are distributed equally across the three trophic clusters. If so, then one-third should group in the biotrophic cluster and two-thirds in clusters 2 and 3. However, as Table 2 shows, 29 of 34 terricolous Hymenochaetales group in the biotrophic cluster dominated by known ECM samples. Results from Fisher's exact test strongly reject a null distribution (P < 0.0001). Therefore, terricolous Hymenochaetales are biotrophic, and likely mycorrhizal, more often than expected by chance. Do terricolous Hymenochaetales in cluster 1 (biotrophic) have similar stable isotope values to those of known saprotrophic and NS-NE taxa? Tukey-Kramer post-hoc tests (Table 4) significantly reject stable isotope similarities of terricolous Hymenochaetales to C and N values of known saprotrophic taxa and N values of NS-NE taxa. Thus, the stable isotope values of members of this guild are most similar to known ECM taxa.

Phylogenetic analyses of the Hymenochaetales and the Rickenella clade

Two hundred twelve new DNA sequences were produced during this study from herbarium specimens (66 for 28S, 54 for 18S, 33 for rpb2, and 59 for ITS; Appendix S2). No strongly supported conflict was observed between individual gene trees (rRNA and rpb2; data not shown but trees available at TreeBASE or from the corresponding author). The Hymenochaetales supermatrix of 18S, 28S, and rpb2 included 157 taxa and 3880 sites. 22,502 trees were sampled from the posterior distribution after the burn-in and used to calculate PP values in the BI analysis.

Members of the Rickenella clade sensu Larsson et al. (2006) do not form a monophyletic group (Fig. 3; Appendix S3). Rather, we recover a poorly supported clade of lineages (Fig. 3A) belonging to the genera Alloclavaria Dentinger & McLaughlin, Atheloderma Parmasto, Blasiphalia Redhead, Cantharellopsis Kuyper, Contumyces Redhead et al., Cotylidia P. Karst., Globulicium Hjortstam, Leifia Ginns, Loreleia Redhead et al., Muscinupta Redhead et al., Odonticium Parmasto, Peniophorella P. Karst., Rickenella Raithelh., and Sphagnomphalia Redhead et al. (Fig. 3A). All of these are bryophilous or terricolous with the exception of Globulicium, Leifia, Odonticium, and Peniophorella, which are lignicolous. Note that Sidera Miettinen & K.H. Larss. (lignicolous and presumed saprotroph) and Kurtia Karasiński (lignicolous, but ericoid biotroph) were not included in the study.

Details are in the caption following the image
Maximum likelihood (ML) phylogeny of the order Hymenochaetales based on analyses of a 28S, 18S, and rpb2 supermatrix. (A) Rickenella clade. (B) Remainder of the Hymenochaetales and outgroups. Taxon names in quotes are mislabeled. An asterisk (*) indicates lineages recovered in the Rickenella clade sensu Larsson et al. (2006). Values >50% above or below branches represent proportions from 500 bootstrap replicates. Posterior probabilities >0.95 are also indicated at internodes. Blue bold taxon labels group in trophic cluster 1 (ECM). Black bold taxon labels group in trophic cluster 2 (saprotrophic). Green bold taxon labels group in cluster 3 (biotrophic other). Tips labeled black but not in bold (excluding outgroups) lack stable isotope data. A lower case delta symbol (δ) next to a tip indicates that stable isotope data were produced from the sequenced specimen (Table 2). B, L, and T next to labeled tips indicate bryophilous (B), lignicolous (L), and terricolous (T) substrates. Ancestral states with >0.9 posterior probability are indicated with filled circles at or along internodes leading to more than one species. Black circles indicate saprotrophic states and blue circles ECM states.

Examination of the Rickenella clade (Fig. 3A) also reveals the paraphyly of Contumyces with respect to Loreleia, and polyphyly of the genus Cotylidia and the species Odonticium romellii. Loreleia postii (Fr.) Redhead et al., a species implicated as a parasite of the liverwort Marchantia (Kost, 1988), was excluded because blast results from the single sample supported an alliance with the Agaricales, namely, Omphalina Quél., outside the fungal order of focus for this study. Rickenella minuta comprises two strongly supported sister clades (Fig. 3A) suggesting this single morphological species is composed of two phylogenetic species.

The paraphyletic group from which the Rickenella clade is shown as derived (Fig. 3B) is dominated by lignicolous saprotrophs with the exception of the ECM lineage Coltricia. Taxa mostly considered by Larsson et al. (2006) as members of the Rickenella clade can be found in this portion of the tree and include Repetobasidium J. Erikss., Hyphoderma capitatum J. Erikss. & Å. Strid, Tsugacorticium Nakasone & Burds., Resinicium Parmasta, Mycoacia Donk, Skvortzovia Bononi & Hjortstam, and Phlebia georgica Parmasto. The Hymenochaetaceae is recovered as a paraphyletic group (85% ML support / PP > 0.95) including members of the Schizoporaceae Jülich – Basidioradulum Nobles, Hyphodontia J. Erikss., Schizopora Velen., and Xylodon (Pers.) Gray (per Index fungorum) – and taxa of uncertain position in the order – Fibricium J. Erikss., Oxyporus (Bourdot & Galzin) Donk (recently regarded as Rigidoporus Murrill), and Trichaptum Murrill.

MP ancestral state reconstruction (ASR) analyses support several switches to an ECM state from the ancestral saprotrophic state of the Hymenochaetales (Fig. 3). These are inferred to have occurred in Coltricia and in some taxa of the Rickenella clade. The exact number of transitions in the Rickenella clade is not entirely clear as it is not possible to reconstruct ancestral states with confidence in this portion of the tree. The evolution of ECM Coltricia from lignicolous, white-rot, saprotrophic ancestors is strongly supported. In the Rickenella clade several other lineages are inferred as ECM or ECM-like as well: Alloclavaria purpurea (O.F. Müll: Fr.) Dentinger & D.J. McLaughlin1, Cotylidia diphana (Appendix S1, not shown in Fig. 3), Blasiphalia Redhead, Muscinupta Redhead, Lücking & Lawrey, Loreleia marchantiae (Singer & Clémençon) Redhead et al., Contumyces rosellus (M.M. Moser) Redhead et al., Rickenella minuta (Singer & Digilio) Raithelh, and R. swartzii (Fr.) Kuyper. Putative ECM lineages that also include trophic samples that cluster as parasites (cluster 3) include Cantharellopsis prescotii (Weinm.) Kuyper, Muscinupta laevis (Fr.) Redhead, and Rickenella fibula (Appendix S1). The latter is primarily indicated as biotrophic but stable isotope samples of this species were found in all three trophic clusters. Thus, R. fibula appears capable of multiple trophic modes (see below).

Inferred trophic mode of species of Hymenochaetales

Table 5 summarizes 26 species of Hymenochaetales, including taxonomic synonyms, sampled in the global data set and Table 1, their ecology, trophic cluster, inferred tropic mode, and phylogenetic placement by clade or family in the Hymenochaetales. Generally, ecology is a useful predictor of trophic mode for terricolous and lignicolous species in the order. Almost all terricolous Hymenochaetales are ECM including Alloclavaria purpurea, species of Coltricia, Cotylidia diaphana (Schwein.) Lentz, and Rickenella minuta. An exception to this is Cotylidia undulata (Fr.) P. Karst., which is inferred as saprotrophic. All lignicolous Hymenochaetales sampled are saprotrophic. These include Odonticium romellii (S. Lundell) Parmasto, Onnia tomentosa (Fr.) P. Karst., Onnia vallata, and species of Oxyporus (recently revised as Rigidoporus; Wu et al., 2017), and Trichaptum.

Table 5. Overview of species of Hymenochaetales with stable isotope data, their ecology, trophic cluster, inferred trophic mode, and phylogenetic placement
Species Ecology Trophic cluster (No. samples) Inferred trophic mode Phylogenetic placement in Hymenochaetales
Alloclavaria pupurea Terricolous 1 (5), 3 (1) Biotrophic – ECM Rickenella clade
Blasiphalia pseudogrisella Bryophilous 1 (3), 2 (1) Biotrophic – ECM with liverworts (Blasia) Rickenella clade
Canthrellopsis prescotii (=Gerronema albidum) Bryophilous 3 (5), 1 (1) Biotrophic – other Rickenella clade
Coltricia cinnamomea Terricolous 1 (2) Biotrophic – ECM Hymenochaetaceaea
Coltricia montagnei Terricolous 1 (2) Biotrophic – ECM Hymenochaetaceaea
Coltricia perennis Terricolous 1 (4) Biotrophic – ECM Hymenochaetaceae
Contumyces rosellus Terricolous 1 (1) Biotrophic – ECM Rickenella clade
Contumyces vesuvianus Bryophilous 3 (1) Biotrophic – other Rickenella clade
Cotylidia diaphana Terricolous 1 (1) Biotrophic – ECM Rickenella clade
Cotylidia pannosa Terricolous 1 (1), 2 (1) Biotrophic – ECM or saprotrophic Rickenella clade
Cotylidia undulata Terricolous 2 (3) Saprotrophic Rickenella cladeb
Loreleia marchantiae Bryophilous 1 (2) Biotrophic – ECM with liverworts (Marchantia, Conocephalum, Lunularia) Rickenella clade
Loreleia postii Bryophilous 1 (1), 2 (1) Biotrophic – ECM or saprotrophic Agaricales
Muscinupta laevis Bryophilous 1 (2), 3 (1) Biotrophic – ECM or other Rickenella clade
Odonticium romellii Lignicolous 2 (1) Saprotrophic Rickenella clade
Onnia tometnosa (=Inonotus tomentosus) Lignicolous 2 (4) Saprotrophic Hymenochaetaceae
Onnia vallata Lignicolous 2 (1) Saprotrophic Hymenochaetaceae (lacking DNA verification)
Oxyporus cuneatus (=Rigidiporus cuneatus) Lignicolous 2 (1) Saprotrophic Oxyporus cladec
Oxyporus populinus (=Rigidiporus populinus) Lignicolous 2 (2) Saprotrophic Oxyporus clade
Phellinus nigricans Lignicolous 2 (2) Saprotrophic Hymenochaetaceae
Rickenella fibula Bryophilous 1 (9), 2 (2), 3 (12) Biotrophic – ECM or other, infrequently saprotrophic Rickenella clade
Rickenella minuta Terricolous 1 (14) Biotrophic – ECM Rickenella clade
Rickenella swartzii (=R. setipes) Bryophilous 1 (4), 3 (2) Biotrophic – ECM or other Rickenella clade
Sphagnomphalia brevibasidiata (=Gerronema cinctum) Bryophilous 2 (1) Saprotrophic Rickenella clade
Trichaptum biforme Lignicolous 2 (1) Saprotrophic Incertae sedisd
Trichaptum fuscoviolaceum Lignicolous 2 (2) Saprotrophic Incertae sedisd

Notes

  • aper Larsson et al., 2006; bper Sjökvist et al., 2012; cper Wu et al., 2017; dper NCBI taxonomy.

The trophic mode of bryophilous Hymenochaetales varies depending on the species. Two species that produce basidiomes on liverworts such as Blasia and Marchantia are supported as mycorrhizal-like sharing similar biotrophic signatures with ECM Agaricomycetes. Contumyces rosellus, most samples of Rickenella swartzii, and nearly half of our R. fibula samples are also inferred as having a mycorrhizal-like trophic status but produce basidiomes on mosses. Only one bryophilous species is inferred as saprotrophic – Sphagnomphalia brevibasidiata (Singer) Redhead et al. Other bryophilous Hymenochaetales are neither ECM, NS-NE, or saprotrophic and are likely candidates as endophytes or parasites. These include Cantharellopsis prescotii, Contumyces vesuvianus (V. Brig.) Redhead et al., and half of our samples of Rickenella fibula.

A few species are ambiguous with respect to their trophic state, due to assignment to multiple trophic clusters and low sampling, or exhibit dual trophic signatures. These include Cotylidia pannosa, Loreleia postii (one sample of which is confirmed as Agaricales), Muscinupta laevis, Lücking & Lawrey, Rickenella fibula, and R. swartzii (=R. setipes (Fr.) Raithelh. (Appendix S1).

Trophic modes do not appear to be phylogenetically conserved. Biotrophic ECM and saprotrophic taxa are distributed in both the distantly related Rickenella clade and Hymenochaetaceae supporting the contention that biotrophy evolved on multiple occasions in the Hymenochaetales. However, most species sampled in the Rickenella clade are inferred as biotrophic, either as ECM or “other”; none of these feature trophic signatures similar to biotrophic (NS-NE) Hygrophoraceae and Clavariaceae (Appendix S1; Fig. 2B–C).

Decomposition enzyme repertoire

The genomes of Rickenella fibula and R. mellea are very similar overall to those of white-rot saprotrophic Hymenochaetales in terms of presence of enzymes or suites of enzymes used in plant cell wall and lignin degradation (Appendix S4). Some exceptions include the presence of invertase, involved in sucrolytic activity and indicative of plant parasitism and endophytism, in both Rickenella genomes but absent from all other saprotrophic Hymenochaetales genomes. Xylanase, involved in the breakdown of hemicelluose, was found only in the genomes of R. fibula and Porodaedalea niemelaei.

When searching CDD with the invertase amino acid sequences of Rickenella fibula and R. mellea, three active sites were identified consistent with those reported in Parrent et al. (2009). Both appear to be extracellular invertase based on blastp searches at the National Center for Biotechnology Information (NCBI). The laccase enzyme of R. fibula and R. mellea possesses multiple active sites of cupredoxin domains of laccases similar to Tv-LCC from Trametes versicolor (L.) Lloyd (Polyporales). The ligninase enzymes (class II peroxidases) share multiple heme, substrate, Mn, and Ca binding sites similar to Mn peroxidases of Sistotrematsrum niveocremeum (Höhn. & Litsch.) J. Erikss. (Trechisporales) and other Agaricomycetes. The peroxidase enzymes are similar to cytochrome C peroxidase in Schizopora paradoxa (Schrad.) Donk (Hymenochaetales) and share the possession of numerous heme, substrate, and K+ binding sites with ascorbate and cytochrome C peroxidases. No active sites were identified by CDD in the xylanase of R. fibula; however, the enzyme was identified as a member of glycosyl hydrolase family 10 and is similar to other GH10 family proteins of other Agaricomycetes.

DISCUSSION

Revealing unknown trophic diversity in the Hymenochaetales

This is the first study to use stable C and N isotope data to predict or affirm the trophic status of numerous fungi with various ecologies in the Hymenochaetales, an order otherwise dominated by lignicolous saprotrophs. We used stable isotope evidence (Table 5) to infer at least 15 biotrophic non-lignicolous lineages of Hymenochaetales (12 of which are shown in Fig. 3), all of which are ECM or exhibit other modes of biotrophy that depart from previous trophic characterizations (Hobbie et al., 2001; Mayor et al., 2009; Seitzman et al., 2011; Birkebak et al., 2013; Sánchez-García and Matheny, 2017). Moreover, a few taxa such as moss-inhabiting species of Rickenella are characterized by multiple trophic modes, expressing ECM-like and/or possibly parasitic or endophytic signatures, or, to a lesser extent, saprotrophic modes of nutrition (Bresinsky and Schötz, 2006; Chen et al., 2018). Data were grouped in three different components or clusters according to the best-fit model in mclust. The results discussed below are not sensitive to model choice with respect to Hymenochaetales samples as somewhat less fit models cleaved the ECM and ECM-like cluster into two components reflecting high δ13N ratios among samples in the fourth cluster. Only samples of the known ECM Coltricia lineage clustered into this fourth component.

Aside from Coltricia and its close ally Coltriciella Murrill (Tedersoo et al., 2007), most biotrophic Hymenochaetales are concentrated in the Rickenella clade (Fig. 3A; Table 5). However, it is not clear how many shifts to biotrophy (clusters 1 and 3) occurred in the Rickenella clade (Fig. 3A) due to uncertainty about phylogenetic relationships in this portion of the phylogeny. Future studies will need to sample additional gene regions to ascertain the extent that evolutionary shifts to biotrophy may be phylogenetically conserved. Nonetheless, these results reinforce a general trend observed elsewhere in the Agaricomycetes that support the evolution of biotrophic lineages from saprotrophic ancestors (Martin et al., 2016), as the most recent common ancestor of the Hymenochaetales is reconstructed as saprotrophic with robust support (Fig. 3B).

Five non-lignicolous species in the Rickenella clade, viz. Alloclavaria purpurea, Contumyces rosellus, Cotylidia diaphana, Loreleia marchantiae, and Rickenella minuta, possess stable isotope signatures similar to those of ECM Agaricomycetes. Studies on the morphology of roots or rhizoids colonized by these fungi are needed to confirm, or have confirmed in some cases (Redhead, 1981), the presence of anatomical features consistent with an ECM habit. Nevertheless, Alloclavaria purpurea is most likely an ECM lineage with Pinaceae. In general the literature suggests A. purpurea is a moss associate (Dentinger and McLaughlin, 2006). However, three collections from North America were collected on sandy soil under pines, on ground among moss near fir, and on a moss-covered bank under ericaceous shrubs, bayberry, and pines. Corner (1950) noted that A. purpurea fruits exclusively near coniferous trees. Walker et al. (2012) recovered ITS matches of A. purpurea from ECM root tips of Douglas fir, but interpretation of stable isotope data produced in that study was ambiguous depending on the method of analysis used. This led Tedersoo and Smith (2013) to regard A. purpurea as a likely endophyte or saprobe. Our results do not support the suggestion that this species is endophytic or saprotrophic because five of six samples of A. purpurea from North America and Europe consistently produced a stable isotope ECM signature. We thus conclude A. purpurea is a novel ECM lineage in the Hymenochaetales.

ITS sequences of Rickenella minuta from basidiomes match those sampled from ECM root tips of Nothofagaceae in Argentina (Nouhra et al., 2013; data not shown). In addition, basidiomes of R. minuta are produced directly on soil in ectotrophic forests, although at times they can also be found among mosses (P.B. Matheny and G.R. Smith, personal observation). Thus, at least two lines of evidence support the ECM status of R. minuta – root tip molecular data and stable isotope signatures. This confirms the third known ECM lineage in the Hymenochaetales.

Contumyces rosellus (as Omphalina rosella (M.M. Moser) M.M. Moser) was documented by Redhead et al. (1995) from variable habitats, but it most commonly occurs on soil, generally in lawns among grasses or bryophytes. Stable isotope data from one sample of this species suggests an ECM status as well; however, additional collections from different populations need to be assayed to confirm this mode of nutrition.

The ecology of the genus Cotylidia is poorly understood, and prior evidence regarding its trophic status was lacking (Kout and Zíbarova, 2013). Species in the genus have been variably described as terrestrial and producing basidiomes on mineral soils often among mosses or as possibly bryophilous (Redhead et al., 2002). However, bryophilous associations are not constant among the few species described in the genus (Kout and Zíbarova, 2013). Moreau and Audet (2008) suggested C. carpatica (Pilát) Huijsman is a parasite of mosses, but C. pannosa (Sowerby) D.A. Reid has not been reported in association with mosses. We produced stable isotope data from three different species of Cotylidia. One of these, C. diaphana, is terricolous and bears an ECM signature (Table 5). However, C. undulata (Fr.) P.Karst., which is also terricolous, is inferred as saprotrophic. Two samples of C. pannosa were inconsistent regarding trophic assignment (one ECM, one SAP). Additional samples, including from different species, are required to determine how consistent and well supported such inferences are. Note that Cotylidia is not monophyletic in our phylogenetic tree (Fig. 3A). Accordingly, variation in trophic modes among Cotylidia species is not unexpected.

Loreleia marchantiae also groups in the predominantly ECM cluster 1. This is consistent with observations that the fungus penetrates rhizoids of the thalloid liverwort Marchantia L. (Bresinsky and Schötz, 2006), of which it was considered a parasite (Kost, 1988). However, Bresinsky and Schötz (2006) suggested the fungus enables N exchange from cyanobacteria to its liverwort associate. If L. marchantiae were acquiring N from a live autotroph, we would expect N isotope samples of this fungus to group it in trophic cluster 3. This, however, is not the case, as its N signatures are more like those of saprotrophs (Table 4). Loreleia marchantiae is found producing basidiomes on liverworts in the genera Marchantia, Conocephalum Hill, and Lunularia Adans., in particular on live M. polymorpha L. (Kost, 1988; Knudsen and Vesterholt, 2012). Marchantia is also associated with arbuscular-mycorrhizal (AM) fungi in the genus Glomus (Kottke and Nebel, 2005; Russell and Bulman, 2005), and mycorrhizal-like exchange of nutrients (P, N) has been reported in Marchantia palaeacea Bertol (Humphreys et al., 2010). However, at this time there is no evidence to support a N-exchange mutualism between L. marchantiae and its liverwort associate.

Most isotope samples of the bryophilous Blasiphalia pseudogrisella (A.H. Sm.) Redhead, Muscinputa laevis, and Rickenella swartzii also feature ECM trophic signatures. This would suggest these fungi receive photosynthates as C input and exchange N in return in a nutrient exchange mutualism as is typical for the ECM symbiosis (Smith and Read, 2008). Indeed, Kowal et al. (2018) confirm such a mutualism between liverworts and Ascomycota. However, bryophilous Hymenochaetales that group in the predominantly ECM cluster 1 (Table 4) share similar N signatures with saprotrophic fungi. This pattern suggests that bryophilous Hymenochaetales in this cluster may not be engaging in N exchange. As such, hypotheses that these fungi are commensals cannot be dismissed; unless it can be shown that they convey other fitness benefits to their bryophyte associates (Davey and Currah, 2006). Parasitism (Redhead et al., 2002; Larsson et al., 2006) does not seem likely if the fungi are passively receiving C in the form of glucose from their bryophyte associates (Parrent et al., 2009). Redhead (1981) did observe appressoria produced by Blasiphalia pseudogrisella, anatomical structures typical of plant pathogens, which penetrate the rhizoids of the liverwort Blasia pusilla L. However, Kost (1988) suggests this structure is analogous to those produced by ECM fungi, so-called “palmetto-structures” and refers to infected caulonemata and rhizoids of bryophytes as “mycorrhizoids”. If B. pseudogrisella were found to lack invertase involved in active breakdown of sucrose to glucose, then this would support our conclusion of an ECM or ECM-like signature for this fungus.

Muscinupta laevis occurs on live mosses, especially Polytrichum Hedw. (Redhead et al., 2002). Ryvarden (2010) and Vizzini (2010) considered the species to be a moss parasite. Stable isotope data we analyzed (Table 2) confirm a biotrophic mode of nutrition in this species, either as ECM-like (cluster 1) or as a parasite or endophyte (cluster 3) (Table 5). More sampling from diverse locations is needed to confirm these results. Rickenella swartzii is a third biotrophic bryophilous species that fits in this biotrophic ECM-like functional group. It has been found on a wide range of bryophytes (Bresinsky and Schötz, 2006) and considered a parasite invading chloronemata or caulonemata of mosses forming “lignituber-like” structures similar to ericoid mycorrhizas (Kost, 1988). The former authors suggested this species is most likely saprotrophic or forms endomycorrhizas (biotrophic). Stable istotope data affirm a biotrophic signature with four of six samples consistent with an ECM-like state (Fig. 3A) and two in cluster 3 indicative of parasitism or endophytism (Appendix S1; Table 5).

Kost (1988) and Bresinsky and Schötz (2006) predicted contrasting modes of nutrition for Rickenella fibula, which is of particular interest since numerous samples of this species support multiple trophic modes. Kost (1988) suggested R. fibula is a parasite (biotroph) forming “lignitubers” on moss rhizoids as in R. swartzii, whereas Bresinsky and Schötz (2006) considered R. fibula as a saprotroph but could not dismiss an “endomycorrhizal” ecology. Similar to R. swartzii above, but with denser stable isotope sampling, we found evidence that R. fibula is a biotroph—either ECM-like (cluster 1) or parasitic or endophytic (cluster 3). Only two of 23 stable isotope samples support a saprotrophic state for R. fibula. Both biotrophic states do not appear to co-vary with phylogeny (Fig. 3A) and are consistent with a report that R. fibula can be found throughout living and senescent moss gametophytes exhibiting dual trophic modes (Chen et al., 2018). However, R. fibula does not appear to be saprotrophic to a major extent in contrast to the suggestion made by Chen et al. (2018). Redhead (1981) observed peg-like haustoria produced by R. fibula on the rhizoids of a moss indicative of parasitism, behavior not inconsistent with our stable isotope results.

Some authors have suggested that the saprotrophic Phellinus igniarius (L.) Quél. can colonize young roots of Norway spruce producing hyphal structures similar to a Hartig net and thus are capable of “facultative biotrophy” switching between saprotrophic and biotrophic nutritional modes (Smith et al., 2017). We produced stable C and N isotope data from two samples of Phellinus nigricans (Fr.) P. Karst., a member of the P. igniarius complex, but found no evidence of a biotrophic signature in this species (Appendix S1; Table 5). Indeed, stable isotope data confirm a saprotrophic mode of nutrition in the lignicolous species of Phellinus, Odontcium, Onnia, Oxyporus (=Rigidoporus), and Trichaptum. Among non-lignicolous Hymenochaetales, saprotrophic signatures are suggested for the bryophilous Sphagnomphalia brevibasidiata (Singer) Redhead, which occurs on live Sphagnum (Redhead et al., 2002), and the terricolous Cotylidia undulata, discussed above. Additional stable isotope results are needed from Sphagnomphalia brevibasidiata that confirm or reject the saprotrophic signature from our single sample.

Phylogenetic relationships in the Hymenochaetales and the Rickenella clade

Future strategies that increase both gene and taxon sampling are required to resolve relationships of major groups within the Hymenochaetales. Increasing taxon sampling (compared to that shown in Fig. 3) reduced phylogenetic resolution overall in the Hymenochaetales (590 tips represented by at least one gene region), probably due to the large amount of missing data. Inferences about phylogenetic relationships drawn from whole genome data at MycoCosm (Grigoriev et al., 2014) suggest Rickenella is sister to the rest of the Hymenochaetales, consistent with the topology shown here (Fig. 3). However, the taxa forming a grade relative to Rickenella and other Hymenochaetales (Rickenella clade p.p.; Fig. 3B) are lacking whole genome data. If the ML phylogenetic hypothesis is correct, then biotrophic lineages of Hymenochaetales are evolutionarily derived.

Future research efforts should target sequencing whole genomes from ecologically diverse Hymenochaetales such as Alloclavaria purpurea (terricolous ECM), Loreleia marchantiae (bryophilous ECM-like), Rickenella minuta (terricolous ECM), Coltricia or Coltriciella (terricolous ECM), Cantharellopsis prescotii (bryophilous parasite), and Blasiphalia pseudogrisella (bryophilous ECM-like). Phellinus igniarius could be targeted as well (Smith et al., 2017). At present, only two genomes are available from non-lignicolous Hymenochaetales: Rickenella fibula and R. mellea, species that appear capable of multiple trophic modes (ECM-like and parasitic or endophytic) discussed below. Furthermore, the phylogenetic affinities of Loreleia postii need to be evaluated based on additional taxon sampling as our single result suggests an affiliation of this species with Omphalina in the Agaricales.

Genomic traits of bryophilous Hymenochaetales

Different fungal trophic modes are also characterized by different genomic traits, such as presence or number of enzymes involved in cellulose and lignin degradation (Read et al., 2004; Talbot et al., 2008, 2015; Kohler et al., 2015). Biotrophic ECM fungi produce plant cell wall degradative enzymes such as cellobiohydralase, cellulase, chitinase, polygalacturonase, and xylanase, in addition to lignin degradation enzymes such as laccase and peroxidases. Biotrophic ericoid mycorrhizal fungi often produce plant cell wall degradative enzymes arabinosidase, galactosidase, and mannosidase, in addition to those found in ECM fungi. Ericoid mycorrhizal fungi also feature lignin degradative enzymes laccase, lignase, and class II peroxidases. However, mycorrhizal fungi in general contain a much lower number of these enzymes compared to saprotrophs, and ECM fungi lack the plant cell wall degradative enzyme glucanase (Zhao et al., 2013; Kohler et al., 2015). Biotrophic parasitic fungi vary in their functional repertoire depending on whether the fungus is a plant or animal pathogen (Zhao et al., 2013). Invertases, for example, enzymes with sucrolytic activity, are found in plant parasitic fungi such as Heterobasidion irregulare Garbelotto & Otrosina (Olson et al., 2012) and endophytes in general but have been lost in animal parasites and are absent in most mycorrhizal lineages (Parrent et al., 2009; Martin et al., 2016; Strullu-Derrien et al., 2018). Since the study by Parrent et al. (2009), genomic studies generally support a negative correlation between invertase presence and most ECM fungi. Tuber, an ECM lineage in the Ascomycota (Martin et al., 2010), and Sebacina incrustans (Pers.) Tul. & C. Tul., an ECM lineage (Parrent et al., 2009; Weiß et al., 2016) are two exceptions.

Given that bryophytes do not produce wood, it would stand to reason there would be a lack of selection to maintain lignin degradative enzymes in biotrophic Hymenochaetales under the assumption they are derived from white-rot ancestors. Contrary to our expectations, the genomes of Rickenella are characterized by the presence of a suite of enzymes used to degrade lignin (Lundell et al., 2010), cellulose, and hemicelluose (xylanase in R. fibula). Bryophytes are known to contain lignin-like polymers but not lignin itself (Ligrone et al., 2008), which could explain the maintenance of lignin degradation in bryophilous Hymenochaetales. The presence of invertase in the genomes of R. fibula and R. mellea, and its absence in white-rot Hymenochaetales genomes, is more consistent with a non-mycorrhizal biotrophic lifestyle, such as that of a parasite or endophyte (Parrent et al., 2009), as exemplified by the assignment of these taxa in cluster 3 (Fig. 2; Table 4). It remains to be explained why some bryophilous Rickenella samples cluster more closely with known ECM fungi (Table 4) based on similar C utilization patterns, other than these Rickenella species are capable of multiple trophic modes (Olson et al., 2012; Smith et al., 2017; Chen et al., 2018).

ACKNOWLEDGEMENTS

The authors thank the Department of Ecology and Evolutionary Biology at the University of Tennessee, in particular Ken McFarland, and the Missouri Mycological Society and the University of Tennessee Botany Excellence Fund for financial support. The U.S. Department of Energy Joint Genome Institute produced the whole genome and transcriptome for Rickenella fibula HBK330-10. We acknowledge Igor Grigoriev, Joseph Spatafora, and Kerry Barry for their support as participants in the 1000 Fungal Genomes Program. The following herbaria provided specimen loans: CMMF, CORD, FLAS, H, TENN, and TUR. We also thank Renée Lebeuf, Steve Trudell, and Mike Wood for sharing specimen-vouchers and Meg Staton for comments on an earlier version of this manuscript. Mike Wood is particularly acknowledged for the photos shown in Fig. 1. We thank the University of Tennessee Genomics Core for sequencing assistance, Jacob Edwards for laboratory assistance, and staff at the University of New Hampshire Stable Isotope Lab for sample preparation. Marisol Sánchez-García, Brian Looney, and Brian O'Meara provided assistance with phylogenetic and related methods. We thank Joshua Birkebak for pointing out and correcting nomenclatural concerns. Portions of this study were supported by funds from NSF DEB-1354802 awarded to MES and PBM. The basis of this work was originally published as a Master's thesis, copyright Hailee B. Korotkin 2017, University of Tennessee. We thank the associate editor and reviewers for their helpful comments and suggestions to earlier versions of this manuscript.

    DATA ACCESSIBLITY

    DNA sequence data, alignments, and phylogenetic trees are deposited at GenBank and TreeBASE (21259; www.treebase.org) and are also available at http://mathenylab.utk.edu/Site/Alignments_%26_Data_Sets.html.

    Note

  1. 1 The citation of Clavaria purpurea Fr. (Syst. mycol. (Lunndae) 1: 480. 1821) by Dentinger and McLaughlin (2006) as the basionym for Alloclavaria purpurea, type species of Alloclavaria, is an indirect reference to Clavaria purpurea O.F. Müll (1780), cited by Fries, as permitted by ICN Art. 40.3 (Turland et al., 2018). Direct citation of the basionym is Clavaria purpurea O.F. Müll: Fr. Thus, Alloclavaria purpurea should be cited as A. purpurea (O.F. Müll: Fr.) Dentinger & D.J. McLaughlin. The name, because it is sanctioned in Fries’ 1821 work, is conserved against the earlier Clavaria purpurea Schaef. (1774).