Fungal Diversity Arizona gasteroid fungi I: Lycoperdaceae (Agaricales, Basidiomycota) Bates, S.T.1*, Roberson, R.W.1 and Desjardin, D.E.2 1School of Life Sciences, Arizona State University, Tempe, Arizona 85287, USA 2Department of Biology, San Francisco State University, 1600 Holloway Ave., San Francisco, California 94132, USA Bates, S.T., Roberson, R.W. and Desjardin, D.E. (2009). Arizona gasteroid fungi I: Lycoperdaceae (Agaricales, Basidiomycota). Fungal Diversity 37: 153-207. Twenty-eight species in the family Lycoperdaceae, commonly called ‘puffballs’, are reported from Arizona, USA. In addition to widely distributed species, understudied species (e.g., Calvatia cf. leiospora and Holocotylon brandegeeanum) are treated. Taxonomic descriptions and illustrations, which include microscopic characters, are given for each species, and a dichotomous key is presented to facilitate identification. Basidiospore morphology was also examined ultrastructurally using scanning electron microscopy, and phylogenetic analyses were carried out on nrRNA gene sequences (ITS1, ITS2, and 5.8S) from 42 species within (or closely allied to) the Lycoperdaceae. Key words: Agaricales, euagarics, fungal taxonomy, gasteroid fungi, gasteromycete, Lycoperdaceae, puffballs. Article Information Received 22 August 2008 Accepted 25 November 2008 Published online 1 August 2009 *Corresponding author: Scott T. Bates; e-mail: [email protected] Introduction Agaricales, Boletales, and Russulales. Accordingly, a vigorous debate concerning the Lycoperdaceae Chevall. has traditionally path of evolution within these series ensued been placed within the Gasteromycetes; a class (e.g., see Singer, 1962; Heim, 1971; Thiers, that workers have long suggested did not 1984). represent a natural assemblage of fungi (Heim, With the close of the twentieth century, 1971). By the early twentieth century, molecular phylogenetic analyses using systematic mycologists began to produce sequence data from the nuclear genes encoding alternative taxonomies, based in evolutionary ribosomal RNA presented convincing evidence theory, which included members of the that not only were the Gasteromycetes a Gasteromycetes among basidiomycetous agari- polyphyletic grouping, but that gasteroid forms coid (mushroom-like) fungi. For example, the had multiple independent origins from within classification of Heim and colleagues (1934) the Agaricomycotina (Hibbett et al., 1997). did not recognize the Gasteromycetes, but Taxa that were formerly placed within the rather placed both the Gastérales, which Gasteromycetes are now typically referred to as included Lycoperdaceae, and the Agaricales, ‘gasteroid’ fungi (e.g., Pegler et al., 1995; which exclusively contained agaricoid species, Binder and Bresinsky, 2002) or ‘gastero- in the Homobasidiés (Basidiomycetes). mycetes’ (e.g., Reijnders, 2000). Molecular Throughout the middle of that century, several phylogenetic analyses continue to support the evolutionary ‘series’ were proposed that linked position that Lycoperdaceae is a gasteroid taxa within the Gasteromycetes, secotioid lineage which has arisen from within the species (mushroom-like fungi having an lepiotoid mushrooms of the Agaricales (Hibbett unexpanded pileus that encloses malformed et al., 1997; Krüger et al., 2001; Moncalvo et lamellae), and agaricoid members of the al., 2002; Vellinga, 2004). It is also interesting 153 to note that as early as the mid-1960s Kreisel in Arizona. These biomes include xeric alpine (1967a) suggested this relationship and pointed tundra at extremely high elevations (above out similarities between Lepiota lycoperdoides 3500 m); mesic subalpine and montane conifer Kreisel [=Cystolepiota fumosifolia (Murrill) forests at high elevations (ca. 2000–3500 m); Vellinga] and Lycoperdon species. Recent arid adapted chaparral in the mid elevations (ca. studies (Krüger et al., 2001; Krüger and Gargas, 1050–2000 m); and xeric desertscrub in the 2008; Krüger and Kreisel, 2003; Larsson and lower elevations (below 1050 m). It is also Jeppson, 2008) based on nrDNA sequence data notable that boundaries of four major desert indicate that the type genus Lycoperdon is ecoregions (Chihuahuan, Great Basin, Mojave, polyphyletic and question the stability of and Sonoran) come in contact with some boundaries between the genera Bovista Pers., portion of the state. Calvatia Fr., Handkea Kreisel, Lycoperdon Members of the Lycoperdaceae have a Pers., Morganella Zeller, and Vascellum F. widespread distribution and are commonly Šmarda. found in temperate, arid, and tropical climates Prior to the twenty-first century, mention (Pegler et al., 1995). A large part of southern of puffballs from the state of Arizona in Arizona is occupied by the Sonoran Desert mycological literature is limited. Most which is arid with considerable tropical significant is Demoulin’s doctoral thesis (1972), influences, and the mid-elevations of the state which includes records of four Lycoperdon are covered by large tracts of temperate forest species occurring in the state. Other publica- (Brown, 1994). Considering these factors, it is tions (Morgan, 1890; Smith, 1974; Demoulin, not surprising that a biologically diverse and 1993) include only single records. In addition species rich assortment of these gasteroid fungi to these publications, numerous specimens of are present within the state. Arizona puffballs have been collected and This monograph documents 28 species of deposited in herbaria by the few field puffballs from nine of the biotic communities mycologists who have worked in the state, present in Arizona. Each of these gasteroid namely W.H. Long, R.L. Gilbertson, and J.S. species in the Lycoperdaceae is described as States. To supplement that material, specimens well as illustrated, and SEM micrographs have also been collected by one author (STB) typifying spore morphology are presented. over the course of four seasons in the field. Of Ribosomal RNA gene sequences were also the 28 species presented here, 22 were obtained from several of the specimens considered to be new records for the state examined as part of this study. These data were (Bates, 2004) and were subsequently published supplemented with rDNA sequences obtained as such in a checklist of Arizona macrofungi from GenBank, and were used in subsequent (Bates, 2006). In that publication, Calvatia phylogenetic analyses. In addition to this bicolor (Lév.) Kreisel was reported for the first publication, the reader may refer to Bates (2004) time from North America and Disciseda for further discussion of morphology and verrucosa G. Cunn. was reported for the first taxonomy of the Lycoperdaceae. time from the United States. In addition to those records, this publication also documents Materials and methods rare and/or understudied species, such as Calvatia cf. leiospora Morgan and Holocotylon Monography brandegeeanum Lloyd. Macro-morphological data are based on Arizona possesses a wide range of habitat. original descriptions of herbarium material or Brown (1994) outlined 27 major biotic collections made in the field. Micro- communities (biomes) that occur in the morphological data were gathered using an Southwestern United States and Northwestern Olympus BH2 conventional bright field Mexico. Of these, 14 biomes with 2 microscope. Illustrations of spores, capillitial subdivisions of Sonoran desertscrub are found threads, endoperidial hyphae, and exoperidial 154 Fungal Diversity cells were drawn using an Olympus U-DA Herbarium acronyms followed the Index drawing attachment. For microscopic examina- Herbariorum (Holmgren and Holmgren, 1998). tion, all material was first infiltrated with 95% aqueous ethanol before being mounted on glass Basidiospore ultrastructure slides. Observations of characters, such as Observations of spore ultrastructure were spore ornamentation and capillitial structure, made on a XL30 (FEI, Hillsboro, OR, USA) were made from glebal material mounted environmental scanning electron microscope separately in H O (with a small amount of (ESEM) operated under high vacuum in 2 detergent added as a surfactant), 3% aqueous scanning mode, at an accelerating voltage of 3– KOH, and/or Lactophenol Cotton Blue. In 5 kV. Spores were fixed in 2% aqueous addition, a small portion of the exoperidium glutaraldehyde after re-hydration from 24–48 was mounted in Melzer’s reagent to observe hours in deionized H O. After fixation, spores 2 structure of exoperidial hyphae and/or were washed in a phosphate buffer solution and sphaerocysts. Measurements were made from then subjected to a dehydration series from material mounted in KOH at 1000× using an 70% to 100% ethanol. After dehydration, ocular reticle with units calibrated to an optical spores were critical-point-dried using a CPD micrometer. All spore measurements include 020 (Balzer Union, Liechtenstein). Once ornamentation, and spore statistics include: X¯, removed from the critical-point-dryer, the the arithmetic mean of the spore length by spores were affixed to carbon coated aluminum spore width (± SD) for n spores measured; Q, specimen holders with poly-l-lysine and sputter the quotient of spore length and spore width in coated with palladium-gold for 5 min at 10 any one spore, indicated as the range of milliamps in a Technics Hummer II (Anatech variation in n spores measured; Q the mean of LTD., Hayward, CA, USA). m Q values. Standard classical taxonomic and DNA extraction descriptive methodologies for gasteroid fungi A small portion (15–20 mg) of glebal were used (see Calonge, 1998; Cunningham, material was removed from the interior parts of 1944; Demoulin, 1972; Moyersoen and dried specimens, placed in 2.0 ml micro- Demoulin, 1996; Pegler et al., 1995). The centrifuge tube, and ground using a small classification of biotic communities cited sterile plastic pestle. Genomic DNA was then followed the system outlined by Brown (1994). extracted using the protocol of a commercial kit Color descriptions adhered to the nomenclature (Qiagen DNeasy Plant Mini Kit, Germantown, and codification system of Kornerup and MD, USA). The extracted DNAs were eluted Wanscher (1967). Additional color names, from the mini spin column in 100 μl of ‘AE’ indicated by quotations, followed the buffer heated to 65oC. To precipitate the DNA, nomenclature given by Ridgway (1912). Taxa 1/10 volumes of 3M sodium acetate were added were arranged alphabetically and taxonomy, for in order to adjust the concentration of the most part, followed that outlined in Kirk monovalent cations. The extracted DNA was and colleagues (2001) and CABI Index then precipitated in 100% ethanol and Fungorum (http://www.indexfungorum.org). recovered by centrifugation. After being rinsed Abbreviations for the authors of fungal names in 70% ethanol, the solution was again are consistent with Kirk and Ansell (1992) and centrifuged, the supernatant removed, and the subsequent additions found in the CABI Index resultant pellet was allowed to air dry for 2–3 of Fungi Supplement: Authors of Fungal Names, hours. The DNA pellet was then re-suspended which is available online (http://www.in in 25 μl of TE Buffer. dexfungorum.org/AuthorsOfFungalNames.htm). 155 PCR amplification of fungal nrRNA genes maximum parsimony (MP) algorithms, in each Internal transcribed spacers (ITS1 and case with a bootstrap analysis of 1000 replicate ITS2) and 5.8S region of the nuclear ribosomal datasets in order to assess clade stability. RNA gene were targeted for PCR amplification Ribosomal RNA gene sequence data using the primers ITS4 and ITS5 of White and (ITS1, ITS2, and 5.8S) from 41 taxa within the colleagues (1990). Each reaction (25 μl) Lycoperdaceae were included in phylogenetic contained: 10–13 μl sterile Milli-Q H O, 2.5 μl analyses, as was Mycenastrum corium (Guers.) 2 50% glycerol, 2.5 μl 10X Taq I buffer, 2 μl Desv. in order to evaluate its position relative to dNTPs (10 mM), 1.25 μl of each primer (10 the clade corresponding to the family. In μM), 0.75 μl MgCl (50 mM), 0.1 μl Invitrogen addition, 8 taxa from the Agaricales were 2 Platinum Taq enzyme (2–5 units/μl), and 2–4 μl included to assist in assessing the placement of DNA template. The PCR reactions were then all the gasteroid taxa represented in this study. carried out in a DYAD Peltier Thermal Cycler Limacella glioderma (Fr.) Maire [= Limacella (Bio-Rad Laboratories, Hercules, CA, USA) delicata var. glioderma (Fr.) Gminder, using the following cycling parameters: initial Amanitaceae, Agaricales] was selected as the denaturing at 92oC for 2 min., 35 cycles at 92oC out-group to root the phylogenetic trees as for 45 s, 55oC for 30 s, and then 72oC for 1 min., Limacella Earle is not contained within the with a final extension at 72oC for 7 min. PCR Lycoperdaceae or the closely allied Agarica- products were checked for quality on 1% ceae Chevall. (Moncalvo et al., 2000). agarose gels and the final concentration was measured on a NanoDrop ND-1000 Spec- Results and Discussion trophotometer (Thermo Fisher Scientific, Wilmington, DE, USA). Automated sequencing Molecular pylogeny of the Lycoperdaceae was carried out at the ASU School of Life Consensus trees, NJ (not shown) and MP Sciences DNA Laboratory using both the (Fig. 1), did not strongly support monophyly forward and reverse primers in separate for any genus within the Lycoperdaceae, reactions. although clades corresponding to Bovista (with the inclusion of Arachnion album Schwein. in Phylogenetic analyses the NJ analysis), Calvatia [including a well Complementary rDNA sequences were supported clade corresponding to the genus compiled and manually edited on Sequencher Langermannia Rostk.: C. bicolor (Lév.) Kreisel, 4.1 software (Gene Codes Corp., Ann Arbor, C. gigantea (Batsch) Lloyd, and C. pachyderma MI, USA). A total of 21 consensus nrRNA (Peck) Morgan], and Vascellum were recovered. gene sequences were then submitted to The inclusion of Arachnion album, Disciseda GenBank (see Fig. 1 for accession numbers). candida (Schwein.) Lloyd, Holocotylon Additional sequences from closely related brandegeeanum, and Mycenastrum corium species within the Lycoperdaceae were culled within Lycoperdaceae was well supported by from GenBank (www.ncbi.nlm.nih.gov) after a our analyses (98–100% bootstrap values). Thus, BLASTn search (http://blast.ncbi.nlm.nih.gov/). the family could be circumscribed as a Their accession numbers are also cited in Fig. 1. monophyletic group united by the gasteroid Sequences were aligned using the CLUSTAL habit. It should be noted, however, that the W module in the MEGA 4 software package basal member, M. corium, exhibits several (Tamura et al., 2007) and were then manually morphological characters (e.g., spore corrected. The subsequent alignment was ultrastructure; see Bronchart and Demoulin, submitted to TreeBASE (#S2247, http://www. 1973) that are markedly different from the treebase.org/). Phylogenetic trees were remaining members of the Lycoperdaceae clade constructed with the neighbor-joining (NJ) and and is typically included in a separate 156 Fungal Diversity Lycoperdonniveum(DQ112563) LLyyccooppeerrddoonncnfi.vdeeucmipi(eDnQs1(D12Q516132)586) 63 LLyyccooppeerrddoonnmcfa.mdemciifpoiremnes((ADJQ213172652816)) 63 LLyyccooppeerrddoonnmmaomllmei(fEorUm8e33(A66J22)3*7621) LLyyccooppeerrddoonnpmuolcllhee(rEriUm8u3m36(6E2U)8*33663)* LLyyccooppeerrddoonnrpimuluclhaeturrmim(EumU8(3E3U68634)3*663)* LHyacnodpkeeardeoxncipriumlifuolramtuism(A(JE6U1873439616)4)* HHaannddkkeeaafeuxmcoipsualif(oErUm8is33(A65J56)1*7491) HLyacnodpkeeradofnumlaomsbain(EonUii8(3D3Q651152)*576) LLyyccooppeerrddoonnelacmhibniantounmii((ADJQ21317622527)6) LLyyccooppeerrddoonnneicvheiunmatu(DmQ(1A1J22539796)22) LLyyccooppeerrddoonnncivf.eduemrm(DoQxa1n1t2h5u9m9)(FJ438478)* 93 LLyyccooppeerrddoonnucmf.bdreinrmumox(aEnUt8h3u3m66(F5)J*438478)* 93 LHyacnodpkeeardlyocnopuemrdboriidneusm(E(UEU83833635666)5*)* HBoavnidsktaeadelyrmcoopxearndtohiade(DsQ(E1U128537396)56)* 68 BHoovloisctaotdyelormnobxraanntdheag(eDeQan1u1m25(7E9U)833660)* 68 HBoovloiscteoltlaylroandibcaratan(dAeJg2e3e7a6n2u4m) (EU833660)* 99 BHoavnidskteelalaurtaridfoicrmatias((AEJU283373662549))* 99 HHaannddkkeeaasuutrbifcorremtaisce(aEU(E8U3833635695)*8)* 98 HHaannddkkeeaassuubbccrreettaacceeaa((EEUU883333665578)*)* 9682 HVaasncdekleluamsuibncterermtaecdeaiu(mEU(E8U3386335676)*7)* 62 VVaasscceelllulummpirnatteernmseed(AiuBm06(7E7U258)33667)* VMaosrgcealnluemllapsruabteinncsaern(aAtBa0(A67J2732756)26) MLyocrogapneerdllaonsumbainrcgainrnaatutam(A(EJU238736362661))* 70 LLyyccooppeerdrdoonnpemrlaartugmina(AtuJm237(E6U278)33661)* 70 LCyaclvoapteiardgoignapneterlaat(uAmJ6(1A7J429327)627) 96 CCaallvvaattiiaagbiigcaonlotera (E(AUJ863137645912)*) 96 CCaallvvaattiiaapbaiccohlyodre (rEmUa83(E3U658313)*653)* CCaallvvaattiiaacpfa.clehiyodsepromraa((EEUU883333665523)*)* 99 CCaallvvaattiaiaccyfa.tlheifioorsmpiosr(aAJ(E48U68836366)52)* 99 99 CCaallvvaattiiaafcrayagtilhisifo(ArmJ4is86(A8J7418)6866) 9990 CBoavlvisattaiapflruamgbiliesa(A(AJJ428367867219)) 90 BBoovviissttaappalulumdboesaa((AAJJ223377663209)) BBoovviissttaappuasluildlao(sAaJ(2A3J72633716)30) BBoovviissttaappoulsyimllaor(pAhJa23(A7J623317)613) BBoovviissttaaapeoslytimvaolrisph(DaQ(A1J122367206)13) 98 99 98 66 BBoovviissttaaaaeessttivivaalilsis(D(EQU181323662500))* 99 66 BBoovviissttaaaaeessttiivvaalliiss((FEJU48338346757)0*)* BDoisvciissteadaaecsatinvdaildisa((FEJU483383467574))** 98 DLyiscocpiseerddoancpaynrdifoidrmae(E(AUJ823337665240))* 98 LAyraccohpneirodnoanlbpuymrif o(ErmUe83(A36J4293)7*620) AMryaccehnnaiosntraulmbucmo r(iEuUm8 (3E3U684393)*666)* MMyaccreonleapsitortuamclecloanriduimi(A (FE4U8823833686)6)* MAgaacrriocluespcioatmapcelesltarinsd(iDi(QA1F8428523833)8) 61 AEgndaoripctuyschcuammpaegsatrriicso(iDdeQs1(8A2F543832)837) 61 98 ECnhdlooropptyhcyhlluummamgoalyribcdoiitdeess(A(AFF448822883367)) 98 CLehploiorotaphcryisllutamtam(AoFly3b9d1it0e4s8()AF482836) LCeypsitootleapcioritsatactyas(tiAdFio3s9a1(0C4C8U)85333) CLeypsitootlaepciloytpaeocylasrtiiadi(oAsYa1(7C6C3U618)5333) LLeimpaiocteallcalygplioeodelarrmiaa((AAYY117766346513)) Limacellaglioderma(AY176453) Fig. 1. Strict consensus maximum-parsimony tree based on the analysis of nrRNA gene sequences (ITS1, ITS2, and 5.8S) of gasteroid species within (or allied to) the Lycoperdaceae and other agaricoid fungi. GenBank accession numbers are given in parentheses. Sequences produced as a part of this study are indicated by a single asterisk (*) and those originating from Arizona specimens are in bold. Bootstrap values are expressed as percentages of 1000 replications, and clades supported by bootstrap values of 60% or more are indicated. 157 monotypic family, Mycenastraceae Zeller. The species. Our results were also similar to those distinction between Lycoperdaceae and of Larsson and Jeppson (2008), which included Mycenastraceae may, however, be a mute point a much larger sampling of Lycoperdaceae as contemporary workers tend to recognize species from northern Europe and included only a monophyletic Agaricaceae that includes sequence data for both ITS and LSU rDNA. In the above mentioned gasteroid taxa (Metheny order to accommodate a monophyletic Lycoper- et al., 2006). don, those authors applied a very broad concept A core Lycoperdon clade that included of the genus and recognized clades more or less L. cf. decipiens Durieu & Mont., L. cf. corresponding to traditional segregate genera dermoxanthum Vittad., L. echinatum Pers.: (i.e., Vascellum and Morganella) as subgenera. Pers., L. lambinonii Demoulin, L. mammiforme Many of the clades recovered in that study, Pers.: Pers., L. molle Pers.: Pers., L. niveum however, were not well supported, which Kreisel, L. pulcherrimum Berk. & M.A. Curtis, ultimately led the authors to conclude that and L. rimulatum Peck ex Trel. was recovered structure within the Lycoperdaceae is still in both analyses; however, Handkea unresolved. Although convincing arguments for excipuliformis (Scop.: Pers.) Kreisel and H. new taxonomies based on molecular phylo- fumosa (Zeller) Kreisel also clustered with genetic analyses can be made, taxonomic these species. Apart from this clade was revisions should not be formally considered another that contained L. perlatum Pers.: Pers. until the Lycoperdaceae clade is more fully and L. marginatum Vittad. as well as species in resolved. This may ultimately depend on the genera Morganella and Vascellum. multiple gene phylogenies as well as other lines Handkea (see Kreisel, 1998) was clearly a of evidence (e.g., morphological and polyphyletic entity not closely related to ultrastructural data). Until relationships within Calvatia; its members being allied rather to the Lycoperdaceae clade are more certain, we Lycoperdon species. chose to follow a more traditional approach to The placement of L. pyriforme Schaeff.: the group; recognizing genera such as Bovista, Pers. was not clearly resolved in our study; Calvatia, Lycoperdon, and Vascellum in the however, the MP (Fig. 1) and NJ (not shown) classical sense. consensus trees did suggested a more basal In this study we also examined the position for this species within the phylogenetic position of two small superficially Lycoperdaceae clade. Lycoperdon pyriforme similar puffballs, Bovista aestivalis (Bonord.) did not cluster with Morganella species in our Demoulin and B. dermoxantha (Vittad.) De analyses, nor did it in the studies of Krüger and Toni. Sequences originating from Arizona Gargas (2008) or Larsson and Jeppson (2008). specimens of Bovista aestivalis, including one We, therefore, do not currently follow Krüger specimen that had been incorrectly identified as and Kreisel (2003) in placing this species in B. dermoxantha and exhibited intermediate Morganella; however, the shared lignicolous type eucapillitial threads upon re-examination habit is a compelling argument supporting their (FJ438477; collection #STB00124; Bates, inclusion in the genus as Morganella pyriformis. 2004), closely matched (99% identity in a Sequence data from a broader sampling of BLASTn search) that of a European specimen Morganella species, more prevalent in tropical of the same species (DQ112620; see Larsson and subtropical parts of the world, may help in and Jeppson, 2008). These species subsequently clarifying this issue. clustered together in the consensus trees (see The taxonomic implications suggested for Fig. 1) along with B. polymorpha (Vittad.) Lycoperdon by our study, and those of others as Kreisel, often cited as a synonym for B. well (Krüger et al., 2001; Krüger and Gargas, aestivalis. 2008), are particularly significant as L. Sequence from an Arizona specimen perlatum, the type species of the genus, appears previously included under B. dermoxantha to fall outside of the core clade of Lycoperdon (FJ438478; collection #STB00624; Bates 2004), 158 Fungal Diversity on the other hand, did not closely match (97% septate. The walls of paracapillitial and identity) its European counterpart (DQ112579; eucapillitial threads are often, but not always, see Larsson and Jeppson, 2008). Sequence from perforated with small- to large-sized pores. The the Arizona material more closely matched two types can be distinguished easily when a (98% identity) several Lycoperdon species, small portion of the gleba is mounted on a slide which were also included in our phylogenetic in Lactophenol Cotton Blue. At first analyses. The consensus trees (see Fig. 1) paracapillitial threads will stain blue (Fig. 2b) suggested the Arizona material is allied closely and pigmented eucapillitial threads will remain with L. niveum (DQ112599, collected from the more-or-less unchanged. However, upon type locality); however, that specimen is boiling the mountant for a few seconds, apparently not closely related to those from paracapillitial threads become clear and Iceland and Norway identified under the eucapillitial threads take on a deep blue current European concept for the species (see (cyanophilous) staining reaction (Kreisel and Larsson and Jeppson, 2008), a result confirmed Dring, 1967). in our analyses. Furthermore, we found the In addition to these generalized European B. dermoxantha (DQ112579, distinctions, several capillitium types have also collected in Sweden) to be closely allied with been established. Kreisel (1967b) recognized Holocotylon brandegeeanum (sharing 97% three types: the ‘Bovista’ type, the identity in a BLASTn search). Considering the ‘Lycoperdon’ type, and the ‘intermediate’ type results of our study and those of Larsson and (used in this study) or ‘transitional’ type (see Jeppson (2008), it is clear, however, that the Krüger et al., 2001). The Bovista type (Fig. 2d) specimens from Sweden and Arizona originally has elastic, aseptate eucapillitial threads that cited as B. dermoxantha are more closely allied lack pores and have centralized main stems with Lycoperdon species than those of Bovista. with numerous ramified branching appendages. Therefore, we retain our material under this This type is found in Bovista as well as species eptithet, as Lycoperdon cf. dermoxanthum, until in Calbovista Morse ex M.T. Seidl and the uncertainty surrounding the above Bovistella Morgan. The Lycoperdon type (Fig. mentioned taxa can be more fully resolved. 2f) is composed of long, elastic eucapillitial threads with occasional dichotomous or Capillitial threads and capillitium types irregular branches. In addtion, the capillitial Most species in the Lycoperdaceae threads in this type normally have pores, but develop specialized hyphae, called capillitial lack centralized main stems. The Lycoperdon threads, internally along with spores as the type capillitium is found in all Lycoperdon puffball matures. En masse, interwoven species as well as some species in Bovista and capillitial threads form a structure called the Vascellum. The intermediate type capillitium capillitium. The interior portion of a puffball (Fig. 2e), as the name implies, is a transitional that includes the capillitium and spores together form between the Bovista and Lycoperdon is called the gleba. types. It contains capillitial threads, often pored, There are two general types of capillitial with several thick main stems that are threads found in the gleba of Lycoperdaceae connected by numerous branches. This type is species. The first type, termed eucapillitial mainly found in Bovista species, B. aestivalis threads (Fig. 2a), consist of sterile, persistent, for example, that do not exhibit the Bovista thick-walled hyphal units, which are normally type capillitium. pigmented and only occasionally septate. The Two additional capillitium types (used second type, paracapillitial threads (Fig. 2b), here) were later established by Krüger and consists of sterile, flaccid, thin-walled hyphal colleagues (2001): the ‘Mycenastrum’ type and units, that are not pigmented and are frequently the ‘Calvatia’ type. The Mycenastrum type 159 Fig. 2. Capillitial threads (hyphal units) and capillitium types: a) thick-walled, pigmented eucapillitial threads; b) thin- walled, hyaline paracapillitial threads, mounted in Lactophenol Cotton Blue (prior to boiling); c) Calvatia type capillitium – fragile eucapillitial threads, disarticulating at the septa; d) Bovista type capillitium – elastic, aseptate eucapillitial threads with thick main stems and numerous branching appendages; e) intermediate type capillitium – elastic to subelastic, aseptate eucapillitial threads with thickened main stems and repeated dichotomous branching; f) Lycoperdon type capillitium – elastic to subelastic, aseptate eucapillitial threads with occasional dichotomous or irregular branching; a–b = 1000×; c, e–f = 400×; d = 100×. 160 Fungal Diversity contains capillitial threads that lack pores and verrucae (see Fig. 16d, L. pyriforme). Spores have thick main stems that are highly ramified with minute, but discernible, ornamentation are with spinose appendages (see Figs 11b, 17f). termed asperulate or asperate (‘faiblement This type is found exclusively in the genus verruqueuses’ of Demoulin). Asperulate spores Mycenastrum. The Calvatia type capillitium appear smooth at 400×; however, minute (Fig. 2c) is similar to the Lycoperdon type in verrucae or echinae are just visible under high that it has capillitial threads that occasionally magnification. Under SEM, small rounded or exhibit dichotomous branching, normally have irregular-shaped verrucae or echinae with pores, and lack centralized main stems. This pointed apices are clearly discernable (see Fig. type, however, is typically septate and often 15e, L. lividum). Asperate spores have verrucae fragile with capillitial threads that disarticulate that are barely visible at 400×, and are more at the septa. apparent at high magnification. Under SEM, It should also be mentioned that Krüger small- to medium-sized rounded or irregular- and colleagues (2001) included an additional shaped verrucae are present (see Fig. 15d, L. cf. type, the ‘Handkea’ type capillitium, described dermoxanthum). as being similar to the Lycoperdon type but Spores with ornamentation of a medium with “long slit-like pits” in the capillitial wall. grade coarseness are described as verruculose This type has mainly been associated with or echinulate (‘verruqueuses’ of Demoulin). Handkea (Kreisel, 1998); however, this genus Here, the verrucae or echinae are somewhat is not accepted here (see comments above) and visible at 400×, but are clearly discernable at its members have not been reported from 1000×. Under SEM, irregular-shaped, broadly Arizona. conical to subcylindrical verrucae or echinae with pointed, rounded, or truncated apices are Spores readily observed (e.g., see Fig. 14e, D. candida; The majority of basidiospores found in Fig. 16b, L. perlatum). Lycoperdaceae are globose to subglobose, and Spores with a more coarse form of the structure of ornamentation covering them is ornamentation are described as being verrucose taxonomically significant. A brief description or echinate (‘fortement verruqueuses’ of of spore ornamentation types, which basically Demoulin). This type of ornamentation is correspond to the four categories of Demoulin clearly visible at 400×, and in very coarsely (1972), follows. It should be noted that the ornamented spores (e.g., see Fig. 16e, L. terminology used here applies only to spores as rimulatum) the verrucae are even discernable at viewed under the light microscope; however, 100×. At higher magnification under SEM, further explanation of each type as they appear further distinctions are readily made in the under SEM is also given to assist in clarifying ornamentation. Subcylindrical to cylindrical these terms. verrucae can be distinguished as well as various Spores lacking evident ornamentation at other characteristics of this type of spore even the highest power of magnification ornamentation. For example, Lycoperdon molle (1000×) under the light microscope are termed (Fig. 16a) has spores with cylindrical verrucae smooth (‘apparemment lisses à ponctuées’ of and truncated apices; L. rimulatum (Fig. 16e) Demoulin). Under SEM, higher magnification has subcylindrical verrucae with rounded apices; (10,000–20,000×) may also reveal an entirely and long slender echinae are visible in the spore smooth surface (see Fig. 14c, C. pachyderma), ornamentation of species like Calvatia aff. one that is slightly roughened (see Fig. 15c, H. rugosa (Fig. 14d). brandegeeanum), or one with very minute 161 Artificial Key to the Arizona Lycoperdaceae if attached then not forming a cup around lower 1/5 to 1/2 of gasterocarp.............................................11 1a. At maturity, gleba comprised of labyrinthiform, 8a. Spores having long pedicels (5.6–18.0 μm in length) sinuous tramal plates; spore mass covering the ..........................................13. Disciseda hyalothrix surface of a true hymenium; spores subglobose to 8b. Spores having short pedicels (less than 2.0 μm in ovoid (4.0–4.8 μm in diam. × 4.0–5.6 μm in length), length), or pedicels lacking.....................................9 smooth to asperulate ................................................ .............................15. Holocotylon brandegeeanum 9a. Spores globose to subglobose (4.0–5.6 μm in diam. 1b. At maturity, gleba not comprised of labyrinthiform, × 4.0–5.6 μm in length), smooth to asperulate......... sinuous tramal plates; spore mass pulverulent and ...............................................12. Disciseda cervina interwoven with paracapillitial and/or eucapillitial 9b. Spores globose, verrucose....................................10 threads; spores variable..........................................2 10a. Endoperidium with reticulate pattern on basal 2a. Paracapillitial threads present; eucapillitial threads portions, formed of felted to flocculent ridges; absent from the gleba.............................................3 spores 4.0–6.4 μm in diameter ................................. 2b. Eucapillitial threads present; paracapillitial threads ..............................................11. Disciseda candida present in the gleba or not......................................4 10b. Endoperidium lacking reticulate pattern; spores 8.0–12.8 μm in diameter....14. Disciseda verrucosa 3a. Exoperidium verrucose; verrucae broad, pyramidal; spores globose to subglobose (3.2–4.8 μm in diam. 11a. Spore mass violaceous..........................................12 × 4.0–5.6 μm in length), asperulate to asperate 11b. Spore mass brownish yellow, olive-yellow, olive- ......................................26.Vascellum intermedium brown to light brown............................................14 3b. Exoperidium rugulose to minutely verruculose, lower portions more noticeably verruculose to 12a. Spores asperulate (4.0–4.8 μm in diam.).................. echinulate; spores globose (4.0–5.6 μm in diam.), ...........................................8. Calvatia cf. leiospora asperulate to aspirate............28. Vascellum texense 12b. Spores verrucose...................................................13 4a. Eucapillitial threads having numerous ramified 13a. Subgleba present and prominent, comprising lower branches.................................................................5 1/3 to 1/2 of gasterocarp, cellular and composed of 4b. Eucapillitial threads lacking ramified branches, medium- to large-sized cells; exoperidium having exhibiting thick main stems connected by branches, some sphaerocysts which contain pigments............. occasional dichotomous branching, or lacking ...........................................6. Calvatia cyathiformis branches altogether.................................................6 13b. Subgleba absent or reduced, if present then composed of compacted or small-sized cells; 5a. Branches short with spinose tips; spores globose exoperidium lacking pigments in sphaerocysts........ (8.8–12.8 μm in diam.), verrucose to reticulate; ....................................................7. Calvatia fragilis pedicel lacking; endoperidium rigid, thick (up to 4 mm).................................25. Mycenastrum corium 14a. Spores globose (5.6–7.2 μm in diam.), strongly 5b. Branches long with attenuate tips; spores ovoid echinate; endoperidium persistent in mature (4.8–5.6 μm in diam. × 5.6–6.4 μm in length), gasterocarps................................3. Calvatia bicolor asperulate to asperate; pedicel long (10.0–16.0 μm 14b. Spores globose to subglobose, smooth to asperate, in length); endoperidium papery to rigid, thin (less or moderately echinate; endoperidium breaking than 2 mm)................................2. Bovista plumbea apart in mature gasterocarps.................................15 6a. Capillitium of the Calvatia type; eucapillitial 15a. Peridia staining yellow when fresh, developing an threads fragile, frequently breaking into shorter orange pigmentation with age; gasterocarp small- to units........................................................................7 medium-sized (20–60 mm in diam.); spores globose 6b. Capillitium of the Lycoperdon or intermediate type; (3.2–4.8 μm in diam.), echinate................................ eucapillitial threads elastic to subelastic, only ...........................................10. Calvatia aff. rugosa occasionally breaking into shorter units...............18 15b. Peridia not staining yellow when fresh, or only weakly so, not developing an orange pigmentation 7a. Exoperidium comprised of a thick mycelial mat, with age; gasterocarp medium- to large-sized (70– heavily encrusted with particles of soil or sand, 600 mm in diam.); spores globose, subglobose to remaining attached and forming a cup around ovoid, smooth to asperulate..................................16 ‘lower’ 1/5 to 1/2 of gasterocarp............................8 7b. Exoperidium not comprised of a thick mycelial mat, 16a. Spores subglobose to ovoid (3.2–4.8 μm in diam. × not heavily encrusted with particles of soil or sand, 4.0–5.6 μm in length), smooth; peridium glabrous, 162