Mycol. Res. 109 (2): 237-245 (February 2005). O The British Mycological Society DOI: 10.1017/S0953756204001789 Printed in the United Kii~gdom.

------Phylogenetic Studies of Tevfezia pfeilii and Choivomyces ecltinulatus () support new genera for southern African

: Kalaharitrcbev and Eremiomyces -

Yael FERDMAN1, Sharon AVIRAM1, Nurit ROTH-BEJERANO1, James M. TRAPPE2 and Varda KAGAN-ZUR3* l Department of Life Sciences, Ben-Gurion University, POB 653, Beer-Sheva, 84105, . Qepartment of Forest Science, Oregon State University, Corvallis, Oregon, USA 97331-5752. Institutesfor Applied Research, Ben-Gurion University, POB 653, Beer-Sheva, 84105, Israel. E-mail : [email protected]

Received 13 April 2004; accepted 24 October 2004.

The ITS region including the 5.8s rRNA as well as the 5' end of the 28s rRNA gene of hypogeous and Tuberacene were studied to clarify the generic placement of two southern African desert truffles, pfeilii and Clzoiromyces echinulatus. The results show that neither species belongs in the to which it has been assigned on the basis of morphological characters. As expected, two Choiromyces spp. grouped close to the representative of the Ttrberaceae ( melanosporum). However, C. echinulatus diverged from the other Choiromyces species and emerged near members of the genus Terfezia, being even closer to that genus than T. pfeilii. Two new genera and new species combinations, gen. nov. with K. pfeilii fsyn. T. pfeilii) comb. nov. and Eremiomyces gen. nov. with E. echinulatus (syn. C. echinulatus) comb. nov. are therefore introduced to accornodate these taxa. Both genera are closely related to Terfeziia, and thus are placed in the Pezizaceae.

INTRODUCTION (2001), and Diez, Manjon & Martin, (2002). Some of these investigations confirm earlier morphological Truffles are the hypogeous ascomata of fungi belonging findings. OYDonnellet al. (1997) and Hansen et a/. to the Pezizales (Trappe 1979, 1990). The assignment of (200 1) demonstrated that certain hypogeous members genera to a particular family and species to a of the Pezizales show greater affinity to epigeous particular genus was traditionally founded on the members than to other hypogeous ones. Similarly, the morphology of ascomata, asci and spores (e.g. Trappe monophyletic origin of some members of Terfezia and 1979). Thus, Terfezia pfeilii was described from Tirmania was confirmed by Diez et al. (2002). However, Damaraland, Namibia, by Hennings (1897) and other results challenge some of the earlier placements, Choiromyces echinulatus from Cape Province, South for example along with Tirmania, Terfezia belongs to Africa, by Marasas & Trappe (1973). the Pezizaceae rather than the (Norrnan In recent years, however, molecular phylogenetic & Egger 1999; Percudani et al. 1999). Further, research on sequestrate fungi has repeatedly demon- Choiromyces was moved from the Terfeziaceae to the strated that morphology of hypogeous fungi can be (O'Donnell et al. 1997, Percudani et al. misleading. Evolution of hypogeous species typically 1999), and Terfezia terfezioides was removed from involves a convergent reduction in macromorpho- Terfezia and reinstated as Mattirolomyces terfezioides logical characters, often entailing loss of features (Percudani et al. 1999, Diez et al. 2002). otherwise useful to distinguish related epigeous taxa. A preliminary molecular study of Terfezia pfeilii More specifically, molecular analyses of the Pezizales and the phylogenetic relations among epi- and hypo- and Choiromyces echinulatus suggested that these two truffles are molecularly close, though geous species have been conducted by O'Donnell et al. currently placed in different families. This apparent (1997), Norman & Egger (1999), Percudani et al. affinity inspired us to study the phylogenetic relation- (1999), Roux et al. (l999), Hansen, Lzssare & Pfister ships between these two species in relation to other Terfezia and Choiromyces species. To this end, we * Corresponding author. analyzed DNA sequences from the internal transcribed Phylogeny of southern Africa truffles

Table 1. Species analyzed for this work.

GenBank Truffle Collection site Date Provided bv DNA frament accession no.

Clzoiron~ycesalveolatu.? Oregon, USA f. Trappe ITS (Trappe 22830) 28s 5' region C. echuzulatw (voucher no. 12) Ghanzi, v. Kugarz-Zur ITS 28s 5' region South West Africa f. Trappe ITS (Namibia) 28s 5' region C. venosus (536) Germany f. Trappe 28s 5' region Mattirolomyces terfezioides Z. Bratek 28s 5' region Terfezia boudieri form 1 Zeelim, Israel ITS 28s 5' region T. boudieri form 2 Zeelim, Israel Bedouins ITS 28s 5' region T. boudieri form 3 ZeeIirn, Israel Bedouins ITS 28s 5' region T. boudieri 42a Zeelim, Israel Bedouins ITS T. claveryi Jarada, Morocco M. Actzouri ITS 28s 5' region T. pfeilii a Kalahari, Botswana ITS T.pfeilii b Kalahari, Botswana ITS 28s 5' region Tuber melanosporum G. Chevalier ITS Mycelium, isolate 'asco ' 28s 5' region

spacer (ITS) of the rRNA gene cluster, that includes In this study we present molecular data that supports ITS1, 2 and the 5.8s gene, as well as the variable Dl removal of Terfezia pfeilii from Terfezia and and D2 5' regions (van Tuinen, Zhao & Gianinazzi- Choiromyces echinulatus from Choiromyces. We Pearson 1998b) of the large subunit (LSU) of the describe two new genera to accommodate these truffles, rRNA gene cluster. still within the Pezizaceae. Much of the molecular research aimed at identifying fungi in general (e.g. Gardes et al. 1991, Henrion, Le Tacon & Martin 1992, Buscot et al. 1996, Mello et al. MATERIALS AND METHODS 1996, Kirtn et al. 1997, Chillali et al. 1998) and truffles in particular (e.g. Henrion, Chevalier & Martin 1994, The fungi Paolocci et al. 1995, 1999) relies on comparison of se- The origin of fruit bodies obtained for the analysis quences and RFLP patterns of the ITS region. Several presented in this work is summarized in Table I. phylogenetic studies of truffles were also based on this Voucher collections of the specimens studied are DNA region (Roux et al. 1999, Diez et al. 2002). Other deposited in the Department of Botany Mycological molecular phylogenetic studies have been based on se- Herbarium, Oregon State University (OSC). Previously quences from the 18s and the 28s (LSU) rRNA published sequences used in this work are presented in with or without the ITS region (e.g. 07Donnell et al. Table 2. 1997, van Tuinen et al. 1998a, Norman & Egger 1999, Percudani et al. 1999, Hansen et al. 2001). The rRNA genes present different levels of conservation whlch can DNA extpaction and ampl$cation be exploited to analyze any desired phylogenetic level The method described by Grube et al. (1995) was (Hillis & Dixon 199 1). employed to obtain small amounts of amplifiable Molecular data now available for the genera Terfezia DNA. DNA was amplified with primers ITS 1 and ITS and Tirmania place them, as mentioned above, in the 4 (White et al. 1990) for ITS or LR 1 and U2 (van Pezizaceae (Norman & Egger 1999, Percudani et al. Tuinen et al. 1998b and Sandhu et al. 1995, respect- 1999). Such molecular data are not yet available for ively) for the large subunit (LSU). Primers were syn- Mycoclelandia (an Australian desert genus with two thesized by Rhenium, Jerusalem. Both ITS and LSU species), but it was assigned to the Pezizaceae because amplifications were performed as described for ITS by of its strongly amyloid asci (Trappe 1979, Trappe & Kagan-Zur et al. (1999). Beaton 1984). Morphology alone has been the basis for putting two other desert species, Choiromyces echinu- Cloning latus and C. aboriginum, in the Tuberaceae (Marasas & Trappe 1973, Trappe 1979) and a third, Phaeangium PCR amplified ITS or the 5' end of the LSU frag- lefebvrei, in the Pyronemataceae (Alsheikh & Trappe ments were cloned into lnsT/AcloneTMPCR Product 1983). Cloning kit No. K1214 (MBI Fermentas, Lithuania). Y. Ferdman and others

Table 2. Sequences obtained from the National Center for Phylogenetic evolutionary trees based on these data Biotechnology Information (NCBI). were constructed along with 1000 bootstrapping Accession no. Species Origin repeats by MEGA version 2.1 (Kumar et al. 2001). Several reshuffled lists of sequences were presented to Clzoiromyces venosus, LSU partial Oregon USA both the ClustalX and Mega programs. The Mega sequence' C. venosus (as meundriforn~is),ITSb Italy program's 'Subtree Swapping' option was used to Mattirolomyces terfezioides 1, ITSC Hungary adjust outputs. Trees were built with and without M. terfezioides 2, ITSC Hungary deleting gaps. Trees presented here were constructed Terfezia leptoderina 1. ITSC using the gap delete options. Phylogenetic analysis of T. leptoderma 2, ITSC Spain the ITS and the large subunit (LSU) sequences were T. arenariu 1, ITSC Spain T. arenaria 2, ITSC Spain carried out by the neighbour-joining (NJ) and maximal T. olbiensis I, ITSd Not specified parsimony (MP) methods. T. olbierzsisI 2, ITSd Not specified T. claveryi 3, ITSd Not specified Tirmania pinoyi, ITS" Morphological analyses T. iiivea, ITSa T. pmoyi, LSU partial sequencee Algeria Dried specimens were hand-sectioned with a razor T. nivea, LSU partial sequenceC Kuwait blade and mounted in 5 % KOH, ~'elzer'sreagent and cotton blue in lactic acid, respectively. Spore dimen- a O'Donnell et al. (1997). Amicucci, Stocchi & Martin (Direct submission). sions are based on at least 30 randomly selected spores Diez, Manjon & Martin (2002). plus the smallest and largest seen, excluding the * Gutierrez, Honrubia & Morte (Direct Submission). ornamentation. Spores were measured in cotton blue " Hansen et al. (2001). mounts, because the walls or ornamentation of many ascomycetes tend to swell in KOH. Other tissues were measured in KOH. The plasmid mixture was transformed into Escherichia coli XL 1-blue in accordance with the manufacturer's recommendations. Transformants were plated on agar solidified Luria Broth (LB) plates (Miller 1972) con- RESULTS AND DISCUSSION taining 50 pg ml-* ampicillin. Phylogenetic analyses ITS analyses Inserts sequencing The ITS phylogenetic tree presented as Fig. 1 was the Single colonies were picked up and grown in LB over- output result of both NJ and MP methods of analysis, night. Plasmids were extracted from these cultures by although there are differences in percent bootstrapping. the alkaline method (Birnboim & Doly 1979). Plasmids In this tree some of the Terfezia group nodes, notably containing inserts were sent to the sequencing unit of separation of T. claveryi and T. arenaria (40130) as a the Hebrew University of Jerusalem, where the Big Dye distinct clade and grouping of the latter with the Terminator kit (Perkin-Elmer, Branchburg, NJ) was T. boudieri clade (47/36), were only weakly supported used in conjunction with the primer pair MI3 by either method. Including sequences leading to (Forward) and pUC (Reverse) (Sambrook & Russell extensive gap formation resulted in an NJ tree very 2001) for bi-directionally sequencing of inserted frag- similar to Fig. 1. The only placing changes observed ments directly from each plasmid. Primers were syn- involved the weakly supported branches (data not thesized at the Molecular Biology Laboratory, shown) within the Terfezia clade. Reshuffling the input Hadassa Hospital, Jerusalem. The sequence was deter- data did not change the grouping presented in Fig. 1 mined with an ABI prism 377 DNA sequencer (Applied but in one or two cases changed the placing order (e.g. Biosystems, Foster City, CA). Sequences were de- putting T. boudieri first, or separating T. claveryi from posited in the National Center for Biotechnology T. arenaria). These are easily achieved or corrected by Information (NCBI) databases (Table I). 'Subtree Swapping' at will. However, examining the best bootstrapping tree (data not shown) revealed Alignments and phylogenetic analysis several other differences : Tuber melanosporum, chosen as an out group, which forms a Tuber-Choiromyces Several sequences were obtained for phylogenetic clade together with C. venosus (submitted to the analysis from the National Center for Biotechnology GenBank as C. meandriformis) and C. alveolatus in Information (Table 2). Sequence alignments were sub- Fig. 1, emerges as a true out-group, though close to the mitted to TreeBASE PIN no. 12889. sequences were above mentioned Choironzyces spp. aligned with ClustalX version 1.81, downloaded for The separation of T. boudieri 2 (and 42a2) from the Macintosh from http://www.embl.de/-chenna/ T. boudieri I and 3 is well supported in Fig. 1, but all 3 clustal/darwin/. The results were manually checked and T. boudieri ITS forms (Holdengraeber et al. 2001) adjusted. stem from the same root in the bootstrapping tree Phylogeny of southern Africa truffles

40130 Terfezia claveryi-3 Terfezia arenaria-1

64/46 r Terfezia boudieri-1 Terfezia boudieri-3

Terfezia boudieri-42a2

Terfezia leptoderma-1 Terfezia olbiensis-2

Tirmania pinoyi Tirmania nivea Choiromyces echinulatus-12 Choiromyces echinulatus-44894 1' Mattirolomyces terfezioidesl I r Terfezia pfeilii-a loollo~ ~erfeziapfeilii-b I I Tuber melanosporum 1- 1- Choirornyces venosus 1001100 Choimmyces alveolatus

Fig. 1. Neighbour-joining consensus phylogenetic tree of ITS sequences. Bold numbers are % of 1000 bootstrapping replicates supporting the NJ tree presented here. Non-bold numbers are % of 1000 bootstrapping replicates supporting the same node by the MP method.

(no hierarchy). Some of the less supported nodes were between the Tirmania and Tuber-Choiromyces groups. also eliminated from the latter, resulting in the loss of The data support reinstatement of Mattirolomy ces the less supported information. Both, however, essen- terjezioides (Fischer 1938) as a replacement for Terfezia tially support removal of C. echintrlatus and Terfezia terjezioides and representative of a separate genus pfeilii from the genera to which they were assigned distinct from Terfezia, as suggested by Percudani et al. - solely on morphological grounds. (1999) and supported by Diez et ai. (2002). C. echinulatus, emerges closer to the Tirmania clade We also performed multiple alignment of the 5.8s. than to the other Choiromyces species, being placed Table 3 shows that out of the 155 bp analyzed for the within the in-group. Most of the Terfezia spp. group 5.8s gene, 23 sites exhibited variability, representing 22 together to constitute a monophyietic Terjezia clade, independent events. Of these, ten were transitivns (5 the closest neighbour of which is the Tirmania lineage. C-T), eight were transversions, two were inversions of T. leptoderma and T. olbiensis form a distinct group as dinucleotides, one was an insertion and one deletion of sister to T. botrdieri, T. areizaria and T. ciaveryi. two nucleotides. However, T. pfeilii, the Kalahari Terfezia, is distant The above sequences fall into two main groups, one from other TePJizia species and forms a separate including T. melanosporum, C. alveolatus, C. venosus lineage. Mattirolomyces terfezioides, T. pjeilii and C. (as meandriformis) and C. venosus, and the other echinulatus, belonging to separate clades, are placed Te$ezia spp., C. echinulatus, Tirmania spp., and Y. Ferdman and others

-...Y M. terfezioides. Both show relatively small differences 3- within and relatively large differences between the -r.= El groups. Detailed analysis of the 5.8s gene revealed 2 three variable sites within the Tuber-Choiromyces R group (three sequences), and eight sites variable solely 2 3 TI-7 within the Terfezia-Tirmnia group (ten sequences). An ?E'" 2 additional seven. sites were identical within each group 4 .=C .z= but varied between them. T. pfeilii and C. echinulatus 5 displayed greater conformity to the 5.8s gene con- s- sensus sequence (emerging from this study) than $$ T. arenaria and T. claveryi. F; - Analysis of the 5' end of the LSU $- 2-.a To further test these results, the divergent domains Dl F-; and D2 of the LSU rRNA gene were also analyzed. This DNA region is better conserved than the ITS but .s .r 52ta =a still variable enough to show differences between h -s species (Hillis & Dixon 199 1, van Tuinen et al. 1998b). The LSU NJ tree (Fig. 2), like the ITS tree, shows C. '?h' .'5 alveolatus and C. venosus branching together with T. F-; melanosporum to forrn a Tuber-Choiromyces clade, an outgroup; this tree is supported by the bootstrapping. .-Q - As in the ITS tree, the genus Tirmania is placed closer :Q 5 .g to the Terfezia clade, while C. echinuiatus, M. terfe- - zioides, and T. pfeilii, form separate clades where C. 2 echinulatus is closer to Terfezia spp. than to other .O Choiromyces spp. However, the MP trees were some- 8- what different. One of the better supported trees (Fig. 3) % shows a non-hierarchical relationship of M. terfe- zioides, C. echinulatus, Tirmania spp. and the Terfezia Y 2"2 clade, while T. pfeilii emerges in a hierarchically once 3 2 removed clade. The non-hierarchical branching is .g .o -.a pa rather poorly supported (59 % bootstrap value), but -G % $, the separation of T. pfeilii is well supported. All MP trees, however, placed T. melanosporum on its own, as -1- $- an outgroup, while C. alveolatus and C. venosus formed 90 a separate clade. Essentially, the two rRNA gene clus- G ter regions examined in our phylogenetic analysis of

'? Terfezia and C?zoiromycesspecies, namely the ITS region 2 including ITS1, 2 and the 5.8s rRNA gene, as well as L? the 5' end of the LSU gene, generated similar results. F; However, as GenBank contains rather few 28s .srr 'Z= sequences, the range of the 28s (LSU) gene tree was 2' 2 2 necessarily limited as compared with the ITS tree. The LSU result concerning Terfezia boudieri was somewhat 4 % unexpected, as this region is rather better conserved than -g the ITS, we had expected a resolution along species 2 2 lines rather than within the species. While T. pfeilii 3 9 a and b had identical 28s sequences and T. boudieri ITS .-$5 0 -3 -c forms 1 and 3 were also almost identical to each other Cl Li- in their 28s sequences, T. boudieri ITS 2 deviated from - the latter. The hypothesis that these fo~s,though 0 3 they may seem morphologically indistinguishable, 19 4 3 actually represent two different species, deserves fur- g ther investigation. All trees presented here provide strong support for a 0 2 close relationship between T. melanosporum and the iij Choiromyces clade, while equally strongly supported is Phylogeny of southern Africa truffles

Terfezia boudieri 2 Terfkzia claveryi Tirmania pinoyi A- Tilmania nivea x Choiromyces echlnulatus-I2 Choiromyces echinulatus 100 44894 I MattiroIomyces terfezioides I ITerfezia pfeilii I I I-Tuber melanosponrm IChoiromyces alveolatus -Choiromyces venosus U42688 Choiromyces venosus 536

Fig. 2. Neighbour-joining consensus phylogenetic tree of the Dl-D2 5' end of the 28s (LSU) rRNA gene sequences. Numbers are % of 1000 bootstrapping replicates supporting the tree presented here.

Terfezia boudieri 1 Terfezia boudieri 3 Terfezia boudieri 2

Tirmania pinoyi 59 I Tirmania nivea Choiromyces echinulatus 12 , ,001 Choiromyces echinulatus 44894 I I I Mattirolomyces terfezioides rl Terfezia pfeili Choiromyces alveolatus Choiromyces venosus U42688 100 Choiromyces venosus 536 I Tuber mclmosponun

Fig. 3. A representative Maximal Parsimony (MP) phylogenetic tree of the Dl-D2 5' end of the 28s (LSU) rRNA gene sequences. Numbers are % of 1000 bootstrapping replicates supporting the tree presented here. the placing of C. echinuiatus outside this clade within the ingroup. T. pfeilii, although located within the ingroup is phylogenetically farthest from the Terfezia Kalaharituber Trappe & Kagan-Zur, gen. nov. clade. To conclude, our findings evidence that the assign- Etym. : Kalahari- + -tuber (Latin, 'truffle'), 'truffle of ment of C. echinuiatus to the genus Choiromyces and the Kalahari. ' therefore to the Tuberaceue is incorrect. It is also clear that T. pfeilii does not belong to the Terjezia clade and A Tet;fiziu sporis spinis minutissimis dense congestis, ordiniibus DNA abhorrentibus, distributione in Hemisphaerio Meridionali its classification needs to be revised. In spite of the differt. molecular resemblance between T. pfeilii and C. echinu- Typus: Terfezia pfeilii Henn. iatus, which initiated this study, the finding that the two species occupy distinct clades in all types of analyses Kaliharituber pfeilii (Henn.) Trappe & Pagan-Zur, comb. nov. (Fig. 4) precludes their placement in a single genus. We there- fore propose two new genera, described below, to ac- Basionym: Tuber pfeilii Henn., Bot. Jahrb. 22: 75 cornmodate the above findings. (1 897). Y. Ferdman and others 243

Figs 4-5. Ascospores. Fig. 4. Kalaharituber pfeilii in KOH. Fig. 5. Eremiopnyces echinulatus in Cotton blue.

Etym. : in honor of Count Pfeil, the original collector claveryi, Storey 1958, Taylor et al. 1995) and various of the type species. other plants, including Citrullwr; vulgaris (Kagan-Zur Ascomata hypogeous, smooth, unpolished, minutely et al. 1999), April through June. pubescent, turbinate to obpyriform or subglobose, to Illustrations (as Terfezia claveryi or T. pfeilii) : Pole- 6.5 x 6 cm. Peridium ca 1 mm thick, wrinkled Evans (1918: pl. 7), Mattirolo (1922: fig. 28), Leistner (especially on the upper ascoma surface), dark brown, (1967: pl. 50), Marasas & Trappe (1973: fig. 2), the wrinkles yellowish. Gleba solid, fleshy, marbled Alsheikh (1 994 : pl. 2-1 8). with sterile, white veins and yellowish white to brown, Comments: The spore ornamentation of minute, fertile pockets, Odour strongly fungoid. Spores globose, densely crowded spines of K. pfeilii was recognized as 16-22 (-26) pm diam, at first hyaline, in age light different from that of any other Terfezia spp. by brown, the walls + 1.5 pm thick, 2-layered, light yellow Mattirolo (1922), Marasas & Trappe (1973), and to light orange in Melzer's reagent; ornamentation Alsheikh (1994). Nonetheless, it has been variously appearing in KOH as a minutely papillose or granu- misidentified as Tefezia boudieri or T. claveryi (Marloth lose, mucilaginous epispore but in cotton blue clearly 1913, Pole-Evans 1918, Ceruti 1960). Marasas & seen as densely crowded, acute spines 0.5-1.5 Trappe (1973) ascribe the early confusion of K. pfeilii (-2) x 60.5 pm. Asci nonamyloid, with (4) 5-8, with other Terfezia spp. to the paucity of good loosely arranged spores, globose to ellipsoid or specimens for study and the nature of its spore orna- obovoid, 70-100 x 50-80 pm, sessile to substipitate, the mentation : spines so minute and crowded that they can walls + 1 pm thick at maturity, randomly arranged in be discerned only when stained, even with the oil im- fertile pockets. Ectal excipulum 180-220 pm thick, of mersion objective. Use of K. pfeilii for food by natives brown hyphae 4-12 pm broad at septa but the cells of the Kalahari was reviewed by Trappe (1990). inflated to 15-30 ym. Ental excipulum of generally Specimens exainined: Botswana: Ghanzi, G. Scholtz (PRE circumferentially aligned, hyaline to pale yellow 41 869). Namibia : Damaraland, Graf Pfeil (S - holotype of hyphae 4-10 ym broad at the septa, the cells often Terfezia pfeilii); Gibeon, Burger (PRE 42082, 42203); slightly inflated. Glebal hyphae of sterile veins and fertile Gobabis, Verbuch'chlcin (PRE 4390); Klein Karas, 20 Apr. pockets undifferentiated, hyaline, thin-walled, 5-12 pm 1921, A. E. Hill (OSC, TO); loc. cit., 8 June 1923 (BPI 38083, PRE 17799, OSC, TO); Windhoek, Gies (PRE 42076). South broad at septa but the cells generally inflated to 20 pm. Africa: North Cape: Kakamas, June 1931, M. J. Oosthuizen Distribution, and Season : Kalahari Desert of (PRE 26335, OSC); Kalahari Gemsbok National Park, May southern Africa: Namibia, Botswana, and adjacent 1956, R. Story 5616 (PRE 41602); Postmasburg, May 191 1, areas of , In compact to fairly compact, F. L. Hunter (PRE 11293, OSC); Prieska, MacCleod (PRE calcareous, white and pink sands (Leistner 1967), 11619); Upington, 0.A. Leistner 2610 (PRE 42201); associated with spp. (Marloth 1913, as Terjezia Vryburg, Stephens 527 (PRE 36103); unspecified locality, Phylogeny of southern Africa truffles 244

Le Rich (PRE41870) - KaIahari Desert: unspecified locality, sequences place E. echinulatus in the Perizaceae, Bottomley (PRE 443 10);herb. G. Bresadola (BPI, as Terfezia whereas Choiro~nycessensu stricto is in the Tuberaceae boudieri); 1913, R. Marloth (K, as Terfezicr claveryi);Nmh (OYDonnellet 1997, this study). (PRE 44254); Weintraub (PRE 32394). Specimens examined: South Africa: North Cape: Gordonia near Upington, 1 June 1961, 0. A. Leistrrer 2612 (PRE 42202 - holotype of Eremiomyces Trappe & Kagan-Zur, gen. nov. Cltoiromyces echinulatus). Elynz. : Eremi- (Greek, ' desert ') + -myces (Greek, ' '), ' desert fungus. ' A Choiromyces sporis virgis conicis rectis obtusis, cellulis magnopere ACKNOWLEDGEMENTS inflatis, ordiniibus DNA abhorrentibus (ad Pezizaceas, non Participation in this study by J.M.T. was supported in part by the Tuberaceas, affinis) et distributione in desertis Hernisphaerio Meridionali differt. U.S. Forest Service, Pacific Northwest Research Station, Forestry Typus: Choiromyces echinulatus Trappe & Marasas. Sciences Laboratory, Corvallis, Oregon.

Eremiomyces echinulatus (Trappe & Marasas) Trappe & Kagan-Zur, comb. nov. (Fig. 5) REFERENCES Basiosym: Choirmyces echinulatus Trappe & Marasas, Alsheikh, A. M. (1994) Taxonomy and Mycorrhizal Ecology of the Bothalia 11: 139 (1973). Desert TntfRes in the Genus Terfezia. PhD thesis, Oregon State Etynz: 'echinulatus' (Latin, 'having small spines'). University, Corvallis. Alsheikh, A. M. & Trappe, J. M. (1983) Desert truffles: the genus Asconzata pale cream colored and subglobose when Tirmania. Transactions of the Britiih Mycological Society 81: fresh, as dried with a black peridium and solid gleba 83-90. Birnboim, H. C. & Doly, J. (1979) A rapid alkaline extraction pro- marbled by dark brown veins embedding hymenial cedure for screening recombinant plasmid DNA. Nucleic Acids palisades ; opposing palisades deformed from pressing Researclz 7: 1513-1 523. against each other. Buscot, F., Wipf, D., Di Battista, C., Munch, J. C., Botton, B. & Spores globose, 10-14 ym diam, light yellow to light Martin, F. (1996) DNA polymorphism in morels: PCR/RFLP brownish yellow, the walls 1 pm thick, light blue in analysis of the ribosomal DNA spacers and microsatellite-primed PCR. Mycological Reseurclz 100 : 63-7 1. cotton blue; ornamentation of straight, obtuse rods Cemti, A. (1960) Icotzogruphiue Mycologicae. ab. J. Bresadolae. and cones 1-2 x 0.5-1 (-1.5) ym, pale yellow in KOH Supplemento 2. Trento. and light blue in cotton blue, ca 20-25 around the spore Chillali, M., Wipf, D., Guillaumin, J.-J., Mohammed, C. & Botton, circumference, unconnected except for some spores on B. (1998) Delineation of the European Armillaria species based on which barely perceptible lines on the spore surface join the sequences of the internal transcribed spacer (ITS) of ribosomal DNA. New Phytologist138: 553-561. occasional ornaments. Asci hyaline, thin-walled, mostly Diez, J., Manjon, J. L. & Martin, F. (2002) Molecular phylogeny of 8-spored, in youth cylindric to ellipsoid or saccate, by the mycorrhizal desert truffles (Trrjezia and Tirmaniu), host speci- maturity mostly cylindric, 140 x 12-17 pm, astipitate or ficity and edaphic tolerance. Mycologia 94: 247-259. with a short basal projection, the spore arrange- Fischer, E. (1938) Tuberineue. Naturlicherz Pffaizzrnjkilien 8 : 1-42. ment occasionally uniseriate but mostly incompletely Gardes, M.. White, T. J., Fortin, J. A., Bruns, T. D. &Taylor, J. W. (1991) Identification of indigenous and introduced symbiotic fungi biseriate to irregular. Paraphyses thin-walled, hyaline, in ectomycorrhizae by amplification of nuclear and mitochondria1 +4 pm broad. Ectal excipulum of light yellow, thin- ribosomal DNA. Canadian Journal of Botatry 69: 180-190. walled hyphae 4-8 pm broad at the septa; ental Gmbe, M., Depriest, P. T., Gargas, A. & Hafellner, I. (1995) DNA excipulunz similar but the hyphal cells often inflated to isolation from lichen ascomata. Mycological Research 99 : 15-50 Fm broad. Glebal veins of subparallel-interwoven 1321-1324. Hansen, K., Lzssne, T. & Pfister, D. H. (2001) Phylogenetics of the hyphae 3-6 pm broad at the septa, the cells mostly Pezizuceae, with emphasis on Peziza. Mycologiu 93: 958-990. inflated. Hennings, P. (1897) Fungi camerunenses I. Botmischer JaArhlTclzer Distribution, Habitat and Season : North Cape, South fur Systemutik, PJlanzengeschichte utid Pfinzertgeograplzie 22: Africa, and Botswana. June, in sand dunes. 72-1 1 1. Illustrations (as Choironzyces echinulatus): Marasas Henrion, B., Chevalier, G. & Martin, F. (1994) Typing truffle species by PCR amplification of the ribosomal DNA spacers. Mycofogicul & Trappe (1 973 : fig. l), Trappe (1979 : fig. 24a). Research 98 : 37-43. Comments: This rarely collected fungus mimics Henrion, B., Le Tacon, F. & Martin, F. (1992) Rapid identification of Choiromyces of the Tuberaceae in its macro- genetic variation of ectomycorrhizal fungi by ainplification of morphology. However, its spore ornamentation of ribosomal RNA genes. New P/zytologist 122: 289-298. straight, obtuse, rods and cones differs from the Hillis, D. M. & Dtxon, M. T. (1991) Ribosomal DNA: molecular evolution and phylogenetic inference. The Quurterly Rrvieirr of sinuous rods with apical depressions of Choiromyces Biology 66 : 4 1147. venosus, the type species of that genus. The other Holdengraeber, S., Martin, F., Roth-Bejerano, N. & Kagan-Zur, V. Choiromyces spp. have pitted spore surfaces (Trappe (2001) Some lobed fruit bodies of Terfeziu boudieri (a desert truffle) 1979). Eremiomyces echinuiatus is a desert dweller, and combine two ITS-RFLP patterns, suggesting independent origins. like most other desert truffles has tissues with greatly In Scierrcr et Culture a'e iu Tqe: Actes rfrc P' Congr6s Iitter~lutional:78-81. Imprimerie IPSO, Marseille. inflated cells, a character not found in C/zoiromyces Kagan-Zur, V., Kuang, J., Tabak, S., Taylor, F. W. & Roth- species of temperate forests. Of course, the DNA Bejerano, N. (1999) Potential verification of a host plant for the Y. Ferdman and others

desert truffle Teflezia pfeiiii by molecular methods. M~~cological Roux, C., Sejalon-Delrnas, N., Martins, M., Parguey-Leduc, A., Researcfz 103: 1270-1274. Dargent, R. & Becard, G. (1999) Phylogenetic relationships KArCn, O., Hogberg, N., Dahlberg, A., Jonsson, L. & Nylund, J.-E. between European and Chinese traes based on parsimony and (1997) Inter and Intraspecific variation in the ITS region of rDNA distance analysis of ITS sequences. FEMS Microbiology Letters of ectomycorrhizal fungi in Fennoscandia as detected by endo- 180: 147-155. nuclease analysis. New Plzytologist 136: 313-325. Sambrook, J. & Russell, D. W. (2001) Molecular Cloning. (3rd edn.) Kumar, S., Tamura, K., Jakobsen, I. B. & Nei, M. (2001) MEGA2: Vol. 2. Cold Spring Harbor Laboratory Press, Cold Spring lnolecular evolutionary genetics analysis software. Bioi~tJormatics Harbor, NY. P8.115. 17: 1244-1245. Sandhu, G. S., mine, B. C., Stockman, L. & Roberts, G. D. (1995) Leistner, 0. A. (1967) The 'Plant Ecology oj' the Southern Kululzuri. Molecular probes for diagnosis of fungal infection. Journal of [Botanical Survey Memoir No. 38.1, Republic of South Africa Cli~zicalMicrobiology 33 : 29 13-29 19. Botanical Research Institute, Pretoria. Storey, R. (1958) Some Plants Used by the Bushmen in Ob~aiiringFood Marasas, W. F. 0. & Trappe, J. M. (1973) Notes on southern African and Water. [Botanical Survey Memoir No. 30.1 Union of South Tuberales. Botlzalia 11: 139-1 4 1. Africa Department of Agriculture, Pretoria. Marloth, R. (1913) The Flora of South Africa. Vol. 1. Darter Taylor, F. W., Tharnage. D. M.! Baker, N., Roth-Bejerano, N. & Brothers, Cape Town. Kagan-Zur, V. (1995) Notes on the Kalahari desert truffle Terfezia Mattirolo, 0. (1922) Osservazioni sopra due ipogei dell Cirenaica e pfeilii. Mycological Researclz 99: 874-878. considerazioni intorno ai generi Tirmania e Terfezia. Reale Thomas, M. T. (1974) The Harmless People. Alfred A. Knopf, Accarlemiu Nuzionule dei Lincei 5: 545-569. New York. Mello, A., Nosenzo, C., Meotto, F. & Bonfante, P. (1996) Rapid Trappe, J. M. (1979) The orders, families, and genera of hypogeous typing of truffle mycorrhizal roots by PCR amplification of the Ascomycotina (truffles and their relatives). Mycotaxon 9: 297-340. ribosomal DNA spacers. 6 : 417421. Trappe, J. M. (1990) Use of truffles and false-truffles around the Miller, J. H. (1972) Experiments in Moleculur Ge~retics.Cold Spring world. In Atti del Seco~rdoCongresso Internzrionale sul Tarrufo, Harbor Laboratory Press, Cold Spring Harbor, NY. Spoleto. Iraly (M. Bencivenga & B. Granetti, eds): 19-30. Norman, J. E. & Egger, K. N. (1999) Molecular phylogenetic analysis Comunita Montana dei Monti Martani e del Serrano, Spoleto, of Peziza and related genera. Mycologia 91: 820-829. Italy. O'Donnell, K., Cigelnic, E., Weber, N. S. & Trappe, J. M. (1997) Trappe, J. M. & Beaton, G. W. (1984) Mycoclelwzdia nom. nov. Phylogenetic relationships among ascomycetous truffles and the (hypogeous Ascomycotina), a replacement for the pre-empted true and false morels inferred from 18s and 28s ribosomal DNA generic name Clelandiu. Transuctions of the British Mycological sequence analysis. Mycologia 89: 18-65. Society 83: 535-536. Paolocci, F., Cristofari, E., Angelini, P., Granetti, B. & Arcioni, S. van Tuinen, D.. Jacquot, E., Zhao, B., Gollotte, A. & Gianinazzi- (1995) The polymorphism of the rDNA region in typing Pearson, V. (1998a) Characterization of root colonization profiles and ectomycorrhizae of truffle species. In Biotechnology of by a microcosm community of arbuscular mycorrhizal fungi using Ectom~~corrhizae(V. Stocchi, P. Bonfante & M. Nuti, eds): 28s [DNA-targeted nested PCR. Molecular Ecology 7: 879-887. 171-184. Plenum Press, New York. van Tuinen, D., Zhao, B. & Gianinazzi-Pearson, V. (1998b) PCR Paolocci, F., Rubini, A., Angelini, P., Granetti, B. & Arcioni, S. studies of AM fungi, from studies to application. In Mycorrhizu (1999) Rapid molecular approach for a reliable identification of Manual (A.Vama, ed.): 387-400. Springer-Verlag, Berlin. Tuber spp ectomycorrhizae. FEMS Microbiology and Ecology 28: White, T. J., Bruns, T., Lee, S. & Taylor, J. (1990) Amplification and 23-30. direct sequencing of fungal ribosomal RNA genes for phylo- Percudani, R., Trevisi, A., Zambonelli, A. & Ottonello, S. (1999) genetics. In PCRP Protocols. A Guide to Methods and Amplifi- Molecular phylogeny of truffles (Pezizales: Terfeziaceae, cations. (M. A. Innis, D. H. Gelfand, J. J. Sninsky & T. J. White, Tuberuceae) derived from nuclear rDNA sequence analysis. eds) : 3 15-322. Academic Press, San Diego. Moleculur Plrylogenetics ar7d Evolution 13 : 1 69-1 80. Pole-Evans, I. B. (19 18) Note on the genus Terfezia, a truffle from the Kalahari. Tra~tsactionsof the Royal Sociery of South Africa 7: 117-1 18. Corresponding Editor: H. T. Lunzbsclz