A brief chronicle of the genus cordyceps fr., the oldest valid genus in cordycipitaceae (hypocreales, ascomycota).

The earliest pre-Linnaean fungal genera are briefly discussed here with special emphasis on the nomenclatural connection with the genus Cordyceps Fr. Since its valid publication under the basidiomycetous genus Clavaria Vaill. ex L. (Clavaria militaris L. Sp. Pl. 2:1182, 1753), the genus Cordyceps has undergone nomenclatural changes in the post-Linnaean era, but has stood firmly for approximately 200 years. Synonyms of Cordyceps were collected from different literature sources and analyzed based on the species they represent. True synonyms of Cordyceps Fr. were defined as genera that represented species of Cordyceps Fr. emend. G. H. Sung, J. M. Sung, Hywel-Jones & Spatafora. The most common synonyms of Cordyceps observed were Clavaria and Sphaeria Hall, reported in the 18th and in the first half of the 19th century, respectively. Cordyceps, the oldest genus in the Cordyceps s. s. clade of Cordycipitaceae, is the most preferred name under the "One Fungus = One Name" principle on priority bases.

In the 16th century, Fungus Tourn. ex Adans. and Tuber P. Micheli ex F. H. Wigg. were the only two generic fungal names (Table 1) [1]. At that time, fungi were considered as nothing more than the superfluous humidity of soil, trees, rotten wood, and other decaying substances [2]; in fact, potatoes and legume root nodules were considered to be relatives of truffles [3]. At the end of the 17th century and the beginning of the 18th century, Tournefort (1656~1708), considered the father of the modern generic concept, added five new fungal genera, including Agaricus L. of Bauhin (Table 1) [2, 4, 5], which were studied by successive authors of the early 18th century (Table 1) [6, 7, 8]. Subsequently, approximately 30 new fungal genera were proposed by Micheli [9], the father of mycology, which were studied by Von Haller [10, 11] who added the new genera to the list (Table 2). According to some authors [9, 12, 13, 14, 15], Cordyceps militaris (L. : Fr.) Fr., the type species of Cordyceps Fr., was already described in the 17th and early 18th century literature under old generic names, including: Funguli clavati; ex gracili caule paulatim crassiores redditi; ad digit minimi fere longitudinem pallid accedentes (tentative translation: clavate, slender, gradually broadening, somewhat longitudinally similar to finger or toe, pale toward tip) [16, 17]; Fungus parvus luteus ad ophioglossoiden nigrum accedens (tentative translation: small, yellow, tongue-like, black toward tip) (common English name: yellow adderstongue mushroom) [18, 19]; and Fungoides clavatum minus [6].

Table 1

Early fungal genera

aFungal groups described under each genus are as follows (in parenthesis): Fungus (agaricus, boletus), Boletus (morels, clathrus, and phallus), Agaricus (polypores, bracket fungi), Lycoperdon (puff-balls), Coralloides (clavarioid forms), and Tubera (truffles).

Table 2

Major fungal genera of Micheli [9] and Von Haller [10, 11]

OTHER GENERIC NAMES FOR CORDYCEPS FR.

Vaillant [8] reserved the name Coralloides Tourn. for branched forms and erected a new genus, Clavaria Vaill. ex L., for unbranched species that also included Cordyceps militaris (L. : Fr.) Fr. (as Clavaria militaris, crocea Vaill.), thus becoming the first extant generic name used for Cordyceps s. l. Clavaria became such a popular name for basidiomycetous genera that throughout the 18th century it was commonly used for several species of Cordyceps s. l. such as Clavaria capitata Holmsk. and Clavaria pileata Retz. for Cordyceps capitata (Holmsk.) Link (current name Elaphocordyceps capitata (Holmsk.) G. H. Sung et al.); Clavaria lutea minima Mich., Clavaria militaris L., C. lavaria simplex oblonga pulvinata Schmidel, Clavaria squamosa Lam., and Clavaria granulosa Bull. for Cordyceps militaris; and Clavaria parasitica Willd. and Clavaria radicosa Bull. for Cordyceps ophioglossoides (Ehrh.) Link (current name Elaphocordyceps ophioglossoides (Ehrh.) G. H. Sung et al.); and Clavaria sobolifera Hill ex Watson for Cordyceps sobolifera (Hill ex Watson) Berk. & Broome (current name Ophiocordyceps sobolifera (Hill ex Watson) G. H. Sung et al.). However, Buxbaum [12] retained the genus name Coralloides Tourn. for Cordyceps, e.g., Coralloides clavata, lutea, minor was used for Cordyceps militaris.

Von Haller [11] established a new ascomycetous genus, Sphaeria Hall., to distinguish species with rounded, entire perithecia, furnished at the apex with a minute orifice, and asci converging and dissolving at length. Based on these characteristics, several Cordyceps species s. l. were transferred from Clavaria into Sphaeria thus departing from the traditional basidiomycetous classification, including: Sphaeria agariciformis Bolton/S. capitata (Holmsk.) Pers. (=/≡ Cordyceps capitata, current name Elaphocordyceps capitata); S. entomorrhiza Dicks. (≡ Cordyceps entomorrhiza (Dicks.) Fr., current name Ophiocordyceps entomorrhiza (Dicks.) G. H. Sung et al.); S. gunnii Berk. (≡ Cordyceps gunnii (Berk.) Berk.); S. huegelii Corda/S. larvarum Westwood (≡/=Cordyceps huegelii Corda, current name Ophiocordyceps larvarum (Westwood) G. H. Sung et al.); S. innominata R. Taylor/S. taylorii Berk. (=/≡ Cordyceps taylorii (Berk.) Sacc., current name Ophiocordyceps taylorii (Berk.) G. H. Sung et al.); S. robertsii Hook. (≡ Cordyceps robertsii (Hook.) Berk., current name Ophiocordyceps robertsii (Hook.) G. H. Sung et al.); S. militaris (L.) J. F. Gmel. (≡ Cordyceps militaris); S. ophioglossoides J. F. Gmel. and S. radicosa De Candolle (≡ Cordyceps ophioglossoides, current name Elaphocordyceps ophioglossoides); S. sinensis Berk. (≡ Cordyceps sinensis (Berk.) Sacc., current name Ophiocordyceps sinensis (Berk.) G. H. Sung et al.); S. sobolifera (Hill ex Watson) Berk. (≡ Cordyceps sobolifera, current name Ophiocordyceps soboliera); and S. sphecocephala Klotzsch ex Berk. (≡ Cordyceps sphecocephala (Klotzsch ex Berk.) Berk. & M. A. Curtis, current name Ophiocordyceps sphecocephala (Klotzsch ex Berk.) G. H. Sung et al.). By the early 19th century, Sphaeria had already become a large genus consisting of more than 500 species, necessitating its division into Sections, Tribes, and Series [20].

ORIGIN OF THE GENUS NAME CORDYCEPS AND ITS SYNONYMS

Cordyceps was first coined as a genus in Pyrenomycetes by Fries [21] from a combination of the Greek word cordyle, meaning a club, and the Latin word caput, meaning a head. However, Fries [20] himself downgraded Cordyceps to the tribe level of Sphaeria and described it as stroma erectum, caulescens, simplex, ramosum, imarginatum, stipite sterili suffultum, perithecia peripherica, demum prominentia, ostiolis aequalibus (tentative translation: having erect stroma with a simple or branching stem, a sterile stalk supporting the perithecia at the periphery, and projecting with openings at the apex). The treatment of Fries [20] was soon reversed by Link [22], who reaffirmed the genus-level status of Cordyceps. However, the treatment of Fries [20] had such a profound effect on taxonomists that Cordyceps species were named under Sphaeria for many decades.

The genus Cordyceps was first conserved in the Seattle Code (1972) with citation Cordyceps (E. M. Fries) Link, Handb. 3: 346 (1833). However, in subsequent Botanical Codes, citations for Cordyceps differed from Cordyceps E. M. Fries, Observ. Mycol. 2: 316 (1818, cancel page) in Sydney Code (1983) to Cordyceps Fr., Observ. Mycol. 2 (revis.): 316 (1824) in Tokyo Code (1994) [23]. The citation of the genus Cordyceps in the Sydney Code is considered to be correct, i.e., Cordyceps Fr., Observ. Mycol. 2(revis.): 316, 1818, nom. cons. [23, 24, 25].

Type species: Cordyceps militaris (L. : Fr.) Fr., Observ. Mycol. 2(revis.): 317 (1818).

The phylogenetic clade of Cordyceps s. s. was recently delimited [25]. Nearly 20 different genera have been reported as synonyms of Cordyceps Fr. in various sources (http://www.speciesfungorum.org, http://www.mycobank.org, http://www.cybertruffle.org.uk). We recognize only those genera that represent Cordyceps militaris or any other Cordyceps species in the clade Cordyceps s. s. of Sung et al. (for instance C. tuberculata) as true synonyms of Cordyceps Fr., [25] and they are listed below.

Cordyceps Fr

≡ Clavaria Vaill. ex L., Sp. Pl. 2: 1182 (1753).

≡ Sphaeria Hall., Hist. Stirp. Helv. 3: 120 (1768).

≡ Hypoxylum Juss., Gen. Pl.: 6 (1789).

≡ Cordylia Fr., Observ. Mycol. 2(original): 316 (1818), nom. illegit., non Pers. (1807) [Caesalpiniaceae].

≡ Xylaria Hill ex Grev., Scott. Crypt. Fl. (Edinburgh) 2: 2 (1823).

≡ Corynesphaera Dumort., Comment. Bot. (Tournay): 92 (1822).

≡ Sphaeria "trib." Cordyceps (Fr.: Fr.) Fr., Syst. Mycol. 2: 323 (1823).

≡ Kentrosporium Wallr., Beitr. Bot. 1: 163 (1844).

≡ Akrophyton Lebert, Z. Wiss. Zool. (Leipzig) 9: 449 (1858).

≡ Torrubia Lév., in Cesati & de Notaris, Comm. Soc. crittog. Ital. 1: 192 (1863).

≡ Phytocordyceps C. H. Su & H. H. Wang, Mycotaxon 26: 338 (1986).

However, we do not recognize the following genera as true synonyms of Cordyceps s. s. [25]:

Mitrasphaera Dumort., Comment. Bot. (Tournay): 92 (1822).

Cordyliceps Fr., 1832.

Campylothecium Ces., 1846.

Polistophthora Lebert, Z. Wiss. Zool., (Leipzig) 9: 452 (1858).

Racemella Ces., Comm. Soc. crittog. Ital. 1: 65 (1861).

Tettigorhyza G. Bertol., Mem. R. Accad. Sci. Ist. Bologna, Ser. 3 5: 574 (1875).

Campylothecium Ces. represented Cordyceps myrmecophila Ces., Polistophthora Lebert represented Cordyceps sphecocephala, and Racemella Ces. represented Cordyceps robertsii, all of which are now transferred to the new genus Ophiocordyceps Petch emend. G. H. Sung et al. (Ophiocordycipitaceae G. H. Sung et al.) [25]. Similarly, Mitrasphaera Dumort. represented Cordyceps capitata that is now transferred to a new genus, Elaphocordyceps G. H. Sung & Spatafora (Ophiocordycipitaceae) [25]. Tettigorhyza G. Bertol. and Cordyliceps Fr. are not considered true synonyms as it is not clear which Cordyceps species they represent.

SUBGENERIC CLASSIFICATIONS AND AMENDMENT OF CORDYCEPS

As the number of species of Cordyceps s. l. increased, successive authors began arranging the genus on the basis of host and morphological characters. Here, the era of subgeneric classification of Cordyceps is tentatively divided into so-called artificial and natural classifications that correspond to the 19th and 20th century, respectively, although it is difficult to make a clear demarcation line.

Artificial classification.

The insect/fungal host, shape of the stromata, and position of the perithecia were the principal characters used for the subgeneric classifications of Cordyceps s. l. in the 19th century (Table 3). Some of the major works under this system are mentioned here. Wallroth [26] seemingly organized entomophyte species of Cordyceps s. l. (under Kentrosporium Wallr.) into capitata (stromate sphaerico, 3 spp.) and clavata (stromate clavato, 2 spp.). Fries [27] included both entomogenous and mycogenous species in Cordyceps and divided them into Entomogenae (2 spp.) and Mycogenae (2 spp.), solely based on the nature of the host. Tulasne and Tulasne [28] classified Cordyceps species under a different genus, Torrubia Lév., into Entomogenae (8 spp.) and Mycogenae (2 spp.), which were further subclassified, depending on the shape of the stroma, into Clavatae (3 spp.) and Capitatae (5 spp.), and Claviformes (1 sp.) and Capitatae (1 sp.), respectively. In slight contrast to preceding authors, Saccardo [29] classified Cordyceps species into three main groups: Eucordyceps Sacc. (entomogenous with immersed perithecia, 21 spp.), Racemella Ces. (entomogenous with superficial perithecia, 7 spp.), and Cordylia Tul. (mycogenous with immersed perithecia, 2 spp.). Cooke [30] based his classification of entomogenous species of Cordyceps on the stroma shape and branching pattern: capitulum globose or elliptical (18 spp.); stroma furcate, capitulum subglobose (3 spp.); stroma simple, clubs elongated (14 spp.); stroma furcate, clubs elongated (4 spp.); and clubs various, perithecia free (6 spp.).

Table 3

Subgeneric classificationsa of Cordyceps s. l. in the 19th century

aDelimitation of similar subgenera may differ from author to author.

bNumber of species are shown in parentheses.

None of the former authors designated the divisions or groups of Cordyceps as subgenera. It was Schroeter [31] who formally ranked the divisions of Cordyceps (under Torrubia Lév.) into the subgenera Racemella Ces. (species with perithecia partly embedded, 4 spp.) and Cordylia Tul. (both entomogenous and mycogenous species having entirely embedded perithecia, 4 spp.), and then further divided Cordylia into Eucordyceps Sacc. (entomogenous, 2 spp.) and Cordylia Tul. (mycogenous, 2 spp.) (Table 3). Massee [32], however, only recognized entomogenous species under Cordyceps and transferred any mycogenous species to Cordylia sensu Tul. He refrained from subgeneric classification of Cordyceps, but rather separated the species based on perithecial position such as perithecia entirely or partly immersed (40 spp.) and perithecia superficial (12 spp.) as well as on the septation of ascospores (Table 3). Broadly arranging Cordyceps species by the host type, Lindau [33] established two subgenera, Eucordyceps Lindau (16 spp.) for entomogenous species and Cordylia Tul. (2 spp.) for mycogenous species. From Table 3, it can be observed that in the 19th century there were sporadic tendencies of recognizing only entomogenous species under Cordyceps [26, 32, 34], although these tendencies were almost completely abandoned by the 20th century.

Natural classification.

By the middle of the 19th century, microscopic details of fructifications in Cordyceps species had been observed, such as the number of ascospores in an ascus; the mode of arrangement of ascospores in the ascus in one or two rows or in a crowded or irregular manner; the shape and color of ascospores; septation either unicellular and simple or divided by transverse, transverse and longitudinal, or even oblique septa; and ascospore length [35]. Tulasne and Tulasne [28] and others also investigated the micro-characters of the Cordyceps species; however, such micro-characters were barely utilized in subgeneric classifications of Cordyceps (Table 3).

The use of micro-characters for subgeneric classification of Cordyceps can largely be accredited to T. Petch, E. B. Mains, and Y. Kobayasi in the 20th century (Table 4). Despite his contribution to the description of over 150 species of Cordyceps spanning over three decades in the first half of the 20th century, Petch [36] did not produce his own classification. The exception is the establishment of a new genus, Ophiocordyceps Petch, that he erected to accommodate Cordyceps species with non-fragmenting ascospores [36]. The genus was not recognized by later authors; however, it was recently reinstated and emended, based on the phylogenetic classification of Cordyceps s. l. [25]. In the history of Cordyceps taxonomy, the most detailed revisionary work was conducted by Kobayasi [37], who recognized three subgenera: Ophiocordyceps (Petch) Kobayasi (having non-fragmenting ascospores) (15 spp.), Eucordyceps Kobayasi (entomogenous and mycogenous species having fragmenting ascospores) (116 spp.), and Neocordyceps Kobayasi (having obliquely oriented perithecia) (6 spp.). He further classified these subgenera into 5 sections and 11 subsections in total, which can be considered as a workable natural classification of Cordyceps. Other mycologists contributed revisions to Cordyceps classification based on regional distributions (Table 4). For example, based on the North American species, the genus Cordyceps was divided into four subgenera: Racemella (Ces.) Sacc. (perithecia superficial and free) (6 spp.), Cordyceps (perithecia partly or completely immersed) (33 spp.), Cryptocordyceps Mains (perithecia developing in a palisade-like layer) (1 sp.), and Ophiocordyceps (Petch) Kobayasi (1 sp.), of which the subgenus Cordyceps was further divided into four sections [38]. In a somewhat similar manner [37], Moureau [39] recognized three subgenera in Cordyceps based on African species: Torrubiella Boud. (2 spp.), Eucordyceps Lindau (34 spp.), and Ophiocordyceps Petch (1 sp.). Koval [40, 41] divided Cordyceps into four subgenera: Ophiocordyceps Petch (10 spp.), Cordyceps Kobayasi (43 spp.), Fusicordyceps Koval (10 spp.), and Neocordyceps (Kobayasi) Koval (2 spp.), which were further classified into nine sections. Kobayasi [42], in his second revision of Cordyceps, retained his previous three subgenera Ophiocordyceps (19 spp.), Eucordyceps (255 spp.), and Neocordyceps (8 spp.), but reduced the number of sections to seven. The subgeneric concept of his second revision [42] was not consistent with the previous revision [37]; for example, C. sinensis was classified in the subgenus Ophiocordyceps in the previous revision, but was transferred to the subgenus Eucordyceps in his second revision. Currently, around 400 entomopathogenic and mycoparasitic species are estimated in Cordyceps s. l. [25].

Table 4

Major subgeneric classificationsa of Cordyceps s. l. in the 20th century.

aDelimitation of similar subgenera may differ from author to author.

bNumber of species are shown in parentheses.

The revisionary work continued until the 1980s. Two new subgenera were added to Cordyceps; a new subgenus, Bolacordyceps O. E. Erikss., was proposed to include species that produce bola-ascospores, e.g., C. bifusispora O. E. Erikss. [43], whereas Zhang et al. [44] established a new subgenus, Megalocordyceps K. Zhang, C. Wang et M. Yan, to include species with unicellular ascospores, e.g., C. gansuënsis K. Zhang, C. Wang & M. Yan (current name Ophiocordyceps gansuënsis (K. Zhang, C. Wang & M. Yan) G. H. Sung et al.). However, the identification of O. gansuënsis is questionable and is also considered a synonym of O. sinensis [45].

In addition, new genera were erected to include Cordyceps species with abnormal characters. For example, two new genera, Sphaerocordyceps Kobayasi and Wakefieldiomyces Kobayasi, were erected to incorporate Cordyceps species having atypical ascospores [46]. Cordyceps palustris Berk. & Broome (= C. hormospora Möller) and C. ussuriensis Koval were transferred to Sphaerocordyceps due to their spherical secondary ascospores, whereas C. peltata Wakef. was transferred to Wakefieldiomyces based on the constriction at the middle of the filamentous ascospores and their subsequent division into two septate fragments.

Among the subgenera of Cordyceps s. l., the subgenus Eucordyceps Kobayasi [37, 42] is the largest (Table 4). However, Sung et al. [25] showed that the subgenera of Cordyceps (Table 4) were not only phylogenetically distant, but that their morphological characters, including those of their hosts, were also evolutionarily unrelated. Thus, the generic concept of Cordyceps Fr. was emended. As a result, the species of Cordyceps s. l. were reshuffled among four phylogenetic genera: 1) Cordyceps Fr. emend. G.H. Sung et al. (40 spp.) (Cordycipitaceae Kreisel ex G. H. Sung et al.); 2) Ophiocordyceps Petch emend. G. H. Sung et al. (146 spp.); 3) Elaphocordyceps G. H. Sung & Spatafora (21 spp.) (Ophiocordycipitaceae); and 4) Metacordyceps G. H. Sung et al. (6 spp.) (Clavicipitaceae (Lindau) Earle ex Rogerson). Among the four phylogenetic genera, Ophiocordyceps is the largest and is comprised of species distributed in all three morphological subgenera of Kobayasi [37, 42], indicating that the subgenera are polyphyletic. Recently, another phylogenetic genus, Tyrannicordyceps Kepler & Spatafora (Clavicipitaceae), was erected to accommodate six species of Cordyceps s. l. [47]. While nearly 175 species remain in the residual group of Cordyceps s. l. in Cordycipitaceae and many more species were missing from the phylogenetic analyses of Sung et al. [25], there is a high likelihood of establishing new phylogenetic genera in the future to accommodate those in the residual group as well as the missing ones. For example, Polycephalomyces Kobayasi was recently emended [48] to incorporate three residual Cordyceps species of Sung et al. [25] and four Ophiocordyceps species.

CONCLUSIONS

The present review of Cordyceps species s. l. highlights the old and deserted generic names given in the literature of the pre-Linnaean era as well as the generic synonyms associated with this genus in the post-Linnaean era. The genus Cordyceps has firmly stood against all its synonyms and has been used for approximately 200 years since its publication in 1818, proving itself as the oldest valid genus in Cordycipitaceae. Cordyceps, as a teleomorph fungus, has a clear generic concept based on characters such as perithecium, ascus, ascospore, part-spore, position of perithecia on stroma from lateral to apical, perpendicular to oblique, or superficial to immersed, and host specificity, among others [37, 38, 42]. Furthermore, Cordyceps is a traditionally used generic name for numerous medicinal insect fungi, especially in the eastern Asian countries [49, 50]. Recent amendment to Article 59 of the Melbourne Code has proposed protecting a single genus name for each fungal clade among the competing names on a priority basis, irrespective of the teleomorph or anamorph states of the organisms [51, 52, 53, 54, 55]. Cordyceps Fr., being the oldest valid genus name in the Cordyceps s. s. clade, ranks as the most preferred name over its competing names on the priority basis. Shrestha et al. [56] has recently discussed the phylogenetic relationship between Cordyceps Fr. and Beauveria Vuill. in connection with the nomenclatural issue of these two genera.