JOURNAL OF INVERTEBRATE PATHOLOGY 43, 133-139 (1984)

Entomopathogenic Species of the Hyphomycete Genus Tolypocladium

ROBERT A. SAMSON

Centraalbureau voor Schimmelcultures, P.O. Box 273, 3740 AG Baarn, The Netherlands

AND

GEORGE G. SOARES, JR. University of Florida, AREC Lake Alfred, Florida 33850 Received January 18, 1983; May 23, 1983

Two entomopathogenic species of the hyphomycete genus Tolypocladium are described in detail. Tolypocladium extinguens sp. nov. was found on larvae of Arachnocampa luminosa (Diptera; Mycetophilidae) in caves in New Zealand. Tolypocladium cylindrosporum W. Gams is shown to be pathogenic to Aedes sierrensis and A. australis and is reported from Plecia nearctica (Diptera; Bibionidae). The occurrence of entomopathogenic species in this normally soil-borne genus is discussed. KEY WORDS: Fungal pathogens, Tolypocladium cylindrosporum, T. extinguens, sp. nov., Diptera, Aedes.

INTRODUCTION showed that this fungus is Tolypocladium cylindrosporum (Soares et al., 1979; The fungal genus Tolypocladium was Soares, 1979).We iser and Pillai (1982)have erected by Gams (1971) with the type spe- described a similar fungus from Aedes aus- cies T. inflatum. The genus contained three tralis from New Zealand. Kish et al. (1974, soil-borne species resembling those of the 1977) isolated T. cylindrosporum from the genus Beauveria. The species are charac- love bug, Plecia nearctica (Diptera; Bibio- terized by white slow-growing cottony col- nidae), in Florida, and observed larval mor- onies. The conidiophores are usually short tality after reinfectivity tests. In this paper and bear lateral or terminal whorls of phial- a detailed description of this still little- ides, which have a swollen, sometimes cy- known entomopathogenic fungus is given lindrical base and thin, often bent necks. on the host and from pure cultures. The conidia are one celled, hyaline, and Recently another fungal pathogen on the formed in slimy heads. In addition to the glowworm, Arachnocampa luminosa (Dip- soil-borne species, Barron (1980, 1981) de- tera; Mycetophilidae), at Waitomo, New scribed two new species, T. parasiticum Zealand, was received at the Centraalbu- and T. trigonosporum, from rotifers. Re- cently Bissett (1983) revised the genus and reau voor Schimmelcultures. This isolate also belongs to Tolypocladium, but could added three species to it. not be placed among the described taxa and During the last years several Tolypo- is proposed as new. cladium isolates from insects have become available. In a study of the natural enemies Tolypocladium extinguens spec. nov. of the western treehole mosquito, Aedes sierrensis, in California, a fungal pathogen (Figs. 1,2) was observed (Sanders, 1972; Pinnock et Corpora hyphalia in hospite lo-25 x 4- al., 1973)and tentatively identified as Beau- 5 Frn. Coloniae in agaro malti (2%) post 14 veria tenella. However, later studies dies, ad 1 cm diam, primum albae pulver-

133 0022-2011/84$1.50 Copyr@ht 0 1984 by Academic Press, Inc. A11 n’nldr nF rsnmA..rtinn ;n 51”” +A.,,. rec*“,m 134 SAMSON AND SOARES

ulentae , deinde cremeae . Conidiophora ad several regularly verticillate branches, 100 pm longa, stipites 3-5 p,rn lati, iden- which terminate in whorls of phialides. tidem verticillati, in verticillos phialidum Phialides small, consisting of a swollen exeuntes. Phialides e ventro inflato et col- base and a thin neck, 3.5-9 x 1.5-2 km, lulo angusto constantes, 3.5-9 x 1.5-2.0 or sometimes reduced to a lateral conidi- km, nonnumquam subterminales collulum oferous peg or with one sympodial prolif- laterale ferentes vel semel sympodialiter eration, giving rise to conidia in slimy proliferentes; conidia in capitulis mucidis heads. Conidia one celled, ellipsoidal, hya- congregata, continua, elipsoidea, hyalina, line, smooth walled, 2-2.5 x 1.5-2 km; levia, 2.0-2.5 x 1.5-2.0 urn; in vitro ma- sometimes larger, subglobose to ellipsoidal jora ellipsoidae ad cylindrica, nonnumquan conidia are also formed, 3.5-4 x 3- reniformia, 3.0-5.5 x 1.5-2.0 pm. 3.5 km. Holotypus in Arachnocampa luminosu, Morin Cave, Waitomo in Nova Zelandia, Description in Vitro lectus a D. Williams et P. Johnston, in Herb. CBS praeservatus, cultus CBS Colonies on agar media grow very slowly 345.77, isolatus a G.J. Samuels. but sporulate abundantly within 7 days. On Etymology. Lat. extinquere = to extin- 2% malt extract agar reaches a diameter of guish (the light of the glowworms). 1 cm within 14 days, at first white powdery, later changing to cream. On YpSs and Sa- Description on the Host bouraud glucose agar similar growth but Host filled with hyaline hyphae, 1.5-4 p,rn colony color changing to pale aveilaneous. wide and cylindrical hyphal bodies, Conidiogenous structures as on the host but lo-25 x 4-6 km; surface covered by a conidia larger and ellipsoidal to cylindrical, dense felt of hyaline hyphae, 2.5-4 pm sometimes slightly curved, 3-5.5 x 1.5- wide, bearing numerous complex brarmhed 2.3 pm. On mealworm agar conidia are conidiophores. Conidiophores up to 100 both subglobose to ellipsoidal and more pm long, with stipe, 3-5 p,rn wide, bearing elongated and curved. On this medium the ENTOMOPATHOGENIC Tolypocladium SPECIES 135

FIG. 2. Camera lucida drawing of T. extinguens: (a) conidiophore on host, (b) conidia on host, (c) conidiophore on mealwormagar, (d, e) conidia on mealwormagar, (0 hyphal bodies within Aruchno- campa host. colonies change from white to pale yellow- T. extinguens differs from the described brown. Teleomorph not observed. taxa by its very slow growth in pure culture and by its subglobose to ellipsoidal, some- Material Examined times kidney-shaped, conidia. Only one Holotype on glowworm, Arachnocampa isolate was available for study. luminosa (Skuse), Mot-in Cave, Waitomo, New Zealand. Leg. D. Williams and P. Tolypocladium cylindrosporum W. Johnston. 27.01.82, in herb. CBS; CBS Gams (Figs. 3-7) 345.77, isolated from glowworm by G. J. Description on the Host Samuels. The fungus seems to be endemic in In mosquito larvae infection usually oc- glowworm populations, which normally curs through the external cuticle or the al- occur in large numbers in the caves of Wai- imentary canal, most commonly the tomo, New Zealand (Richards, 1964). Mor- midgut. Following infection, hyphae grow tality of the larval and worm stages is usu- and ramify first locally, then throughout the ally less than 5%, but it may reach up to hemocoel. Ellipsoidal to cylindrical hyphal 15-20% (G. J. Samuels, pers. commun.). bodies variable in size and shape, 9-25 x 136 SAMSON AND SOARES

3.5-5 p,m, budding from hyphae on short The morphology of the strains examined pegs (Fig. 3). Hyphal bodies or “blasto- is similar to soil-borne isolates (see Cams? spores” (Fig. 4) carried to other parts of 1971) except the phialides and conidia are the hemocoel where they initiate new in- occasionally longer. fection foci. Larvae (fourth instar Aedcs sierrensis) begin to die 3-7 days after ex- DISCUSSION posure to conidia when they have been ex- The isolation of T. extinguens from Ar- tensively colonized by the fungus (Fig. 5). achnocampa represents an additional ex- Larvae become packed with mycelia, ample of a Tolypocladium species infecting giving the cadavers a pale pink to cream an insect host. The realization that species color and cheese-like consistency. Hyphae within the genus Tolypocladium may be en- penetrate out of all parts of the body of the tomopathogens is a recent one and to date cadaver, growing outward in all directions, natural infections by these fungi have been ramifying. On cadavers at the surface of the documented in only four insect species. water, hyphae form a dense tuft of mycelia, When Gams (1971) established the genus usually at the tip of the siphon tube of cul- he examined mostly soil isolates. However, icine larvae (Fig. 6) and over the entire one strain of T. inflatum (CBS 716.70) was dorsal surface of anopheline larvae. Such isolated from Aradus cinnamomius (Hem- an aerial mycelium produces conidiophores iptera; Aradidae). Whether this was acting and conidia in slimy heads as on solid syn- as a true pathogen or as a saprophyte could thetic media. No sporulation occurs under not be established. water. For more detailed data on infectivity The paucity of isolates raises the ques- and pathogenesis see Soares (1982). tion of why Tolypocladium species have not been described more often from insects Description in Vitro in the past. One possibility is that Tolypo- Colonies grow well on 2% malt extract or cladium infections are rare in nature. How- oatmeal agar attaining a diameter of 1.5- ever, several Tolypocladium spp. are rela- 2.7 cm at 25°C in 10 days, forming a white, tively common inhabitants of soil (Bissett, cottony felt. Conidiophores short, hyaline, 1983). Lundgren et al. (1978) found inhibi- relatively complex, consisting of verticillate tory effects of three Tolypocladium spp. to branches bearing lateral or terminal whorls a number of other fungi. They suggested of phialides. Phialides consisting of a that this inhibitory ability may be of con- swollen cylindrical base. Conidia one siderable importance for the survival of celled, hyaline, smooth walled, cylindrical these fungi in the harsh environment of soil. with rounded ends, often slightly bent, The common occurrence in soil may also mostly 4-5 x 1.2-2.0 pm, occasionally indicate that Tolypocladium spp. have an longer, up to 8 pm. Teleomorph not ob- alternate host in the soil, e.g., nematodes. served. The morphological resemblance of T. cylin- drosporum to the nematophagous species Material Examined Harposporium baculiforme and H. si- CBS 989.73 isolated from Plecia nearc- cyodes (Drechsler, 1959), for example, is tica by L. P. Kish, IFAS, University of very striking and the susceptibility of nem- Florida, Gainesville. CBS 612.80 = N8-35; atodes to Tolypocladium is worthwhile ex- 613.80 = NS-29111; 614.80 = N2-B.t.1 iso- amining. lated from larvae of Aedes sierrensis in Cal- Another possible explanation for this ifornia by G. G. Soares. CBS 276.82, iso- scarcity of isolates may be the confusion of lated from Larvae of Aedes austrulis, Qoint Tolypocladium spp. with other similar and Point, Dunedin Otago Coast, New Zealand, better known fungal entomopathogens. by J. S. Pillai. How often in the past, for example, has in- ENTOMOPATHOGENIC Tolypocladium SPECIES

FIG. 3. Hyphal bodies being budded from hyphae of Tolypocladium cylindrosporum within the hemocoele of an infected Aedes sierrensis L2 larva. Phase contrast 640 x . FIG. 4. Hyphal bodies from the hemocoele of an infected Aedes sierrensis larva (L4). Phase con- trast, 440 x . FIG. 5. Hyphae of Tolypocladium cylindrosporum within the siphon tube hemocoele of an Aedes sierrensis L2 larva. H, hemocoele; T, tracheal trunk within siphon tube. Phase contrast, 640x. FIG. 6. Aedes sierrensis L4 larva killed by Tolypocladium cylindrosporum. Note sporulating surface tuft of mycelium arising from siphon and anal papillae of host. 138 SAMSON AND SOARES

FIG. I. Camera lucida drawings of Tolypocladium cylindrosporum: (a, b) canidiophores, (c-f) conidia, (c) CBS 614.M from A. sierrensis, (d) CBS 613.80 from A. sierrensis, (e) 983.73 hxn Pkcia nearctica, (f) 276.82 from A. australis, (g) Hyphai bodies within fourth instar larvae of A. sierrensis, (h) hyphal bodies budding from mycelium within host body.

fection caused by a Tolypocladium sp. been fungi producing a compactthyahue myce- attributed to a Beauveria sp. or other hy- lium. phomycetous fungi of similar morphology? It is interesting to note that all cases of The similarities of T. cylindrosporum and Totypocladiem infections retarded to date T. injlutum to Beauveria brongniartii and have been reported from I&&era of- the B. bassiana, respectively, are particularly same suborder, Nematocera, inchrding the striking and it seems very possible that To- glowworm, Arachnocampa. T&k swsts lypocladium isolates may have been mis- a somewhat greater host spe&fi&y than is identified as Beauveria species in the past normally associated with the hyphomyce- (Soares et al., 1979). Culture morphology tous entomopathogens such as Beauveria of all four species is very similar, with all and Metarhizium. This is important if these four producing compact hyaline mycelia. fungi are eventually to be used in insect All of these fungi also have flask-shaped pest management since they would be less conidiogenous cells with thin necks. Al- likely to infect certain bene&ial insect spe- though Beauveria spp. produce sympodial cies present in a given ecosystem. How- conidia on an elongating rachis, while To- ever, preliminary host range studies have lypocladium spp. produce conidia on phial- shown certain nondipteran hosts to be sus- ides, these differences are less evident in ceptible to T. cylindrosporamm. Notably, young cultures. Moreover, this distinction 100% of Galleria meblonella larvae were is not always as clear as might be supposed. killed by this fungus 10 days after contam- In any event, it is important to point out the ination of the integument with conidia similarities of these fungi to Tolypocladium (Soares, 1979). and stress the importance of careful ex- All entomopathogenic strains of Tolypo- amination of insect specimens infected by c/a&urn have also been isolated only from ENTOMOPATHOGENIC TolypocladiurnSPECIES 139 arthropod hosts occupying cryptic habitats, DRECHSLER, C. (1959). Two new species of Harpo- which are not easily sampled for pathogens. sporium parasitic on nematodes. J. Washington Acad. Sci., 49, 106-112. Thus casual isolations are less likely here. GAMS, W. (1971). Totypocladiwn, eine Hyphomyce- Indeed, the isolations from mosquitoes tengattung mit geschwollene Phialiden. Persoonia, were made as part of long-term ecological 6, 185-191. studies of the host species involved. In the KISH, L. P., ALLEN, G. E., KIMBROUGH, J. W., AND case of A. sierrensis surveys were con- KUITERT, L. C. (1974). A survey of fungi associated with the love bug, Plecia nearctica in Florida. Flu. ducted specifically for the purpose of iso- Entomol., 57 (3), 281-284. lating entomopathogens. KISH, L. P., TERRY, I., AND ALLEN, G. E. (1977). This underscores the importance of long- Three fungi tested against the love bug, Plecia term field surveys of pest populations for nearctica in Florida. Flu. Entomol., 60 (4), 291- the purpose of isolating pathogens. A 295. LUNDGREN, B. BAATH, E., AND SODERSTROM,B. E. greater emphasis on such surveys would no (1978) Antagonistic effects of Tolypocladium spe- doubt result in the isolation of many new cies. Trans. Brit. Mycol. Sot., 70, 305-307. pathogen species and strains, and it seems PINNOCK, D. E., GARCIA, R., AND CUBBIN, C. M. reasonable to assume that such surveys, (1973). Beauveria tenella as a control agent for mos- particularly of soil and aquatic habitats, quito larvae. J. Invertebr. Pathol., 22, 143-147. RICHARDS, A. M. (1964). The New Zealand glow- would also produce new records of Toly- worm. Stud. Speleol., 1, 38-41. pocladium species infecting insects. SANDERS,R. D. (1972). Microbial mortality factors in Aedes sierrensis populations. Proc. Calif. Mosq. Control Assoc., 40, 66-68. ACKNOWLEDGMENTS SOARES,G. G., JR., (1979). “Study of Tolypocladium The authors thank Dr. G. J. Samuels and P R. John- cylindrosporum Gams, a New Naturally Occurring ston (DSIR, Auckland New Zealand) for sending the Fungal Pathogen of Mosquitoes (with Notes on specimens of the killed glowworms and the data about Beauveria bassiana (Bals.) Vuill.).” Ph D. thesis, the disease. Dr. W. Gams kindly provided the Latin University of California, Berkeley. translation, while Mr. Keith Seifert made linguistic SOARES,G. G., JR., (1982). Pathogenesis of infection corrections. by the hyphomycetous fungus, Tolypocladium cy- lindrosporum Gams in Aedes sierrensis and Culex tarsalis (Diptera: Culicidae). Entomophagn, 27 (3) REFERENCES in press. SOARES, G. G. JR., PINNOCK, D. E., AND SAMSON, BARRON, G. L. (1980). Fungal parasites of rotifers, a R. A. (1979). Tolypocladium, a new fungal pathogen new Tolypocladium with underwater conidiation. of mosquito larvae with promise for use in microbial Cannd. J. Bat., 58, 439-442. control. Proc. Mosq. Control Assoc., 47, 51-54. BARRON, G. L. (1981). Two new fungal parasites of WEISER, J., AND PILLAI, J. S. (1981) Tolypocladium bdelloid rotifers. Cunad. J. Bat., 59, 1449-1455. cylindrosporum Gams (Deuteromycetes, Monili- BISSETT, J. (1983). Notes on Tolypocladium and re- ales), a new pathogen of mosquito larvae. Enro- lated genera. Canad. J. Bot. 61, 1311-1329. mophaga, 26 (4), 357-361.