A Disjunct Californian Strain ofEntomophaga aulicaeInfectingOrgyia vetusta

JOURNAL OF INVERTEBRATE PATHOLOGY ARTICLE NO.

68, 260–268 (1996)

0094

A Disjunct Californian Strain of Entomophaga aulicae Infecting Orgyia vetusta ANN E. HAJEK,1 SCOTT R. A. WALSH,1,2 DONALD R. STRONG,3

AND

JULIE C. SILVER2

Department of Entomology, Cornell University, Ithaca, New York 14853-0901 Received December 4, 1995; accepted July 3, 1996

Fungal epizootics occurred in abundant Orgyia vetusta (western tussock moth; Lepidoptera: Lymantriidae) populations on Lupinus arboreus bushes growing on the Pacific coast north of San Francisco, California. The causative pathogen was isolated and identified as Entomophaga aulicae, Group II, based on RFLPs using rDNA and PCR-amplified rDNA products. Inability of this fungus to infect the lymantriid Lymantria dispar (gypsy moth) confirmed its distinction from Entomophaga maimaiga, the only other member of this species complex which predominantly infects lymantriids. Later instar wandering by O. vetusta in outbreak populations and close proximity of larvae in dense populations are characteristics most probably promoting development of E. aulicae epizootics; these life history patterns are also typical of Lymantria dispar populations experiencing epizootics of E. maimaiga. r 1996 Academic Press, Inc.

KEY WORDS: Entomophaga aulicae; Entomophaga maimaiga; fungal entomopathogen; epizootic; Orgyia vetusta; Lymantria dispar; gypsy moth; DNA; PCR.

INTRODUCTION

The Entomophaga aulicae species complex contains strains that have been recorded infecting Lepidoptera from many different families (Hajek et al., 1991b). This species complex includes several members that are well known for the epizootics they cause, e.g., Entomophaga maimaiga in Lymantria dispar (gypsy moth) (Andreadis and Weseloh, 1990; Hajek et al., 1990) and Entomophaga aulicae in Euproctis chrysorrhoea (Speare and Colley, 1912), Estigmene acrea (Young and Sifuentes, 1959), Lambdina fiscellaria (Otvos et al., 1973), Choristoneura fumiferana (Perry and Re´gnie`re, 1986), 1 The first and second authors contributed equally to this study. The order of authors was determined solely alphabetically. 2 Division of Life Sciences, University of Toronto, Scarborough Campus, Scarborough, Ontario, M1C 1A4 Canada. 3 Bodega Marine Laboratory, University of California, Bodega Bay, California 94923.

0022-2011/96 $18.00 Copyright r 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

and three lepidopterous pests of sorghum (Hamm, 1980). All strains within this complex have pear-shaped conidia and hyaline resting spores that are so similar as to not be used to differentiate strains morphologically. At present, this complex contains two validly named species, E. aulicae and E. maimaiga. Recent studies comparing fungal DNA have identified several groups within the E. aulicae species complex (Walsh et al., 1990). It is unknown which group includes the strain infecting the arctiid Hyphoraia aulica in Europe from which E. aulicae was first mentioned in 1869 (MacLeod and Mu¨ller-Ku¨gler, 1973). The 1988 description of E. maimaiga (Soper et al., 1988) was based on differences in isozymes and host specificity, as well as biogeography, and was confirmed further using several different DNA-based analyses (Walsh et al., 1990; Hajek et al., 1991b). Strains within this species complex display a level of host specificity that is rare among entomopathogenic fungi; genetically different fungal strains are not always able to cause infections in hosts which are susceptible to other strains. For example, E. maimaiga, which can infect lymantriids, is generally not able to infect geometrids (Hajek et al., 1995), while none of the E. aulicae strains that have been tested are able to infect L. dispar (Soper et al., 1988) but E. aulicae strains can infect geometrids (A.E.H. and M. Bidochka, unpublished data). The E. aulicae species complex has a Holarctic distribution with records from Europe, Asia, and North America. Around the Pacific Ocean, members have been reported from British Columbia and Japan. The state of California is distant from areas where the E. aulicae species complex has previously been reported. We report the occurrence of epizootics caused by a member of the E. aulicae species complex in populations of the western tussock moth, Orgyia vetusta, on Lupinus arboreus (bush lupine) at Bodega Bay, along the Pacific coast north of San Francisco. The causative pathogen is identified and the biology of this pathogen is described and compared with other members of the E. aulicae species complex.

260

261

E. aulicae INFECTING O. vetusta MATERIALS AND METHODS

Fungal Isolation Larvae of O. vetusta were collected in late July 1993 on L. arboreus growing on the headlands at Mussel Point, Bodega Bay, California. Larvae were reared in the Ithaca laboratory on foliage of Lupinus polyphyllus 3 arboreus hybrids, commercially available garden perennials, and were checked daily for mortality. Any larvae that died were placed in humid chambers and were observed at frequent intervals. Once cadavers produced conidia, 20 healthy O. vetusta larvae were introduced into the chamber for several hours and then removed. The showered larvae were then reared at 20°C and 14:10 (L:D). For the first 2 days after inoculation, larvae were maintained at 100%RH but were then transferred to 60%RH. For 5 days after conidial inoculation, the fungus grew within larvae that had successfully become infected. On Day 5, larvae were surfacesterilized by briefly immersing them in 10% sodium hypochlorite. Under sterile conditions, a proleg was ablated from each larva and approximately 10 µl of hemolymph was collected. Hemolymph from each larva was added to individual tissue culture flasks containing 9.5 ml Grace’s insect tissue culture medium and 0.5 ml fetal bovine serum (GIBCO, Gaithersburg, MD) with 45 µl gentamicin sulfate (50 mg/ml). Flasks were maintained at 20°C without light. If fungal cells began growing, cultures were subcultured 1–2 times/week. All isolates were accessioned in the USDA-ARS Collection of Entomopathogenic Fungal Cultures (ARSEF) in Ithaca, New York. Fungal Identification Fungal isolates used in molecular comparisons. Table 1 lists the strains used for identification of isolates. These strains were obtained from the ARSEF Collection of Entomopathogenic Fungi, Ithaca, New York, and the Forest Pest Management Institute of the Canadian Forestry Service (FPMI), Sault Ste. Marie, Ontario, Canada. All fungal isolates were grown in axenic culture in Grace’s insect tissue culture medium supplemented with 5% (v/v) fetal bovine serum (GIBCO). Isolates were maintained by storage at 4°C between bimonthly transfers. DNA Extraction, Enzyme Digestion, Electrophoresis, and Southern Analyses Entomophaga genomic DNA was isolated from protoplast cultures as previously described (Walsh et al., 1990). Approximately 3 µg of each genomic DNA was digested to completion with a 10-fold excess of restriction enzyme EcoRI or HindIII (GIBCO-BRL). Electrophoresis on 0.8% agarose gels, transfer to Immobilon-N (Millipore, Bedford, MA), and subsequent Southern

TABLE 1 Isolates from the Entomophaga aulicae Species Complex Used for Strain Identification Species Entomophaga maimaiga

Isolate

Host

1390

Lymantria dispar (Lepid.: Lymantriidae) Lymantria dispar (Lepid.: Lymantriidae) Aedia leucomelas (Lepid.: Noctuidae) Lambdina fiscellaria l. (Lepid.: Geometridae) Choristoneura occidentalis (Lepid.: Tortricidae) Choristoneura fumiferana (Lepid.: Tortricidae) Dendrolimus spectabilis (Lepid.: Lasiocampidae) Mamestra brassicae (Lepid.: Noctuidae) Heterocampa guttivitta (Lepid.: Notodontidae) Heliothis spp.? (Lepid.: Noctuidae) unidentified lepidopteran Orgyia vetusta (Lepid.: Lymantriidae) Orgyia vetusta (Lepid.: Lymantriidae) Orgyia vetusta (Lepid.: Lymantriidae) Orgyia vetusta (Lepid.: Lymantriidae) Orgyia vetusta (Lepid.: Lymantriidae)

2775

Entomophaga aulicae (Group I)

198 491 a

599 a

646 a

1655

1751 2843

E. aulicae (Group II) E. aulicae (Group III) Entomophaga spp.

251 770 a 4004 4005 4006 4007 4008

Geographical origin Japan, 1984

CT, USA, 1989

Japan, 1976 BC, Canada, 1972 BC, Canada, 1976 NFLD, Canada, 1978 Japan, 1984

Japan, 1984 VT, USA, 1989

GA, USA, 1978 Switzerland, 1979 CA, USA, 1993 CA, USA, 1993 CA, USA, 1993 CA, USA, 1993 CA, USA, 1993

a Isolate from the Forest Pest Management Institute of the Canadian Forestry Service, Sault Ste. Marie, Ontario, Canada. All other isolates are from the Collection of Entomopathogenic Fungal Cultures at the U.S. Department of Agriculture, Agricultural Research Service (USDA-ARS), Ithaca, New York.

analyses under high-stringency conditions (50% formamide at 42°C) were performed using standard protocols (Sambrook et al., 1989). Probe Preparation The development and characterization of the probes used in the present study will be detailed elsewhere (Walsh and Silver, manuscript in preparation). Briefly, a repetitive DNA sequence hybridizing only to E. maimaiga DNA, EM-13, was identified from cloned genomic DNA of a United States isolate of E. maim-

262

HAJEK ET AL.

aiga. Similarly, an E. aulicae-specific DNA sequence, Q5, was identified from cloned genomic DNA of a Canadian isolate of E. aulicae. The clone containing ribosomal DNA (rDNA) sequences from E. aulicae, H34.0/pp1a-49, included the entire 18S rRNA gene, internal transcribed spacers, the 5.8S rRNA gene, and the majority of the 25S rRNA gene. Probe SW-20 used in RFLP assays was a cloned 1700-bp EcoRI fragment isolated from genomic DNA of a Canadian isolate of E. aulicae. Probes C1 and C4 were cDNA sequences constructed from E. aulicae total RNA isolated from protoplasts grown at 21°C in total darkness, followed by a 60-min exposure to light prior to harvesting. Probe C4 was approximately 1700 bp in length and showed high sequence homology to Hsp60 gene sequences from several organisms. Probe C1 was approximately 5700 bp in length and did not show significant homology with any DNA sequence in GenBank. Cloned inserts were electrophoretically purified from vector DNA, radiolabeled using the random-primer method (Feinberg and Vogelstein, 1983), and used as probes. Dot Blot Assays Approximately 1 µg of each purified genomic DNA was dotted onto nitrocellulose (Schleicher & Schuell [S&S] Inc., Keene, NH) under vacuum in a Manifold System II (S&S). DNA was fixed to the membrane for 2 hr at 80°C. Prehybridizations and hybridizations were performed under high-stringency conditions as previously described (Walsh et al., 1990). Radioactive blots were exposed to Kodak XAR film at 270°C for 6 hr. rDNA Primers and Amplification Conditions Primer 8-1T7 (58-TCCCTGCCCTTTGTACACACCGCCCGTC-38) and primer 12-1T7 (58-CCCCTAATTCCTCTAATCATTCGCTTTACC-38) were constructed according to the E. aulicae 18S and 25S rDNA gene sequences (Walsh and Silver, unpublished data), respectively. When used in the polymerase chain reaction (PCR), these primers resulted in the amplification of approximately 200 bp of the 38 end of the 18S gene, both internal transcribed spacers, the 5.8S gene, and approximately 1100 bp of the 58 end of the 25S gene from genomic DNAs from E. aulicae and E. maimaiga (Walsh and Silver, unpublished data). PCR amplifications were performed in volumes of 100 µl containing 2 mM MgCl2, 200 µM each dNTP (Boehringer-Mannheim), 50 pmol of each primer, 100 ng of template DNA, 10 µl of 103 Taq polymerase buffer (Appligene, Pleasanton, CA), and 1.5 units of Taq DNA polymerase (Appligene), overlaid with 100 µl of mineral oil (United States Biochemicals, Cleveland, OH). Amplifications were performed in a DNA thermal cycler (Perkin–Elmer Cetus, Norwalk, CT) programmed for an initial 93°C denaturation step of 1.5 min, followed by 40 cycles of 93°C for 1.5 min,

65°C for 1 min, and 71°C for 1.17 min. The fastest possible transition between temperatures was selected. The amplifications ended with an 8.5-min extension at 71°C and storage at 4°C. The PCR reaction was extracted with an equal volume of chloroform, and 12 µl of each reaction was analyzed by electrophoresis on 1.2% agarose gels containing ethidium bromide. Spore Measurements O. vetusta could not easily be reared in the Ithaca laboratory so larvae of O. leucostigma were substituted for laboratory studies. Larvae of O. leucostigma were obtained as eggs from the Forestry Canada, Forest Pest Management Institute in Sault Ste. Marie, Ontario, and reared on a high wheat germ diet (Bell et al., 1981). To measure conidia and resting spores of the Bodega Bay fungus, late instar O. leucostigma were infected in the laboratory with a mixture of ARSEF 4004, 4005, and 4006. Ten cadavers were placed in a humid chamber after host death to promote conidial production and discharge. Conidia were measured in a drop of water under a coverslip. To produce conidia in vitro, 100-mm-wide plates of EYSMA (egg yolk/Sabouraud maltose agar; Soper et al., 1988) were inoculated with 1 ml of fungal cells from tissue culture. Cultures required approximately 1 week at 20°C for production of mature mycelium. Fungal mycelium was removed from the media and conidia that were subsequently discharged were collected in water to measure length and width. O. leucostigma cadavers producing resting spores were maintained at room temperature for 1 week after host death for spore maturation and then at 4°C until resting spores were measured. Cadavers were soaked in water for several hours and gently macerated with a glass rod to separate resting spores from body contents. To measure resting spores that had been produced in O. vetusta in the field, soil was collected in February 1993 at Bodega Bay under three lupine bushes that had hosted infected larvae the previous summer. Resting spores were extracted from the soil using density gradient centrifugation (Hajek and Wheeler, 1994). Bioassays Larvae of L. dispar were obtained as eggs from USDA, APHIS, Otis Methods Development Center, and reared on the same diet as O. leucostigma. To conduct bioassays, fifth instar larvae were injected with 10 µl of 1 3 105 fungal cells/ml at the base of a proleg (Soper et al., 1988). Thirty larvae of O. leucostigma and L. dispar were injected with protoplasts for each of the five isolates. After injection, larvae were maintained at 20°C and 14:10 L:D and monitored daily. After host death, cadavers were placed at 100% RH and monitored for conidial production for 3 days. Cadavers were

E. aulicae INFECTING O. vetusta

263

subsequently placed at 4°C until dissection and microscopic evaluation to detect resting spore production. RESULTS

Fungal Isolation In late July, five isolates, ARSEF 4004, 4005, 4006, 4007, and 4008, were obtained from two different sporulating cadavers using the methods described. All attempts to isolate this fungus from conidia were unsuccessful. Isolates all grew as hyphal bodies with cell walls in tissue culture media for variable periods of time but gradually began growing as protoplasts, lacking cell walls, and, after 1–2 months in culture, isolates all grew principally as protoplasts. Fungal Identification Analyses of genomic DNA. In order to determine the species identity of the fungal isolates from O. vetusta, several different DNA-based analyses were carried out. Figure 1 shows the results of dot blot assays comparing the hybridization signals of DNA from these O. vetusta isolates with species-specific probes for either E. aulicae or E. maimaiga. DNA from isolates representative of different groups observed in the E. aulicae species complex (Walsh et al., 1990), and from two isolates of E. maimaiga, were included in the dot blot assays for comparison. As shown in Fig. 1A, the E. aulicae-specific probe hybridized to DNA from the E. aulicae group I (599) and group II isolates (251), but not to DNA from a putative European E. aulicae isolate (770) or that from the E. maimaiga isolates (1390, 2775). The signal obtained in hybridizations of this probe to DNA from the O. vetusta isolates (4004–4008) was as intense as that observed with DNA from the E. aulicae group I and II isolates. To verify that these isolates were not E. maimaiga, the same blot was stripped and hybridized with an E. maimaiga-specific probe (Fig. 1B). Although hybridization was not observed between this probe and any of the E. aulicae or the O. vetusta isolates, hybridization signals were evident with DNA from both E. maimaiga isolates. Since the signals from the two E. maimaiga DNAs (Fig. 1B, 1390 and 2775) were somewhat weak, an additional test was performed to ensure that each dot contained sufficient DNA present to generate a hybridization signal. To test for the amount of DNA loaded, the same blot (Fig. 1B) was stripped and hybridized with a cloned, radiolabeled E. aulicae rDNA probe, which is capable of hybridizing to rDNA sequences from all of the Entomophaga isolates (Fig. 1C). As shown in Fig. 1C, each dot was found to have sufficient DNA. The dot blot analyses with the speciesspecific probes could therefore be used to establish that the O. vetusta isolates belonged to the E. aulicae species

FIG. 1. Dot blots containing approximately 1 µg of each Entomophaga DNA were hybridized with the E. aulicae group I and II species-specific probe Q5 (Walsh et al., 1990) (A). The same blot was stripped and subsequently hybridized with the E. maimaiga speciesspecific probe EM-13 (B). To control for DNA loading, this blot was stripped and reprobed with H34.0/pp1a-49, a cloned E. aulicae sequence representing the coding regions and internal transcribed spacers of the rDNA cassette (C). Numbers above and below the figures represent the isolates listed in Table 1.

complex and were distinct from either E. maimaiga or European E. aulicae. Among isolates of E. aulicae, three groups are observed based on ribosomal DNA banding patterns (ribotypes) in Southern hybridizations, and these groups have been designated E. aulicae Groups I, II, and III, respectively (Walsh et al., 1990). In order to determine to which ribotype group the O. vetusta isolates belonged, Entomophaga DNAs as well as DNAs from the O. vetusta isolates were digested with the restriction

264

HAJEK ET AL.

enzymes HindIII or EcoRI and hybridized with the E. aulicae rDNA probe H34.0/pp1a-49 (please see Materials and Methods). As shown in Fig. 2A, in the HindIII digests, four bands (of approximately 4, 0.7, and a doublet at 0.9 kb) were observed in hybridizations with DNAs from E. aulicae Group I isolates (Fig. 2A, lanes 1–7). HindIII-digested DNA from the only isolate of Group III thus far identified, 770 (Fig. 2A, lane 14), had bands at approximately 3.3, 2.7, and 1.5 kb, while DNA from the Group II isolate, 251 (Fig. 2A, lane 13), showed bands at approximately 4.0, 1.5, and 0.9 kb. DNA from the two E. maimaiga isolates had hybridization patterns distinct from any of the E. aulicae patterns, with bands at approximately 5.2 and 1.3 kb (Fig. 2A, lanes 15 and 16). Finally, the DNA from all five isolates from O. vetusta (Fig. 2A, lanes 8–12) was found to have the E. aulicae Group II banding pattern, i.e., identical to that of isolate 251 (Fig. 2A, lane 13). Furthermore, an E. aulicae Group II banding pattern was also observed in hybridizations of the rDNA probe to EcoRI-digested DNA from the O. vetusta isolates (data not shown). Genomic DNA from the E. aulicae and E. maimaiga isolates and from the isolates from O. vetusta was used as template in the polymerase chain reaction with primers which annealed to the mid-region of the 25S rRNA gene and the 38-end of the 18S rRNA gene (Fig. 2B). The amplified products included the internal transcribed spacers and the 5.8S rRNA gene. As shown in Fig. 2B, an amplification product of approximately 2 kb was observed with target DNA from each E. aulicae Group I (Fig. 2B, lanes 1–8) and Group II isolate (Fig.

2B, lane 14). The Group II but not the Group I amplification products also included a smaller band at approximately 550 bp (Fig. 2B, lane 14). With target DNA from the Group III isolate, a major amplification product of approximately 2.3 kb (Fig. 2B, lane 15) was generated, while DNA from the E. maimaiga isolates (Fig. 2B, lane 16 and 17) resulted in an amplification product of 2.0 kb, as had been seen for the E. aulicae Group I isolates. Target DNA from the O. vetusta isolates (Fig. 2B, lanes 9–13) yielded amplification products similar to those observed using E. aulicae Group II DNA. The similarity between the isolates from O. vetusta and E. aulicae Group II was observed also in RFLP analyses using cloned E. aulicae cDNAs as probes (Fig. 3). EcoRI-digested DNA from E. aulicae Group I isolates probed with clone C1 (Fig. 3A) generated three different patterns, each with two or more bands (Fig. 3A, lanes 1–7). However, EcoRI-digested DNA from the E. aulicae Group II isolate (Fig. 3A, lane 13) resulted in only one band of approximately 10 kb. This same pattern was seen with DNA from the isolates from O. vetusta (Fig. 3A, lanes 8–12). Hybridization between probe C1 and DNA from E. aulicae Group III (Fig. 3A, lane 14), or E. maimaiga (Fig. 3A, lanes 15 and 16), was detectable only when the blots were exposed for more than 5 days and resulted in patterns different from that observed for the O. vetusta isolates (data not shown). When HindIII-digested Entomophaga DNAs were probed with a second cDNA probe, C4, two bands at 1.4 and at 1.2 kb were observed in DNA from both of the E.

FIG. 2. (A) Southern blot analyses of approximately 3 µg of each Entomophaga DNA digested with HindIII and hybridized with the E. aulicae rDNA probe, H34.0/pp1a-49. (B) Ethidium bromide-stained agarose gel electrophoresis of amplification products obtained from Entomophaga DNAs as template with primers 8-1T7 and 12-1T7 (refer to Materials and Methods). Numbers at the top of the lanes refer to the isolates listed in Table 1. Size markers (l DNA digested with DraI) in kilobases (Kb) are indicated on the left.

E. aulicae INFECTING O. vetusta

265

FIG. 3. Southern blot analyses of Entomophaga DNAs digested with EcoRI (A) or HindIII (B) and hybridized with probe C1 (A) or probe C4 (B). Variation in band intensities among the E. aulicae isolates reflect the amounts of DNA loaded. Numbers at the tops of the lanes refer to the isolates listed in Table 1. Size markers (l DNA digested with DraI) in kilobases (Kb) are noted on the left.

aulicae Group I (Fig. 3B, lanes 1–7) and E. maimaiga isolates (Fig. 3B, lanes 15 and 16). Bands migrating at 7.2 and 2.5 kb were observed with DNA from E. aulicae Group III (Fig. 3B, lane 14). However, DNA from the O. vetusta isolates (Fig. 3B, lanes 8–12) exhibited the same RFLP patterns with probe C4 as did DNA from the E. aulicae Group II isolate (Fig. 3B, lane 13) with bands at 1.5 and 1.2 kb. The fungal isolates from O. vetusta appeared identical to those of E. aulicae Group II in dot blot analyses with probes Q5 and EM-13 (Figs. 1A and 1B), in RFLP analyses using the rDNA probe H34.0/pp1a-49 (Fig. 2A), in PCR analyses of part of the rDNA cassette (Fig. 2B), and in RFLP analyses using two different cDNA probes (Figs. 3A and 3B). However, some genetic variation was observed among the isolates from O. vetusta and ARSEF 251 (E. aulicae Group II) when HindIIIdigested DNA from the O. vetusta isolates was probed with either probe C1 (Fig. 4A) or probe SW-20 (Fig. 4B). In particular, in hybridizations with probe C1, a band migrating at approximately 3.9 kb was observed in HindIII-digested DNA from ARSEF 251 (Group II isolate) (Fig. 4A, lane 13), but was not observed in similarly digested DNA from the isolates of O. vetusta (Fig. 4A, lanes 8–12). All other bands in these digests were otherwise identical (Fig. 4A, lanes 8–13). It is noteworthy that probe C1 was able to find polymorphisms also in the DNAs from the different E. aulicae Group I isolates (Fig. 4A, lanes 1–7). When hybridized with probe SW-20, the HindIIIdigested Entomophaga DNAs each exhibited a repeti-

tive banding pattern. DNA from each of the seven E. aulicae Group I isolates (Fig. 4B, lanes 1–7) resulted in seven distinct band patterns. DNA from the O. vetusta isolates showed similar patterns to one another; however, several polymorphisms were observed relative to DNA from ARSEF 251 (Group II isolate) (Fig. 4B, lanes 8–13). Hybridization of probe C4 with HindIII-digested DNA from the E. aulicae Group III isolate (Fig. 4B, lane 14) or from the E. maimaiga isolates (Fig. 4B, lanes 15 and 16) could be observed only with exposure for more than 5 days and resulted in very different patterns from those seen with DNA from the O. vetusta isolates (data not shown). Thus, with the exception of RFLP analyses carried out with HindIII-digested DNA and anonymous probes C1 and SW-20, which were also capable of distinguishing polymorphisms in DNA from E. aulicae Group I isolates, all other RFLP and PCR studies indicated that the fungus isolated from O. vetusta was E. aulicae Group II. Spore size. Hyaline, smooth-walled resting spores (azygospores), and pear-shaped conidia of the Bodega Bay pathogen were of the general appearance characteristic of the E. aulicae species complex (MacLeod and Mu¨ller-Ku¨gler, 1973). The sizes of spores produced both in vivo and in vitro are presented in Table 2. Sizes of both conidia and resting spores were generally in the range of the E. aulicae species complex (MacLeod and Mu¨ller-Ku¨gler, 1973). As is typical of entomophthoralean fungi, spores produced in vitro were larger than

266

HAJEK ET AL.

FIG. 4. Southern blot analyses of HindIII-digested Entomophaga DNAs hybridized with probe C1 (A) or probe SW-20 (B). Numbers at the top of the lanes refer to the isolates listed in Table 1. Size markers (l DNA digested with DraI) in kilobases (Kb) are noted on the left.

in vivo-produced spores. Resting spores were never produced in vitro. Host specificity. None of the L. dispar larvae that were injected with any isolate from Bodega Bay died, while all isolates caused high levels of infection in O. leucostigma larvae (mean 5 99.2%; SE 5 0.8%). O. leucostigma larvae died in a mean of 4.5 days (SE 5 0.3 TABLE 2 Measurements of Conidia and Resting Spores of E. aulicae Isolated from O. vetusta Length (µm) Conidia

n

x 6 SE

Width (µm)

Range

x 6 SE

Range

Source Discharged from O. leucostigma cadaver 60 34.0 6 0.4 27.3–40.3 22.3 6 0.3 18.2–28.6 Discharged from 50 44.4 6 0.8 31.2–59.8 32.1 6 0.8 23.4–46.4 EYSMA a Diameter (µm) Resting spores

n

x 6 SE

Range

Source Extracted from soil from Bodega Bay 31 44.2 6 0.9 32.5–61.1 Produced in O. leuco300 40.1 6 0.4 27.3–54.6 stigma cadavers b a

Wall thickness (µm) x 6 SE

2.8 6 1.0 2.7 6 0.1

EYSMA, egg yolk/Sabouraud maltose agar (Soper et al. 1988). 60 resting spores measured for each of five cadavers. Values presented are means of average/cadaver. b

across isolates) and for all isolates .70% of cadavers produced both resting spores and conidia. These results are consistent with the conclusion that the Bodega Bay isolates were E. aulicae. Unfortunately, further specificity testing with strains originating from O. vetusta became impossible due to contamination of the Bodega Bay strains in storage. Difficulties in the original isolations coupled with low-density populations of O. vetusta after the 1992 epizootic deterred attempts at reisolation. Natural History Caterpillars of the univoltine O. vetusta feed upon the foliage, flowers, and pods of L. arboreus at the study sites on the Bodega Marine Reserve (Bodega Bay, CA) (Barbour et al., 1973; Harrison, 1994). Eggs hatch in April to May from cases overwintering on stems and branches. Larvae pass through five (male) or six (female) instars and then pupate from June to August. A few caterpillars remain on foliage through September. By July, many caterpillars are in the penultimate and ultimate instars, when individuals infected with E. aulicae become conspicuous on the foliage of the host plant. Populations of O. vetusta vary greatly in intensity (individuals per unit host plant) among lupine stands (Harrison and Wilcox, 1995) and fluctuate through several orders of magnitude in intensity among years (Davidson and Barbour, 1977; Strong et al., 1995). The incidence of E. aulicae was particularly high among late instar larvae in July 1992, a zenith year of caterpillar intensity. In years with lower intensities of

E. aulicae INFECTING O. vetusta

O. vetusta, such as 1994, incidence of the disease was so low that infected caterpillars were difficult to find. DISCUSSION

Hybridization of DNA from the O. vetusta isolates with the E. aulicae-specific and E. maimaiga-specific DNA probes identified these strains as members of the E. aulicae species complex, but distinct from E. maimaiga. These strains did not possess sequences homologous to the E. maimaiga-specific DNA probe, further supporting the distinction between these lymantriid fungal strains and E. maimaiga isolated from the lymantriid, Lymantria dispar. Inability of O. vetusta isolates to infect L. dispar supported previous findings that E. aulicae isolates cannot infect L. dispar (Soper et al., 1988; Hajek et al., 1991b). Conversely, the fact that E. maimaiga, another member of the same species complex, can infect L. dispar supports the differentiation of E. maimaiga within this complex and reinforces the specificity of E. maimaiga for L. dispar. The lymantriid fungal isolates from Bodega Bay were found to be distinct from isolates representative of fungal strains causing the nearest recorded epizootics involving members of the E. aulicae species complex. The Japanese and British Columbian E. aulicae isolates had a Group I pattern, identical to that observed with the east coast E. aulicae isolates, while the O. vetusta isolates were found to possess a Group II pattern based on analyses of RFLPs in rDNA, PCRamplified rDNA products, RFLP studies in EcoRIdigested DNA with probe C1, and in HindIII-digested DNA with probe C4. In earlier studies, among 30 E. aulicae isolates representative of seven families of lepidopteran moths and several geographic origins across North America and Japan, only two Group II isolates were observed (Walsh et al., 1990; Walsh and Silver, unpublished). These were isolated from an arctiid moth in Ontario and a noctuid moth in Georgia. This latter isolate was used as the E. aulicae Group II reference isolate in the current study. As a result of the studies herein, E. aulicae Group II isolates are now known from three lepidopteran families, one of which, the Noctuidae, also hosts E. aulicae Group I isolates (Walsh et al., 1990). The Group II isolates from California displayed an ability to infect lymantriids, but demonstrated a differential specificity within this family, being able to infect O. vetusta and O. leucostigma, but not L. dispar. Laboratory bioassays used to conduct these tests maximize chances for infection if a host is permissive for the pathogen. Although clearly a member of E. aulicae Group II based on the analyses above, in some combinations of RFLP probes and restricted DNA, it was possible to distinguish the O. vetusta isolates from the ARSEF 251 (the Group II reference isolate). These same combina-

267

tions were also capable of distinguishing different isolates of E. aulicae Group I. The identification of an E. aulicae Group II isolate infecting lymantriids is interesting in light of immunological studies comparing antigenic differences on protoplast surfaces between members of E. aulicae and E. maimaiga (Hajek et al., 1991a). Although incapable of infecting the lymantriid, Lymantria dispar, one isolate of E. aulicae possessed significant cross-reactivity with E. maimaiga antisera. This isolate was the E. aulicae Group II reference isolate used in the present study. Another isolate of the E. aulicae species complex, taken from an epizootic where only E. aulicae Group I isolates were identified, showed no cross-reactivity to the E. maimaiga sera. It is not known if the E. aulicae Group II isolates identified from O. vetusta also possess crossreactivity to E. maimaiga antisera. Similarly, infectivity tests have not been performed with ARSEF 251 (Group II reference isolate) and lymantriid larvae. However, it is interesting to note that the E. aulicae species complex member that is most immunologically similar to the lymantriid-specific E. maimaiga is an E. aulicae Group II isolate and that the only E. aulicae member outside E. maimaiga known to infect lymantriid larvae also belongs to E. aulicae Group II. Although O. vetusta inhabits a much different habitat than L. dispar, similarities in life history exist between these two species. For both of these univoltine hosts, females are flightless, eggs and larvae generally remain on host plants, and larvae are herbivorous, feeding on foliage of woody plants. The foliage of L. arboreus bushes eaten by O. vetusta is much closer to the ground than the foliage of deciduous trees preferred by L. dispar. However, both moth species reach outbreak densities and, when dense, late instar larvae wander on the ground. E. maimaiga resting spores occur in abundance in soil at the bases of trees (Hajek and Wheeler, 1994) and E. aulicae resting spores were found in soil below L. arboreus bushes. During wandering, late instars might become infected by germ conidia produced by resting spores in the soil. In agreement with this hypothesis, E. maimaiga infections are much more abundant in later instar L. dispar and E. aulicae infections have only been found in later instar O. vetusta. E. aulicae epizootics in C. fumiferana are also assumed to be initiated when later instars emerge from buds and wander on the forest floor (Perry and Re´gnie`re, 1986). For both O. vetusta and L. dispar, when populations are dense, larvae feed and rest in close proximity. For both systems, this would facilitate secondary transmission of the disease (from cadavers to healthy larvae), leading to the development of epizootics. Epizootics of E. aulicae in O. vetusta occurred directly along the coastal headlands at Bodega Bay. Due to dense summer fogs that frequently occur throughout

268

HAJEK ET AL.

the day, this environment provides a cool, moist habitat without abundant ultraviolet radiation (Barbour et al., 1973). Such conditions optimize the production and survival of fungal spores and this, in turn, would promote infection. Summer fogs are also frequently associated with decreased winds which would allow airborne conidia to remain in the localized area and would optimize conditions required for infection. ACKNOWLEDGMENTS We thank J. Perry and M. Wheeler for excellent technical assistance. S. Harrison provided insights concerning the ecology of O. vetusta. J. deBenedictis, D. Dahlsten, R. Humber, Y. Tanada, and G. Thomas provided information about this pathogen and host and confirmed the absence of prior knowledge of this disease in California. This work was supported in part by USDA, NRICGP No. 92-37302-7658 (A.E.H.) and an NSERC (Canada) Strategic Grant (J.C.S.). REFERENCES Andreadis, T. G., and Weseloh, R. M. 1990. Discovery of Entomophaga maimaiga in North American gypsy moth, Lymantria dispar. Proc. Natl. Acad. Sci. USA 87, 2461–2465. Barbour, M. G., Craig, R. B., Drysdale, R. R., and Ghiselin, M. T. 1973. ‘‘Coastal Ecology: Bodega Head.’’ UC Press, Berkeley, CA. Bell, R. A., Owens, C. D., Shapiro, M., and Tardif, J. G. R. 1981. Development of mass rearing technology. In ‘‘The Gypsy Moth: Research Toward Integrated Pest Management’’ (C. C. Doane and M. L. McManus, Eds.), pp. 599–633. U.S. Dept. Agric. Tech. Bull. 1584. Davidson, E. D., and Barbour, M. G. 1977. Germination, establishment, and demography of coastal bush lupine at Bodega Head, California. Ecology 58, 592–600. Feinberg, A. P., and Vogelstein, B. 1983. A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal. Biochem. 132, 6–13. Hajek, A. E., and Wheeler, M. M. 1994. Application of techniques for quantification of soil-borne entomophthoralean resting spores. J. Invertebr. Pathol. 64, 71–73. Hajek, A. E., Butler, L., and Wheeler, M. M. 1995. Laboratory studies of host range of the gypsy moth fungal entomopathogen Entomophaga maimaiga. Biol. Control 5, 530–544. Hajek, A. E., Butt, T. M., Strelow, L. I., and Gray, S. M. 1991a. Detection of Entomophaga maimaiga (Zygomycetes: Entomophthorales) using enzyme-linked immunosorbent assay. J. Invertebr. Pathol., 58, 1–9. Hajek, A. E., Humber, R. A., Elkinton, J. S., May, B., Walsh, S. R. A., and Silver, J. C. 1990. Allozyme and RFLP analyses confirm

Entomophaga maimaiga responsible for 1989 epizootics in North American gypsy moth populations. Proc. Natl. Acad. Sci. USA 87, 6979–6982. Hajek, A. E., Humber, R. A., Walsh, S. R. A., and Silver, J. C. 1991b. Sympatric occurrence of two Entomophaga aulicae (Zygomycetes: Entomophthorales) complex species attacking forest Lepidoptera. J. Invertebr. Pathol. 58, 373–380. Hamm, J. J. 1980. Epizootics of Entomophthora aulicae in lepidopterous pests of sorghum. J. Invertebr. Pathol. 36, 60–63. Harrison, S. 1994. Resources and dispersal as factors limiting a population of the tussock moth (Orgyia vetusta), a flightless defoliator. Oecologia 99, 29–37. Harrison, S., and Wilcox, C. 1995. Evidence that predator satiation may restrict the spatial spread of a tussock moth (Orgyia vetusta) outbreak. Oecologia 101, 309–316. MacLeod, D. M., and Mu¨ller-Ku¨gler, E. 1973. Entomogenous fungi: Entomophthora species with pear-shaped to almost spherical conidia (Entomophthorales: Entomophthoraceae). Mycologia 65, 823– 893. Otvos, I. S., MacLeod, D. M., and Tyrrell, D. 1973. Two species of Entomophthora pathogenic to the eastern hemlock looper (Lepidoptera: Geometridae). Can. Entomol. 105, 1435–1441. Perry, D. F., and Re´gnie`re, J. 1986. The role of fungal pathogens in spruce budworm population dynamics: frequency and temporal relationships. In ‘‘Fundamental and Applied Aspects of Invertebrate Pathology’’ (R. A. Samson, J. M. Vlak, and D. Peters, Eds.), pp. 167–170. Foundation Fourth Internat. Colloq. Invertebr. Pathol., Wageningen, Netherlands. Sambrook, J., Fritsch, E. F., and Maniatis, T. 1989. ‘‘Molecular Cloning: A Laboratory Manual,’’ 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Soper, R. S., Shimazu, M., Humber, R. A., Ramos, M. E., and Hajek, A. E. 1988. Isolation and characterization of Entomophaga maimaiga sp. nov., a fungal pathogen of gypsy moth, Lymantria dispar, from Japan. J. Invertebr. Pathol. 51, 229–241. Strong, D. R., Maron, J. L., Harrison, S., Connors, P. G., and Jefferies, R. L. 1995. High mortality, fluctuation in numbers, and heavy subterranean herbivory in bush lupine, Lupinus arboreus. Oecologia 84, 85–92. Speare, A. T., and Colley, R. H. 1912. ‘‘The Artificial Use of the Brown-tail Fungus in Massachusetts with Practical Suggestions for Private Experiment, and a Brief Note on a Fungous Disease of the Gypsy Caterpillar.’’ Wright & Potter, Boston, MA. Walsh, S. R. A., Tyrrell, D., Humber, R. A., and Silver, J. C. 1990. DNA restriction fragment length polymorphisms in the rDNA repeat unit of Entomophaga. Exp. Mycol. 14, 381–392. Young, W. R., and Sifuentes, J. A. 1959. Biological and control studies on Estigmene acrea (Drury), a pest of corn in the Yaqui Valley, Sonora, Mexico. J. Econ. Entomol. 52, 1109–1111.