JOURNAL OF INVERTEBRATE PATHOLOGY 74, 147-153 (1998)ARTICLE NO. IN984773

o

, Interactions between a Nosema sp. (Microspora" Nosematidae)and Nuclear Polyhedrosis Virus Infecting the Gypsy Moth,

, " Lymantria dispar (Lepidoptera" Lymantriidae)Leah S. Bauer,* Deborah L. Miller,* Joseph V. Maddox,t and Michael L. McManus_

* USDA Forest Service, North Central Forest Experiment Station, Pesticide Research Center and Department of Entomology, Michigan State

University, East Lansing, Michigan 48823; tOffice of AgT_ultural Entomology and lllinois Natural History Survey, Champaign, Illinois 61820;

SUSDA Forest Service, Northeastern Forest Experiment Station, Hamden, Connecticut 06514

Received August 15, 1997; accepted March 11, 1998J

focused on the introduction and establishment of insect

Simultaneous and sequential per os inoculations of parasitoids and predators from its native range in

gypsy moth larvae withthe Lymantria dispar nuclear Europe (Reardon, 1981). Only a single attempt topolyhedrosis virus (LdNPV) and a Nosema sp. from introduce a gypsy moth pathogen is documented. ThisPortugal demonstrated that the interaction of two early biological control effort by Speare and Colleypathogens" during coinfection was variable, ranging (1912) involved the release of the "gypsy fungus," nowfrom synergistic to antagonistic. Susceptibility of gypsy known as Entomophaga maimaiga Humber, Shimazu,moth larvae to viral infection was unaffected by simul- and Soper, from Japan into eastern Massachusetts.

taneous and subsequent microsporidian infection. This Prior to the discovery of widespread E. maimaigaresulted from the comparatively slow pathogenesis ofthe microsporidium when compared to the virus. Viral epizootics in North America (Andreadis and Weseloh,infectivity, however, increased 10-fold when larvae 1990; Hajek et al., 1990), the collapse of local gypsywere preinfected with Nosema sp. per os, or through moth populations typically resulted from the L. clispartransovarial infection. Time to death decreased for nuclear polyhedrosis virus (LdNPV) (Lewis, 1981; Dw-

larvae infected with both pathogens compared to lar- yer and Elkinton, 1993; Elkinton and Liebho]d, 1990).vaeinfected with the virus alone. Polyhedron produc- First reported in North America in 1907, LdNPV wastion was significantly reduced by microsporidian infec- inadvertantly introduced with gypsy moth, as a contami-tion preceding viral infection. In this infection nant of gypsy moth parasitoids, or on imported hostsequence, larvae died at an earlier stage and were less plants (Hajek et al., 1995). Although both LdNPV andthan half the mass of cadavers infected with virus E. maimaiga are now considered vital to the naturaliza-alone. The biological significance of these results on tion of gypsy moth in North America (Hajek et al.,gypsy moth population dynamics and the implication 1993), these two pathogens represent a sma|] propor-

• for use of this Nosema sp. from Portugal in gypsy moth tion of the total number of microorganisms associatedbiological control are discussed in the context of viral with gypsy moth in Eurasia (Weiser, 1987).

• epizootiology. • © 1998AcademicPress Microspora is a ubiquitous group of obligate, intracel-Key Words: Lymantria dispar; gypsy moth; Nosema ]ular, protozoans commonly causing disease in insects

sp. from Portugal; nuclear polyhedrosis virus; biologi- and other invertebrates, and gypsy moth is no excep-cal control; pathogen interactions; gypsy moth viral tion. Five species of microsporidia are described fromepizootic; population dynamics. European gypsy moths, and at least 10 different iso-

lates were collected during the last decade in Europe(Solter et al., 1997). These pathogens, however, are

INTRODUCTION absent from gypsy moths in North America.

Following the accidental introduction of gypsy moth, In their normal host, microsporidia typically causeLymantria dispar (L.) (Lepidoptera: Lymantriidae), chronic disease expressed as prolonged developmental

' into North America over a century ago, control efforts time; reduced adult size, longevity, fecundity, mating,and egg fertility; and increased mortality in all develop-mental stages (Novotny and Weiser, 1993). Because of

i This article reports the results of research only. Mention of a the debilitating, rather than acute nature of microspo-proprietary product does not constitute an endorsement or a recom- ridiosis, these pathogens are considered importantmendation for its use by USDA. cofactors in maintaining gypsy moth population densi-

147 0022- 2011/98 $25.00Copyright © 1998 by Academic Press

All rights of reproduction in any form reserved.

148 BAUERET AL.

ties belo_v tolerance thresholds for long periods of time laboratory bioassays were used to predict what might-(Novotny, 1988, 1989; Weiser, 1987; Weiser and Novot- occur if this Nosema sp. were introduced and estab-ny, 1987; Zelinskaya, 1980). lished in a gypsy moth population where LdNPV is a

A species of Nosema, collected from gypsy moths in primary controlling factor. We compared the effect ofPortugal (J.V.M., unpublished data), is a promising Nosema infection on (1) the infectivity of LdNPV incandidate for biological control of gypsy moth in North microsporidia-free, per os-infected and transovarially °America (McManus et al., 1989). This microsporidium infected gypsy moth larvae, (2) the viral mortalityis receiving considerable research attention; areas be- response, and (3) the impact of coinfection on viral anding investigated include taxonomy and genetics, epizo- microsporidian pathogenesis in the gypsy moth.otiology, tissue specificity, host range, impact on hostbionomics, and interactions with other natural enemies(Jeffords et at., 1987, 1988, 1989; Onstad et al., 1993; MATERIALSAND METHODS

Solter et al., 1997). Favorable attributes of this micro- Insects. Gypsy moth egg masses, obtained from thesporidian species include high infectivity and low viru- USDA Forest Service, Northeastern Forest Experimentlence, permitting some larvae to reach the adult stage Station (USDA FS NEFES, Hamden, CN) laboratoryand transmission of the microsporidium within the colony, were surface sterilized (Bell et al., 1981) andhost egg (transovarially) (L.S.B., unpublished data), allowed to hatch. Neonate larvae were placed at aThus, gypsy moth larvae, hatching in the spring, density of 10 per cup (60 ml clear polystyrene) contain-sustain varying levels of infection. This strategy of ing high wheat germ diet (Bell et al., 1981) and rearedvertical transmission (between generations) better lim- at 24°C, L16:D8.its the microsporidium to its gypsy moth host and Transovarially infected neonates were produced byensures survival at low host population densities, inoculating parental stock as second-instar larvae with

In contrast, the relatively inefficient vertical transmis- Nosema dosages ranging from 4 × 102 to 4 × 103sion of LdNPV is achieved by the liberation of polyhe- spores/larva. Only neonates hatched from egg massesdra into the host environment the previous season after laid by infected females were included in the virallarval cadavers lyse on host trees. Gypsy moth neonate bioassays oftransovarially infected neonates.larvae become infected by ingesting the viral polyhedraduring egg hatch (Woods and Elkinton, 1987). Horizon- Pathogens. Fresh Nosema spores were produced bytal transmission (within a generation) is similar for inoculating second-instar gypsy moth larvae with ca.bothpathogens, involving ingestion of viral polyhedra 1 × 103 spores/larva, as described below. This dosageor microsp0ridian spores that contaminate the environ- assured 100% infection and a high yield of spores.ment. Because LdNPV is highly virulent, both host and Infected salivary glands and fat body were removedpathogen become scarce following a gypsy moth popula- from late fifth-instar larvae, homogenized in a hand-tion collapse, held glass tissue grinder with sterile tap water, and

Field studies of gypsy moth population dynamics in filtered through fine mesh fabric and thick cotton. TheCentral Europe suggest a close association between resulting spore suspension was centrifuged at 100g forLdNPV and microsporidia. Following a viral epizootic, 30 min. The pellet was resuspended in sterile tap watergypsy moth populations begin to recover and microspo- and centrifuged two more times; spores were furtherridian infections increase, reaching 15 to 30% as viral purified using a continuous density gradient of Ludox

• infections •begin to reappear. As host population density (0-100%) (Undeen and Alger 1971). The LdNPV lyophi-lized powder (Hamden isolate, LDP 226), obtained frompeaks, however, LdNPV is the most prevalent patho-

gen, and microsporidian prevalence declines. Postmor- USDA FS NEFES, was rehydrated in sterile tap watertem examination of cadavers reveals a high prevalence using a 1-ml tissue grinder. Concentrations of spores

and polyhedral inclusion bodies (polyhedra) were quan-of mixed vi'ral and microsporidian infections (Weiser, tiffed using a Petroff-Hausser cell counting chamber,1987;.Novotny, 1989, 1991; David and Novotny, 1990;Pilarska and Vavra, 1991; Novotny and Weiser, 1993). and dosages were prepared using serial dilutions.Weiser (19.87) suggests that microsporidiosis may act Viral and microsporidian interactions in microspo-as a stressing agent, increasing the susceptibility of ridia-free larvae. Microsporidia-free gypsy moth lar- ,gypsy moth to virus and lowering the host density vae were inoculated per os with four dosages of twothreshold required for viral epizootic induction. Under- pathogens in all combinations, totalling 16 treatmentsstanding the dynamic interactions between viral and per experiment with 24 larvae per treatment. The ,microsporidian infections is of particular interest in the experiments included (1) simultaneous inoculation ofpresent study. LdNPV and Nosema at third instar (dosages were 0,

Our objective was to evaluate the compatibility of 3 × 101, 3 × 102, 3 × 103 spores/larva and 0, 3 × 103,two obligate, intracellular pathogens for biological con- 3 × 104, 3 × 105 polyhedra/larva); (2) sequential inocu-trol of the gypsy moth. Sequential and simultaneous lation of Nosema at second instar (dosages were 0,

,

LdNPV AND Nosema SP. INTERACTIONS IN GYPSY MOTH 149

5 X 101_5 X 102, 5 X 103 spores/larva) followed by Ld- compare differences in time to death between separateNPV inoculations at third instar (dosages were 0, and dual treatments (Onstad et al., 1986) with the3 x 103, 3 x 104, 3 × 105 polyhedra/larva), resulting in Statistical Analysis System 6.03 (SAS Institute 1990).Nosema inoculations preceding LdNPV by 6 days; and Student's t test was used to test for significant differ-(3) sequential inoculation of LdNPV at third instar ences between the separate and dual pathogen treat-(dosages were 0, 3 x 103, 3 x 104, 3 x 105 followed by ments for larval dry mass, and production of infectiveNosema inoculation at fourth instar (dosages were 0, units (spores and/or polyhedra) per larva with Systat4 x 102, 4x 103, 4 x 104 spores/larva), resulting in 4.0 (Wilkinson, 1987).

' LdNPV inoculations preceding those of Nosema by 5 d.LarVae were inoculated by applying 1-ttl droplets, RESULTS

•containing pathogen dosages, onto 7 x 1-mm diet discsplaced singly in the wells of 24-well tissue culture Viral and microsporidian interactions. The IDso ofplates. Homogeneous cohorts of microsporidia-free sec- LdNPV in gypsy moth larvae was significantly higherond-, third-, or fourth-instar larvae, harvested from (10-fold) than in larvae infected 6 days before with 5 xdiet cups .less than 24 h after molt, were placed 103Nosema spores(Table 1).Atthe lower microsporid-individually into each well. Only larvae that consumed ian dosages, viral infectivity tended to increase as thethe entire dose within 24 h were included in the assay. Nosema dosage increased. No significant differences inAfter inoculation, larvae were reared individually in viral infectivity were measured when the two patho-60-ml cups and monitored daily for development and gens were inoculated simultaneously, or when LdNPVmortality. Cadavers were removed from diet cups daily, infection preceded Nosema infection (Table 1). Slopes ofplaced individually in labeled vials, dried (75°C), the viral ID'5oSwere similar at each dosage and se-weighed, and stored. Cadavers were rehydrated over- quence of Nosema.night in a known volume of water, homogenized in a Time to death for gypsy moth larvae infected with7-ml glass tissue grinder, and diluted sufficiently to only LdNPV was similar within and between the threefacilitate counting of polyhedra and spores using a experiments, with means ranging from 12 to 14 daysCompound microscope (400x) with a Petroff-Hausser (Table 2). Viral lethal times were also similar to thosecell counter. Within each experiment, the number of measured for larvae with dual infections, except whenpolyhedra or spores was counted in cadavers sampled Nosema infection preceded LdNPV infection. In thefrom the separate and dual treatments at the highest latter inoculation sequence, larvae with dual infection

dosages only.

Viral bioassays of larvae transovarially infected with TABLE 1Nosema. A cohort of transovarially infected neonate

Effect ofper osNosema Dosages on the Infectivity ofLdNPVlarvae, harvested from egg masses 24-48 h after initial in Gypsy Moth Larvaeegg hatch, were inoculated with a viral suspensionapplied to the diet surface of 60-ml cups. Each bioassay Nosema dosages LdNPV IDso a

included five 10-fold dilutions of virus (0.05-500 polyhe- (spores/larva) (95%FL)(polyhedra/larva) Slope+_SE

dra/cm 2) and_a water control and 24 larvae/concentra- Simultaneous Nosema andLdNPVtion: Larvae were allowed to feed on the viral-treated

0 1.6 x 104 (9.1 x 103-3.0 x 104) 1.56 _+0.28diet in groups of 4 larvae/cup for 48 h and transferred 3 x 101 7.8 X 103(2.3X 103-2.5X 104) 1.18 _ 0.20individually to 60-ml diet cups. Each larva was moni- 3 x 102 1.1 x 104(1.9x 103-3.8X 104) b 1.50 __0.29

tored for development and mortality. Spores and polyhe- 3 X 103 1.2 X 104 (5.6 X 103-2.2 X 104) 2.05 _ 0.50

dra were quantified from frozen individuals at a later Sequential Nosema preceding LdNPVdate. An uninfected cohort of neonates was also bioas-

0• 1.9 x 104 (1.0 x 104-3.7 x 104) 1.29 _ 0.23sayed simultaneously, in the same manner. Each bioas. 5 x 101 1.7 x 104 (3.4 x 103-5.8 x 104) 1.02 __0.28say was replicated four times. Since not all progeny 5 x 102 1.2 x 104 (2.9 x 103-3.2 x 104) 0.92 _+0.24

were infected with the microsporidium, only those 5 x 103 8.1 X 102(7.2-2.7x 103) b 0.98 __0.41

determined to be infected by postmortem examination Sequential LdNPV preceding Nosema

' were usedin the analyses. . 0 1.2 x 104 (6.8 x 103-2.1 x 104) 1.63 _+0.29

Statistics. Slope and maximum-likelihood estimate 4 x 102 1.3 x 104 (6.9 x 103-2.3 x 104) 1.48 +_0.26' 4 x 103 1.1 x 104 (3.2 x 103-4.0 x 104) 5 1.35 __0.23of viral median infective dosages (ID50s) were calcu- 4 x 104 2.5 x 104 (1.3 x 104-5.0 x 104) 1.27 __0.23lated with the probit option of POLO-PC (LeOra Soft-ware, 1987). The criterion for significance was failure of Note.IDso, medianinfectivedose; FL, fiducial limits; SE, standard

the 95% confidence limits to overlap. The LdNPV IDa0 error.a Each ID_oestimate includes mortality data from ca. 96 larvae perestimates were calculated at 20 days after inoculation bioassay, and was calculated 20 days after inoculation with LdNPV.with the .virus. The Kruskal-Wallis test was used to b90% FLmaximum resolution.

150 BAUER ET AL.

o

TABLE 2 TABLE 4

Time to Death (Days) for Gypsy Moth Larvae Infected Infectivity of LdNPV in Neonate Gypsy Moth LarvaeSeparately, Simultaneously, and Sequentially with Nosema Infected Transovarially with Nosema and Time to Death inand LdNPV Days

Time to death (days) a n LdNPV IDs0 Time toBioassay [mean (range)] (larvae) (90% FL) death

(polyhedra/cm 2 [mean _Simultaneous Larval health diet) Slope _ SE SE (days)]

Nosema 25.3 (19-32) 24 Microsporidia-free 48.2 (8.7-1181.1) 0.70 7.17 + 0.09LdNPV 14.0 (13-19) 24 - -Nosema andLdNPV 13.5 (11-19) 24 Microsporidia-infected a 3.5 (0.4-24.7) 0.71 __+0.14 7.62 _ 0.20

a Neonate larvae were infected transovarially with Nosema sp.Sequentialfrom infected mothers.

N osema 23.9 (5-34) 24LdNPV 12.0" (9-21) 22Nosema preceding LdNPV 10.0* (5-22) 23

, Sequential Cadavers resulting from this inoculation sequence pro-duced ca. 1000x fewer polyhedra than cadavers in-Nosema 29.6 (16-37) 23

LdNPV • 13.5(10-16) 22 fected with virus alone. In addition, production ofLdNPV preceding Nosema 13.7 (12-23) 22 microsporidian spores was significantly lower in cadav-

• ers infected with both pathogens (Table 3). This reduc-e Days to death (mean +_SE) was determined for larvae inoculated tion was grehtest following simultaneous infection.

with the highest dosages within each experiment.• Within an experiment, dual pathogen treatment differed signifi- The larval stage at death and dry mass of coinfected

cantly from single pathogen treatment (x2 = 7.62, P < 0.006)using cadavers were similar to those infected with virusthe Kruskal-Wallistest. alone, except when the microsporidian infection pre-

ceded viral. In the latter experiment, coinfected larvae

died at an earlier stage of development and were less•died within ea. 10 days, significantly less than the 12 than half the mass of cadavers infected with virus alone

days todeath for larvae infected with virus alone. Time (Table 3).to death for larvae infected with only the microspo-

ridium was significantly longer than for the viral or Viral infections in larvae transovarially infected withdual infections (P < 0.0001), with means ranging from Nosema. Neonate larvae, transovarially infected with

ca. 24 to 30 days (Table 2). Nosema, were ca. 10-fold more susceptible to LdNPV

Production of polyhedra was significantly affected than were microsporidia-free neonates (Table 4). Vari-when microsporidian infection preceded viral (Table 3). ability in the initial spore load of transovarially in-

TABLE 3

Distribution of Gypsy Moth Larval Instars at Death (%), Larval Cadaver Dry Mass (Mg), Infective Units/Larva (Mean _+ SESpores and Polyhedra) from the Highest Dosages a within Separate and Dual Pathogen Bioassays

Infective units/larva. Larval stage (%) Larval dry

Bioassay III IV V mass (mg) (n) Spores (n) Polyhedra (n)

Simultaneous: Nosema and LdNPV

Nosema 0 89 11 80.43 _ 6.21 (18)* 5.0 x 109 _ 5.1 x l0 s * (5) --LdNPV 0 100 0 22.81 +_1.75 (19) -- 4.8 x l0 s _ 8.7 x 107 (8)Simultane(ms 0 100 0 20.08 _+2.29 (24)* 1.2 x l0 s _ 7.5 x 107* (14) 6.2 x l0 s +_1.3 x l0 s (14)

Sequential: Nosema preceding LdNPV

N osema " 67 33 0 22.67 - 2.90 (24)* 2.3 x 109 _ 4.1 x 108 * (5)LdNPV 4 96 0 28.30 _ 2.75 (18)* -- 7.7 x l0 s +__1.5 x l0 s * (7)Sequential 83 17' 0 13.87 _ 2.17 (22)* 1.0 x 109 __3.1 x 10s * (15) 8.3 x 105 _+8.3 x 105 * (15) '

Sequential: LdNPV preceding Nosema

Nosema 0 0 100 232.22 _+14.96 (24)* 5.7 x 109 ___7.9 x 10s * (15) -- ,LdNPV 0 100 0 43.98 _ 3.51 (20) -- 1.6 x 109 +__2.1 x l0 s (5)Sequential 0 100 0 42.89 _+3.67 (20)* 2.2 x 109 +_1.2 x 109 * (12) 1.3 x 109 +_3.0 x l0 s (12)

a Results are from highest dosages (ID100)within each treatment.• Within an experiment, dual pathogen treatment differed significantly from single pathogen treatment at the P -<0.05 level of significance

using Student's t test.o

LdNPVANDNosemaSP.INTERACTIONSIN GYPSYMOTH 151

.fected larvae is likely responsible for the variability in will facilitate the natural control of gypsy moths inmortality response to viral inoculations, resulting in North America.overlapping fiducial limits. One critical consideration prior to the release of

Viral time to death in gypsy moth neonates was ca. 7 microsporidia is whether their addition might have andays, irrespective of preexisting microsporidian infec- antagonistic effect on gypsy moth's established natural

J tion. Larvae transovarially infected with microspo- enemies complex. Interactions between entomopatho-ridium; however, died in ca. 35 days as late-stage gens can be independent, antagonistic, or synergistic.larvae, almost five times later than virally infected The implications of these terminologies are quite differ-

' larvae. Unlike LdNPV which is a fatal infection, how- ent when discussing the attributes of a microbialever; over half of the larvae infected with Nosema insecticide as compared to a pathogen that was intro-survived to the adult stage (58%). duced for establishment as a biological control agent.

Most studies are designed to find the pathogen mixtureDISCUSSION. or formulation that achieves the highest mortality in

Pest management objectives for areas of North the shortest time period (for reviews see Krieg 1971,America infested or threatened with infestation by Jaques and Morris 1981). However, such a strategy isgypsymoth attempt to (1) suppress gypsy moth popula- not compatible with the inoculative release of insecttion fluctuations to minimize adverse impacts on forest pathogens introduced as biological control agents be-ecosystems, (2) reduce the nuisance of large, noxious cause periodic "extinctions" of the host may occur. Thus,caterpillars in urban forests and high-value recreation sustainability of the introduced pathogen, as well asareas,(3) slow the spread of the gypsy moth to contigu- other natura_enemies, would be periodically disrupted,ous uninfested areas, and (4) eradicate infestations in leading t*bfrequent and large-amplitude fluctuations of

the host population.noncontiguous uninfested areas. These objectives areaccomplished primarily through the introduction and Our laboratory bioassays indicate that the impact ofestablishment of gypsy moth natural enemies, inten- this Nosema sp. from Portugal on the pathogenesis of

LdNPV in the gypsy moth was variable, ranging fromsive monitoring of populations, and intermittent spray-ing with the relatively narrow spectrum microbial synergistic to antagonistic, depending on the sequenceinsecticide formulated with B. thuringiensis var. of infection. Viral infectivity, time to death, and patho-kurstaki. Because mortality caused by B. thuringiensis genesis were unaffected by simultaneous and subse-results from toxicity to several insecticidal proteins, it quent microsporidian infection. However, larvae with adoes not cycle in the field, and must be reapplied as preexisting microsporidian infection were more suscep-needed. Natural enemies, on the other hand, once tible to LdNPV, particularly at a high per os microspo-introduced from the gypsy moth's native lands, have ridian dose or when infection occurred transovarially.the potential to sustain themselves permanently within Moreover, time to death was decreased, resulting inthe host and its environment, lower polyhedron production by the virus. Varying

The complex of introduced and native natural en- degrees of competition between microsporidia and NPVemies of gypsy moth in North America, including are documented in a variety of lepidopterans (Lipa,pathogens, parasites, and predators collectively, have 1971; Nordin and Maddox, 1972; Fuxa, 1979; Cossen-often failed to maintain gypsy moth populations below tine and Lewis, 1984; Moawed et al., 1987). Antagonismthe human tolerance threshold, as evidenced by a is not unexpected as both pathogens often invade thehistory of episodic outbreaks. The gypsy moth pathogen same tissues, perhaps resulting in competition for

'complex in North America clearly lacks the diversity cellular resources or interference with cellular func-found throughout gypsy moth populations in Europe tions.and Asia. The gypsy moth virus, LdNPV, is long recog- In Europe, where this species of Nosema is endemic,nized as the primary agent responsible for collapse of gypsy moth larvae hatch in the spring with varyinghigh-density populations worldwide, including North levels of microsporidiosis from transovarial transmis-America (Elkinton and Liebhold, 1990). Since 1989, the sion (L.S.B., unpublished data). During eclosion, thesegypsy moth fungus, E. maimaiga, in North America, neonates then ingest viral polyhedra which contami-

, produced widespread epizootics, particularly during nate the egg chorion. Our results suggest that larvaewet, spring weather (Hajeket al., 1993). Conspicuously which are first infected with microsporidia and thenlacking in North American gypsy moth populations are become infected with virus will contaminate the envi-microsporidia. Unlike LdNPV and E. maimaiga, these ronment with a mixture of spores and polyhedra.entomopathogens typically cause chronic disease, result- Although we found that coinfection results in lowering in increased mortality at each life stage, prolonged polyhedron production, we also learned that larvaedevelopment, and reduced fecundity. Based on several infected with microsporidia tend to be more susceptibleEuropean studies (Zelinskaya, 1980; Novotony, 1989), to viral infection. Indeed, mixed infections common inwe feelthat microsporidia possess characteristics that larval populations of gypsy moths in Europe may

152 BAUERET AL.

depress" the density threshold required for induction of REFERENCES

a viral epizootic. Andreadis, T. G., and Weseloh, R. M. 1990. Discovery of Ento-The release of polyhedra and spore mixtures from mophaga maimaiga in North American gypsy moth, Lymantria

gypsy moth cadavers suggests that simultaneous inocu- dispar. Proc. Natl. Acad. Sci. USA 87, 2461-2465.

lation may occur through horizontal transmission in Bell, R. A., Owens, C. D., Shapiro, M., and Tardiff, J. R. 1981. Massthe field. In our laboratory study, simultaneous inocula- rearing and virus production. In "The Gypsy Moth: Research

Toward Integrated Pest Management" (C. C. Doane and M. L.tions confirmed that the rate of pathogenesis for No- McManus, Eds.), pp. 599-600. USDA, Washington. [U.S. Dept.sema waS slower than for LdNPV, minimizing the Agric.For. Serv. Tech. Bull. 1584].impact ofmicrosporidiosis on both viral infectivity and Cossentine, J. E., and Lewis, L. C. 1984. Interactions between 'lethal time. On the other hand, the rapid rate of viral Vairimorpha necatrix, Vairimorpha sp., and a nuclear polyhedrosis

pathogenesis and short lethal time antagonized the virus from Rachiplusia ou in Agrotis ipsilon larvae. J. Invertebr.

micr0sporidium life cycle by reducing the number of Pathol. 44, 28-35.spores produced. Antagonism of the microsporidium David, L., and Novotny, J. 1990. Laboratory activity and field effect ofNosema lymantriae Weiser (Microsporida) on larvae of the gypsyinfection was dose dependent and more evident when moth, Lymantria dispar L. Biologia (Bratislava)45, 73-79.viral infeetion preceded that of the microsporidium. Dwyer, G., and Elkinton, J. S. 1993. Using simple models to predictSimilar antagonism by the virus on microsporiosis is virus epizootics in gypsy moth populations. J. Anim. Ecol. 62,1-11.

also reported in another study (Lipa, 1971). Although Elkinton, J. S., and Liebhold, A. M. 1990. Population dynamics ofthe actual mechanisms of these observed pathogen gypsy moth in North America. Annu. Rev. Entomol. 35, 571-576.interactions were beyond the scope of this study, it Fuxa, J. R. 1979.Interactionsof the microsporidiumVairimorpha

Would be interesting to investigate how these very necatrix with a bacterium, virus, and fungus in Heliothis zea. J.Invertebr. Pa_hol. 33, 316-323.different pathogens utilize the resources, cellular ma-

chinery, and space within coinfected host cells. Studies Hajek, A. E., Humber, R. A., and Elkinton, J. S. 1995. Mysteriousorigins of Entomophaga maimaiga in North America. Am. Ento-using cell culture would greatly facilitate our under- mol. 41, 31-42.

standing of these "interactions and their underlying Hajek, A. E., Humber, R. A., Elkinton, J. S., May, B., Walsh, S. R. A.,mechanisms, and Silver, J. C. 1990. Allozyme and restriction fragment length

. Sustainability of effective and viable populations of polymorphism analyses confirm Entomophaga maimaiga respon-these and other host-specific natural enemies of gypsy sible for 1989 epizootics in North American gypsy moth popula-

moth depends upon the availability of an adequate tions. Proc. Natl. Acad. Sci. USA 87,6979-6982.

population of susceptible hosts. Virulent pathogens Hajek, A. E., Larkin, T. S., Carruthers, R. I., and Soper, R. S. 1993.• • Modeling the dynamics of Entomophaga maimaiga (Zygomycetes:such as :LdNPV theoretically may kill too quickly and Entomophthorales) epizootics in gypsy moth (Lepidoptera: Lyman-become too prevalent, to the detriment of other natural triidae) populations.Environ. Entomol. 22, 1172-1187.enemies. In fact, gypsy moth viral epizootics are so Jaques, R. P., and Morris, O. N. 1981. Compatibility of pathogens

devastating that few, if any, residual egg masses can be with other methods of pest control and with different crops. Infound the following year. Antagonism of virulent patho- "Microbial Control of Insects and Mites" (H. D. Burges and N. W.

•gens such as the gypsy moth virus by microsporidia Hussey, Eds.), pp. 695-715. Academic Press, NewYork.Jeffords, M. R., Maddox, J. V., and O'Hayer, K. W. 1987. Microsporid-

may dampen volatile population oscillations and pro- ian spores in gypsy moth larval silk: A possible route of horizontalvide a continuous supply of host necessary to support a transmission. J. Invertebr. Pathol. 49, 332-333.diversity of natural enemies, thus generating consis- Jeffords, M. R., Maddox, J. V., McManus, M. L., Webb, R. E., and•tent, long-term control. Indeed, microsporidia, because Wieber, A. 1988. Egg contamination as a method for the inoculative• . .

of their chronic nature, high infectivity, and efficient release of exotic microsporidia of the gypsy moth. J. Invertebr.

vertical transmission, are one of the few host-specific Pathol. 51, 190-196.natural enemies detectable when host population den- Jeffords, M. R., Maddox, J. V., McManus, M. L., Webb, R. E., and

Wieber, A. 1989. Evaluation of the overwintering success of twosities are low (Onstad and Maddox, 1989; Onstad et al., European microsporidia inoculatively released into gypsy moth1990). A greater diversity of natural enemies that populations in Maryland. J. Invertebr.Pathol.53, 235-240.

includes microsporidia may reduce the frequency, am- Krieg, A. 1971. Interactions between pathogens. In "Microbial Con-plitude, and duration of outbreaks, thereby maintain- trol of Insects and Mites" (H. D. Burges and N. W.Hussey, Eds.), pp.ing gypsy moth populations at endemic levels. Such 459-468.Academic Press, NewYork.

permanent and self-sustaining control would greatly LeOra Software. 1987. "POLO-PC. A User's Guide to Probit or LogitAnalysis." LeOra, Berkeley, CA.minimize the combined damage caused by both gypsy

moth and human interventions on the biological diver- Lewis, F. B. 1981. Gypsy moth nucleopolyhedrosis virus. In "TheGypsy Moth: Research Toward Integrated Pest Management"

sity and health of NorthAmerican forests. (c. c. Doane and M. L. McManus, Eds.), pp.454-455. USDA,Washington. [U.S. Dept. Agric. For. Serv. Tech. Bull. 1584]

Lipa, J. J. 1971. Studies on the interactions of various pathogens inACKNOWLEDGMENTS one insect host (cutworms). Institute of Plant Protection, Poznan,

We thank G. Heaton, J. Rae, and R. Oyer for laboratory assistance Poland. [Final Rep. E21-Ent-17, FG-PO-194]during this project, and Drs. R. A. Haack, C. N. Koller, and D.W. McManus, M. L., Maddox, J. V., Jeffords, M. R., and Webb, R. E. 1989.Onstad for their critical reviews of the manuscript. Evaluation and selection of candidate European microsporidia for

LdNPV AND Nosema SP. INTERACTIONS IN GYPSY MOTH 153

introduction into U.S. gypsy moth populations. In "The Lymantri- Onstad, D. W., Reissig, W. H., and Shoemaker, C.A. 1986. Influence ofidae: A Comparison of Features of New and Old World Tussock apple cultivar, tree phenology, and leaf quality on the developmentMoths" (W.E. Wallner, Ed.), pp. 455-468. USDA, Washington. [U.S. and mortality of Choristoneura rosaceana (Lepidoptera: Tortric-Dept. Agri. For..Serv. Gen. Tech. Rep. NE-123]. idae). Can. Entomol. 118, 123-132.

Moawed, S. M., Saleh, M. R., and Salem, S.A. 1987. Interaction ofthe Pilarski, D., and Vavra, J. 1991. Morphology and development ofmicr0sporidium Vairimorpha ephestiae with nuclear polyhedrosis Nosema serbica Weiser, 1963 (Microsporida, Nosematidae), para-virus and their effect on Egyptian cotton leafworm, Spodoptera site of the gypsy moth Lymantria dispar (Lepidoptera, Lymantri-litto ralis. Acta Ento mol. Bohemoslov. 84, 422-430. idae). Folia Parasitol. 88, 115-121.

Nordin, G. L., and Maddox, J. V. 1972. Effect of simultaneous virus Reardon, R. C. 1981. Alternative controls. Parasites. In "The Gypsyand microsporidian infection on larvae of Hyphantria cunea. J. Moth: Research Toward Integrated Pest Management" (C. C.Invertebr. Pathol. 20, 66-69. Doane and M. L. McManus, Eds.), pp. 299-421. USDA, Washing-

" ton. [U.S. Dept. Agric. For. Serv. Tech. Bull. 1584]

Novotony, J. 1988. The use of nucleopolyhedrosis virus (NPV) and SAS. 1990. "Procedures Guide Version 6." SAS Institute, Cary, NC.microsporidia in the control of the gypsy moth. Folia Parasit. 35,.199-208. Solter, L. F., Maddox, J. V., and McManus, M. L. 1997. Host specificity

of microsporidia (Protista: Microspora) from European populationsNoy0tony, J. 1989. Natural disease of gyps3}moth in various grada- of Lymantria dispar (Lepidoptera: Lymantriidae) to indigenous

tion phases: In "The Lymantriidae: A Comparison of Features of North American Lepidoptera. J. Invertebr. Pathol. 69, 135-150.

New and Old World Tussock Moths" (W. E. Wallner, Ed.), pp. Speare, A. T., and Colley, R. H. 1912. "The Artificial Use of the101-_Ill. Washington, DC. '[U.S. Dept. Agri, For. Serv. Gen. Tech. Brown-Tail Fungus in Massachusetts, with Practical SuggestionsRep. NE-123] for Private Experiment, and a Brief Note on a Fungous Disease of

Novotony; J. 1991. Influence of gypsy moth larvae age on the the GypsyCaterpillar."Wright& Potter, Boston, MA.

development- of microsporidian diseases. Biologia (Bratislava) 46, Undeen, A. H., and Alger, N. E. 1971. A density gradient method for105-112. fractionating microsporidian spores. J. Invertebr. Pathol. 18, 419-

Novotony, J., and Weiser, J. 1993. Transovarial transmission of 420.Nosema lymantriae (Protozoa: Microsporidia) in the gypsy moth Weiser, J. 1987. Patterns over place and time. In "Epizootiology ofLymantriadispar L. Biologia (Bratislava)48, 125-129. Insect Diseases" (J. R. Fuxa and Y. Tanada, Eds.), p. 215-242.

Onstad, D. W., and Maddox, J. V. 1989. Modelling the effects of the Wiley, New York.microsporidium, Nosemapyrausta, on the population dynamics of Weiser, J., and Novotny, J. 1987. Field application of Nosemathe insect, Ostrinia nubilalis. J. Invertebr. Pathol. 53, 410-421. lymantriae against the gypsy moth, Lymantria dispar L. J. Appl.

Onstad, D. W., Maddox, J. V., Cox, D. J., and Kornkven, E. A. 1990. Entomol. 104, 58-62.Spatial, and temporal dynamics of animals and the host-density Wilkinson, L. 1987. "SYSTAT: The System for Statistics." SYSTAT,threshold in epizootiology. J. Invertebr. Pathol. 55, 76-84. Evanston, Illinois.

Onstad, D. W., Maddox, J. V., Jeffords, M. R., McManus, M. L., Bauer, Woods, S. A., and Elkinton, J. S. 1987. Bimodal patterns of mortalityL. S., Webb, R. E. 1993. Modeling horizontal transmission of from nuclear polyhedrosis virus in gypsy moth (Lymantria dispar)microsporidia in gypsy moth populations. In "Proceedings: Inter- populations. J. Invertebr. Pathol. 50, 151-157.agency Gypsy Moth Research Forum," Jan. 19-22, 1993, Annapo- Zelinskaya, L. M. 1980. Role ofmicrosporidia in population dynamicslis, MD, p. 87. USDA, Washington. [U.S. Dept. Agri. For. Serv. Gen. of the gypsy moth in the forests of the lower Dniepr Valley. Vest.Tech. Rep. NE-179] Zool. 1, 57-62.

6