Changes in the fecundity of tent caterpillars: a correlated character of disease resistance or sublethal effect of disease?

Transcription

1 ecologia (1995) 3: Springer-Verlag 1995 Judith H. Myers 9 Barbara Kuken Changes in the fecundity of tent caterpillars: a correlated character of disease resistance or sublethal effect of disease? Received: 4 April 1994 / Accepted: 1 April 1995 Abstract ver the fluctuation in population density of tent caterpillars, Malacosoma californicum pluviale and M. disstria, fecundity changes from being high at peak density to tow for several years during the decline. During the increase phase, fecundity rapidly returns to moderately high levels with a further increase occurring toward the end of the increase phase. Two hypotheses which might explain these shifts are that (1) mortality from viral disease which is common during population declines selects for resistant individuals with low fecundity as an associated characteristic, and (2) sublethal viral disease reduces fecundity of moths during population decline. In this study we observed rapid shifts in the frequencies of large and small egg masses and in the mean fecundity between different phases of the population fluctuation. Viral disease was more common in caterpillars from small egg masses of the forest tent caterpillar. These observations are consistent with the hypothesis that sublethal effects of virus reduce the fecundity of moths during the population decline, but high fecundity is quickly restored when disease is rare during the population increase. Key words Population cycles 9 Forest lepidoptera - Nuclear polyhedral virus 9 Fecundity 9 Sublethal Introduction Much work on population cycles of forest lepidoptera has focussed on mortality (review in Myers 1988). ften life-table studies consider the fecundity of insects to be constant from year to year because counting eggs is tedious or impossible. Tent caterpillars are particularly good for monitoring shifts in fecundity because their egg masses remain on trees for over a year, and can be collected and counted after hatch. In long term studies of J.H. Myers (~). B. Kukan Centre for Biodiversity Research, Departments of oology and Plant Science, University of British Columbia, Vancouver, Canada, V6T I4 western tent caterpillars, Malacosoma californicum pluviale, we have found a consistent pattern of change in fecundity; females lay the largest egg masses in the year of peak density. The fecundity of moths is then reduced for several generations during the population decline (Myers 1988, 1990, 1993). thers have observed similar patterns of change in fecundity with change in population density of tent caterpillars. In earlier studies of western tent caterpillars in British Columbia, Wellington (1960) found numerous large egg masses at the beginning of the year of peak density and initial decline. Similarly Blais et al. (1955) recorded large egg masses at peak densities of the forest tent caterpillar, M. disstria, in ntario. In Minnesota, fecundity of forest tent caterpillar was the highest during the last two years of population increase (Witter et al. 1975). In the summer of peak caterpillar density, egg masses were numerous (400, ,000 per ha) and large (means of eggs per mass for high-density plots). At peak density, survival of the early larval stages was good while that of late instars was poor. Moths emerging at the end of the summer of peak density had lower fecundity (means of eggs per female among plots), and the populations continued to decline in the next two generations. This observed pattern of highest fecundity at peak population density, and reduced fecundity for several generations into the decline could explain the cyclic population dynamics of these forest Lepidoptera (Myers 1988; Ginzburg and Taneyhill 1994). Therefore, explaining the causes of reduced fecundity during the population decline is a key challenge for fully understanding population cycles. ne of the characteristics associated with the decline of tent caterpillar populations is the widespread occurrence of nuclear polyhedral viruses (NPV). Disease could explain several of the characteristics of population cycles of tent caterpillars. For example, in the 1-st year of the population decline the survival of late instar larvae is poor, while in the 2-nd-year survival of earlier instars is reduced (Myers 1990). This is consistent with the ac-

2 476 ECLGIA 3 (1995) 9 Springer-Verlag cumulation of polyhedra in the environment in the first year causing infection of early-instar caterpillars the next year (Woods and Elkinton 1987). Disease might also be involved with shifts in fecundity of moth populations. Myers (1990) developed the,,disease defense hypothesis" as a possible explanation of the reduction of fecundity during the population decline of tent caterpillars. Mortality from NPV during the population decline might select against susceptible and for more resistant individuals (Martignioni 1957). If reduced fecundity were a correlated characteristic of resistance to disease (Fuxa and Richter 1990), fecundity might be expected to decline in populations after exposure to high mortality from disease. Relaxed selection for resistance in the absence of disease could increase the frequency of susceptible, high-fecundity phenotypes during the phase of population increase. Under this scenerio a rapid reduction in fecundity might be expected from,,hard" selection associated with high mortality from virus and a more gradual increase in fecundity with the,,soft" selection occurring as the population increases in the near absence of virus. An alternative hypothesis is that reduced fecundity during the population decline is associated with sublethal infection by virus. Laboratory (Rothman and Myers 1994) and field studies (L. Rothman, unpublished work) indicate that exposure to NPV can reduce the fecundity of western tent caterpillars. It has thus far proven impossible to test the disease resistance hypothesis for cyclic forest insects directly by challenging caterpillars from populations in different phases with virus, because few or no egg masses can be found before hatch at low density. However, several predictions of the disease resistance hypothesis and of the hypothesis that sublethal infection reduces fecundity can be tested. In this paper we consider three issues: (1) the distribution of egg mass sizes in populations of tent caterpillars at different stages of the population cycle, (2) the rate of change in the fecundity of moths over the increase phase of the cycle, and (3) the occurrence of NPV in caterpillars hatching from small and large egg masses. If the variation in fecundity were genetically determined, selection from higher fecundity (soft selection) following the reduction in viral disease should cause a gradual shift in the distribution of egg mass sizes between generations. A second prediction of the disease defense hypothesis is that caterpillars from small egg masses should be more resistant to disease, and therefore, less likely to die of virus. If small egg masses are from females who were sublethally infected, virus might be expected to be more prevalent among caterpillars from small egg masses (Shapiro and Robertson 1987). Methods and results Population trends Because western tent caterpillars form obvious tents, populations can be counted even when densities are low. P. m MANDARTE WESTHAM j.... i.... i Fig. 1 Number of western tent caterpillar tents counted on Mandarte and Westham Islands We have been counting tents and collecting egg masses from consistent areas at several locations in the lower mainland of British Columbia Canada since 1975 (Myers 1990, 1993). Since females lay all of their eggs in a single mass, egg mass size is equivalent to fecundity. Populations were at peak density in 1976 and in Following the second peak they declined until , and then began to increase. Population trends for two sites, Mandarte and Westham, have diverged after the recent decline with the density at Mandarte continuing to increase while that at Westham has declined after an initial increase (Fig. 1). Distribution of egg mass sizes A dimorphism of low and high fecundity phenotypes might occur if sublethally infected individuals produced fewer eggs than uninfected individuals. If fecundity were genetically determined primarily by a single locus with a dominant allele, a dimorphism with large and small egg masses might also exist. However, if fecundity were inherited as a quantitative trait with multiple loci, a broader range of egg mass sizes might be expected with an approximately normal distribution. Selection could shift the mean fecundity. Environmental conditions such as food limitation or deterioration might influence the fe-

3 Fig. 2 Distributions of egg mass sizes for four generations of tent caterpillars on Mandarte Island. Egg masses are combined into size categories starting with 139 or fewer eggs/mass, and increasing over intervals of 20 eggs. Numbers in each category have been converted to proportions for comparison. Year, mean and sample size are given on each graph. Distributions are of egg masses at the beginning of the year indicated and will have been produced by moths in the previous summer. Density trend of the population is shown in Fig. 1 (N number, X mean) 2 ~e el. N LATE INCREASE ECLGIA 3 (1995) 9 Springer-Verlag N=26 X = EGG MASS SIE CATEGRY BEGINNING 1987 DECLINE N = o.2 o PEAK N=85 Y = 91.q EGG MASS SIE CATEGRY :LINE 1988 N=11 X = ~e o.1 ll EGG MASS SIE CATEGRY EGG MASS SIE CATEGRY Fig. 3 Distributions of egg mass sizes for four generations of tent caterpillars on Westham Island. (See caption of Fig, 2) INCREASE 1985 N=17 X = EARLY PEAK 1986 N=52 X = 224,,e o.2 o.1 [~ el_ ~. o.1 o,, EGG MASS SIE CATEGRY EGG MASS SIE CATEGRY el b-~ g~ g~ LW DENSITY 1989 N= X=173 EGG MASS SIE CATEGRY EGG MASS SIE CATEGRY

4 478 ECLGIA 3 (1995) 9 Springer-Verlag t13.< ~ X < GALIAN 000 T 3_ T 00 T 9 T ~ -J.- " '3 b., -b- r.t3 r.t CYPRESS T I 1 '86 1 ' ' l b- [- b- r13 ~t3 N WESTHAM ;7 19; rar~ b" ~rd b" Fig. 4 Changes in number of tents (line and open circles) and mean (_+2 SE) egg mass sizes (filled squares) for four populations of western tent caterpillar. To assist visual comparison, horizontal lines indicate 200 eggs per mass on each graph. Study sites, all in south-western British Columbia, are described in Myers (1990) cundity of moths, and cause high densities of moths to have smaller egg masses than low densities. The observed distributions of egg mass sizes at different phases of the population fluctuation demonstrate the shifts in fecundity. Distributions of egg mass sizes in two populations of western tent caterpillar are shown in Fig. 2 and 3. In the year of late population increase, 1985, a range in sizes of egg masses occur with a hint of a bimodal distribution eg. a dimorphism of fecundities. At peak population density, 1986, most egg masses have more than 200 eggs. At the beginning of the generation in 1987 most egg masses are small (less than 200 eggs) at Mandarte, but are still large at Westham. At both sites the egg masses remained small with continued population decline or low numbers. The distributions of egg mass sizes do not resemble changes in a normal distribution, but are more compatable with a dimorphism with a predominance of large egg masses at high caterpillar density, a predominance of small egg masses at low density, and a combination of large and small egg masses with fewer intermediate sizes in the late phase of population increase. ;t3 t~.r MANDARTE TTT ' Rate of change in average egg mass size Population trends and mean egg mass sizes for four populations of tent caterpillars from late increase through decline and early recovery are shown in Fig. 4. Mean egg mass sizes are those at the beginning of the generation. In all populations the average of the egg masses declined by eggs following the population peak in 1986 or With the beginning of the population increase several generations later, the means of egg masses increased by eggs in a single year. In three populations egg masses continued to increase for several generations while at Mandarte the size of the egg masses stabilized after the initial increase. After the initial recovery, egg masses remained similar in size for several generations. The rapid shift in fecundity observed between peak to decline and back to the increase phase of the western tent caterpillar is probably too rapid to be explained by a genetic shift. However, without understanding the genetic basis and heritabilities underlying fecundity and without measurements of selection coefficients, further evaluation of the observed shifts in fecundity is not possible. Patterns are compatible with expectations of sublethal viral infection reducing the fecundity of some individuals. or~ [- r.t3

5 ECLGIA 3 (/995) 9 Springer-Verlag 479 ccurrence of disease in caterpillars from large and small egg masses If the reduced fecundity of tent caterpillars in declining populations were associated with a sublethal infection and associated vertical transmission, small egg masses might be more likely to be contaminated with disease. If on the other hand, low fecundity were a correlated character of disease resistance (Fuxa and Richter 1990), viral infection might be expected to be lower among caterpillars arising from small egg masses. In 1989 we obtained 24 forest tent caterpillar egg masses (size range eggs, median 140 eggs) from a population near Edmonton, Alberta that was in the late peak phase of the population fluctuation. Without any disinfection, we reared family groups of caterpillars from these egg masses in the laboratory, in glass jars, on red alder leaves (Alnus rubra). The leaves were washed with chlorox and rinsed before use, and were changed at least every other day. Dead caterpillars were smeared on glass slides, and examined microscopically for the presence of polyhedral inclusion bodies indicating infection with nuclear polyhedral virus. f these 24 families of caterpillars, viral deaths occurred in 13. verall 72% of 11 families originating from egg masses with less than 140 eggs had virus while only 39% of 13 families from egg masses with more than 140 eggs had virus. This difference was not significant by X 2 analysis, but the trend does not support the hypothesis that caterpillars from small egg masses are more resistant to viral disease. The mean percentage of caterpillars dying of virus in families from small egg masses, 30%, was significantly higher than that from large egg masses, % (Mann-Whitney U-test, P<0,05). In seven of eight families infected with virus arising from small egg masses, 25% or more individuals died of virus. For large egg masses only two families had more than 25% die of virus. These results are consistent with the interpretation that viral contamination is more strongly associated with small egg masses. Discussion Fecundity of forest Lepidoptera often changes with density, and reduced fecundity is often observed in crowded and declining populations (Miller 1963; Baltensweiler 1968; Gruys 1970; Mason et al. 1977). The surprising aspect of the change in fecundity of tent caterpillars is that high fecundity is associated with peak density, when it might be expected that crowding or food limitation would reduce fecundity. It is possible that a lag in the impact of high density results in an initial increase in fecundity (Rossiter 1991) and an additional lag keeps the fecundity low for several generations in the decline. This analysis of the distributions of egg mass sizes over a population peak of tent caterpillars shows rapid changes in mean fecundity that are unlikely to be totally explained by genetic selection. While mortality from NPV could in theory select for increased disease resistance (Watanabe 1987), it is not clear how disease resistance and fecundity might be linked. Fuxa and Richter (1990) found reduced fecundity in a strain of Spodoptera frugiperda selected for resistance to NPV. However, Boots and Begon (1993) found that a laboratory population of the grain moth, Plodia interpunctella, exposed to granulosis virus evolved resistance with significant increase in size as a correlate of resistance. Vail and Tebbets (1990), in a comparative study of six populations of P. interpunctella observed that the two populations with the greatest resistance also had the lowest number of eggs per unit body weight. The number of eggs per female in these two resistant strains was very different and showed no correspondence with resistance. Therefore, no clear relationship exists between disease resistance and fecundity among Lepidpotera. Distributions do not expose the underlying mechanism(s) that might cause high and/or low fecundity individuals in a population. A number of mechanisms can influence the fecundity of Lepidoptera. Rossiter (1991) studied the effects of maternal diet on offspring size in gypsy moth (Lymantria dispar). She found that feeding larvae on leaves from defoliated trees reduced the size of resulting moths. In the next generation, however, offspring of females challenged by poor food quality produced larger egg masses than offspring of control females. This delayed maternal effect might explain the jump in egg mass size observed at the end of the population increase of tent caterpillars. Alternatively, moderate densities of caterpillars might stimulate improved foliage quality (Williams and Myers 1984; Roland and Myers 1987) and higher fecundities. Heterogeneity in food quality (Haukioja 1993), presence or absence of microparasites (disease), temperature during larval development (Morris and Fulton 1970), or genetic variation, could all influence fecundity, and the distribution of egg mass size. Ginzburg and Taneyhill (1994) developed models to consider the potential impact of maternal influences on cyclic population dynamics of forest Lepidoptera, but changes in fecundity caused by other mechanisms could have similar impacts on population dynamics as those attributed to maternal influences. The emphasis on the study of disease, and particularly viral disease, in forest Lepidoptera often has been on its impact in causing the population crash, or by the desire to develop insect control through a microbial insecticide (Magnoler 1975). We feel that sublethal impacts of disease deserve more attention as a mechanism for maintaining disease in low density populations, and as an influence on the dynamics of disease transmission to subsequent generations. If caterpillars become infected shortly before pupation, infection could reduce the resources available for egg production and cause low fecundity. Disease might also be transmitted on the eggs of sublethally infected females. Laboratory studies of granulosis virus and P. interpunctelia show sublethal effects on larval development time, egg production and egg via-

6 480 ECLGIA 3 (1995) 9 Springer-Verlag bility (Sait et al. 1994). Similar relationships might exist for other lepidopterans as well. Reduction of disease in low density populations could result in a rapid switch to high fecundity phenotypes after the population decline, and build up in disease could cause a rapid shift to low fecundity phenotypes in the intitial stage of population decline. The existence of sublethally infected individuals and uninfected individuals late in the phase of population increase could be expressed as a bimodal distribution of fecundities as suggested by the data presented here. Determination of the prevalence of virus in caterpillars from small and large egg masses will allow further evaluation of this proposed interaction, and will be the subject of future work. Acknowledgements We wish to thank the many people involved in counting eggs and rearing caterpillars, and Lorne Rothman, Joe Elkinton, and Jamie Smith for comments on the manuscript. Jens Roland kindly collected eggs from forest tent caterpillars. This work was supported by a grant from the Natural Science and Engineering Research Council of Canada. J.H.M. appreciates the hospitality of the Dept. Entomology, University of Massachusetts where this paper was written. References Baltensweiler W (1968) The cyclic population dynamics of the grey larch tortrix, eiraphera griseana Hubner (=Semasia diniana Guenee) (Lepidoptera, Tortricidae). In: Southwood TRE (ed) Insect abundance. Symp Entomol Soc Lond 4:88-98 Blais JR, Prentice WL, Sippell WL, Wallace DR (1955) Effects of weather on the forest tent caterpillar, Malacosoma disstria Hbn., in central Canada in the spring of Can Entomol 87:1-8 Boots MR, Begon M (1993) Trade-offs with resistence to a granulosis virus in the Indian meal moth, examined by a laboratory evolution experiment. Funct Ecology 7: Fuxa J, Richter AR (1990) Reversion of resistance by Spodoptera frugiperda to nuclear polyhedrosis virus. J Invert Pathol 53: Ginzberg LR, Taneyhill DE (1994) Population cycles of forest Lepidoptera: a maternal effect hypothesis. J Anita Ecol 63:79-92 Gruys P (1970) Growth in Bupalus piniarius (Lepidoptera, Geometridae) in relation to larval population density. Verh Rijksinst Natuurbeheer 1:1-127 Haukioja E (1993) Effects of food and predation on population dynamics. In: Stamp NE, Casey TM (eds) Caterpillars: ecological and evolutionary constraints on foraging. Chapman and Hall, New York, pp Magnoler A (1975) Bioassay of nucleopolyhedrosis virus against larval instars of Malacosoma n eustria. J Invert Pathol 25: Martignioni ME (1957) Contributo allconoscenza di una granulosi de Eucosma griseana (Hubner) quale fattore limitante il pullulamento dell'insectto nella Engadine alta. Mitt Schweiz Anst Forsch Verswes 32: Mason RR, Beckwith RC, Paul HG (1977) Fecundity reduction during collapse of a Douglas-fir tussock moth (Lepidoptera: Lymantriidae) population in the Sierra Nevada. Environ Entomol 12: Miller CA (1963) The analysis of fecundity proportion in the unsprayed area. In: Morris RF (ed) The dynamics of epidemic spruce budworm populations. Mem Entomol Soc Can 31:75-86 Morris RE Fulton WC (1970) Models for the development and survival of Hyphantria cunea in relationship to temperature and humidity. Mere Entomol Soc Can 70:1-66 Myers JH (1988) Can a general hypothesis explain population cycles of forest Lepidoptera? Adv Ecol Res 18: Myers JH (1990) Population cycles of western tent caterpillars: experimental introductions and synchrony of fluctuations. Ecology 71: Myers JH (1993) Population outbreaks in forest Lepidoptera. Am Sci 81: Roland J, Myers JH (1987) Improved insect performance from host-plant defoliation: winter moth on oak and apple. Ecol Entomol 12: Rossiter MC (1991) Environmentally-based maternal effects: a hidden force in insect population dynamics? ecologia 87: Rothman LD, Myers JH (1994) Effects of a nuclear polyhedrosis virus (NPV) on the reproductive potential of the western tent caterpillar. Environ Entomol 23: Sait SM, Begon M, Thompson DJ (1994) Long-term population dynamics of the Indian meal moth Plodia interpunctella and its granulosis virus. J Anim Ecol 63: Shapiro M and Robertson JL (1987) Yield and activity of gypsy moth (Lepidoptera: Lymantriidae) nucleopolyhedrosis virus recovered from survivors of viral challenge. J Econ Entomol 80: Vail PV, Hall IM (1969) Susceptibility of the pupa of the cabbage looper, Trichoplusia hi, to nucleopolyhedrosis virus. J Invert Pathol 14: Watanabe H (1987) The host population. In: Fuxa JR, Tanada Y (eds) Epizootiology of insect diseases. Wiley, New York, pp Wellington WG (1960) Qualitative changes in natural populations during changes in abundance. Can J ool 38: Williams KS, Myers JH (1984) Previous herbivore attack of red alder may improve food quality for fall webworm larvae. ecologia 63: Witter JA, Mattson WJ, Kulman HM (1975) Numerical analysis of a forest tent caterpillar (Lepidoptera: Lasiocampidae) outbreak in northern Minnesota. Can Entomol 7: Woods SA, Elkinton JS (1987) Bimodal patterns of mortality from nuclear polyhedrosis virus in gypsy moth (Lymantria dispar) populations. J Invert Pathol 50: