Density dependence and the stabilization of animal numbers

Summary The claim of Varley and Gradwell that the highly density-dependent pupal predation (k ₅) in the population of the winter moth in Wytham Wood, England would keep density within limits (regulate) is compared in this paper with the density limits in the null model: pupal predation causes the same mean generation mortality (35%) as in the field, but is not density-dependent, i.e. either constant or randomly fluctuating between years according to the actual frequency distribution, ceteris paribus. According to this null model the winter moth would have fluctuated between narrower limits than in the field; the claim of Varley & Gradwell must thus be rejected. It is more generally concluded that a regulating factor should be the key factor, but this is not a sufficient condition. It should also prevent in some way the low throughs in the time series that usually accompany the operation of a dominating density-dependent mortality factor.Communication No. 264 of the Biological Station, Wijster     

Escape from pupal predation as a potential cause of outbreaks of the winter moth, Operophtera brumata

The winter moth, Operophtera brumata , shows varying population dynamics in different host plant habitats. Populations in Sitka spruce, Picea sitchensis, plantations and in Scottish moorlands have a tendency to outbreak that is not shown by winter moth in lowland oak woods. Since pupal predators have previously been identified as being important for the regulation of winter moth in a lowland oak wood it was hypothesized that invertebrate pupal predators were failing to control winter moth in outbreak populations. This hypothesis was tested by comparing the abundance of invertebrate predators and patterns of spatially density dependent pupal predation across habitats. Several results supported this hypothesis. Carabid predators of winter moth were one or two orders of magnitude more abundant in oak woods than in moorland or spruce habitats. Staphylinid predators were also more abundant in high winter moth density oak woods than in any other habitat. Beetle predation of tagged cocoons in the field was inversely density dependent in Highland moors in experiments in 1999 and 2000, and in Sitka spruce in 1999. However, in opposition to our hypothesis, pupal predation was also inversely density dependent in oak woods in 2000, although this result may be explained by the low range of winter moth densities in the field that year. These results are discussed in relation to the role of natural enemies in regulating winter moth populations and the differences in life-history of the beetle predators in different habitats.     

Parasitism and predation as agents of mortality of winter moth populations in neglected apple orchards in Nova Scotia

Abstract.

• 1 This study compared the roles of pupal mortality and parasitism in winter moth (Operophtera brumata) population dynamics in Nova Scotian apple orchards and assessed the importance of beetles as pupal predators.
• 2 The component of pupal mortality termed predation accounted for greater stage-specific mortality of winter moth than parasitism by Cyzenis albicans in four neglected orchards.
• 3 Parasitism by Cyzenis albicans was not spatially density-dependent in any orchard, whereas the predation component of pupal mortality was spatially density-dependent in the two orchards most densely populated by winter moth.
• 4 Field experiments indicated that over 60% of pupal predation may be attributed to beetles, and that about 46% of pupal predation occurred within 4 weeks after pupal drop.
• 5 Mortality of introduced populations of winter moth in Nova Scotia resembles that of native populations in England where density-dependent predation regulates the winter moth population and reduces the parasitoid population to minimal levels. The situation in Nova Scotia appears to have changed appreciably since the establishment of parasitoids into the system in the 1950s.

     

Delayed density-dependent parasitism of eggs and pupae as a contributor to the cyclic population dynamics of the autumnal moth

Many populations of forest Lepidoptera exhibit 10-year cycles in densities, with impressive outbreaks across large regions. Delayed density-dependent interactions with natural enemies are recognized as key factors driving these cyclic population dynamics, but emphasis has typically been on the larval stages. Eggs, pupae and adults also suffer mortality from predators, parasitoids and pathogens, but little is known about possible density relationships between mortality factors and these non-feeding life stages. In a long-term field study, we experimentally deployed autumnal moth (Epirrita autumnata) eggs and pupae to their natural enemies yearly throughout the 10-year population cycle in northern Norway. The abundance of another geometrid, the winter moth (Operophtera brumata), increased in the study area, permitting comparisons between the two moth species in predation and parasitism. Survival of autumnal moth eggs and pupae was related to the moth abundance in an inverse and delayed manner. Egg and pupal parasitoids dominated as density-dependent mortality factors and predicted the subsequent growth rate of the host population size. In contrast, effects of egg and pupal predators were weakly density dependent, and generally predation remained low. Parasitism rates did not differ between the autumnal and winter moth pupae, whereas predators preferred winter moth pupae over those of the autumnal moth. We conclude that parasitism of the autumnal moth by egg and pupal parasitoids can be related to the changes of the moth density in a delayed density-dependent manner. Furthermore, egg and pupal parasitoids cannot be overlooked as causal factors for the population cycles of forest Lepidoptera in general.     

Predation of codling moth Cydia pomonella L. by the Silvereye Zosterops lateralis (Latham)

Predation of cocooned larvae of codling moth Cydia pomonella by silvereyes Zosterops lateralis was studied in an apple orchard in Nelson, New Zealand. Apple logs with known larval densities were made available to the birds for known periods of time, either in cages or exposed in the apple orchard. The numbers of silvereyes and the natural predation of codling moth were recorded in the same orchard. Predation was density dependent. On caged logs with an initial high density of 32 larvae, 1.1 larvae were consumed per bird‐hour; in contrast, one larva was consumed per 34.5 bird‐hours at three larvae per log. A curvilinear relationship was demonstrated between larval density and the bird‐hours required for predation; this relationship was consistent with the known density dependence of silvereye predation of codling moth. A regression of the total annual winter bird predation of larvae in the orchard on bird numbers was significant. However, the density dependence of predation resulted in declining rates of predation over the winter as larval density declined; the first birds to arrive in the orchard benefitted from particularly high predation rates. As a consequence, fluctuations in bird numbers during the winter had only a secondary influence on predation rates. The numbers of silvereyes in the orchard showed no relationship to the density of the codling moth population present. This study confirmed the importance of silvereyes in the predation of codling moth and a functional, not numerical, rseponse of these birds to codling moth density.     

Interactions between Operophtera brumata and Tortrix viridana on oak: new evidence from time-series analysis

1. Yearly population estimates of the green oak leaf roller moth, Tortrix viridana (Lepidoptera: Tortricidae), and the winter moth, Operophtera brumata (Lepidoptera: Geometridae), were taken from the pedunculate oak, Quercus robur, from 1951 to 1966 in Wytham Woods, Oxfordshire, U.K. Larval mortality from parasitoids was recorded for both species over the same period. 2. Operophtera brumata population density fluctuated around a constant mean, whereas T. viridana population density exhibited a linear decline over the sampling period. Population counts were subjected to time-series analysis after the linear decline was removed from the T. viridana data by detrending. Multiple regression models were built so that variation in the per capita rate of increase of each population could be partitioned among (a) current and previous population density, (b) current and previous population density of the second moth species, and (c) rates of larval parasitism. 3. Multiple regression analysis suggested that variation in the per capita rate of increase of O. brumata could be explained by negative feedback from O. brumata density at T–2 (32%), a negative relationship with T. viridana density at T–1 (18%), and a positive relationship with parasitism suffered by T. viridana at T–1 (24%). 4. The T. viridana population time-series was dominated by a rapid feedback process such that per capita rate of increase at time T was negatively related to population density at time T–1, explaining 53% of variance in population growth rate. Per capita rate of increase was unrelated to larval parasitism or densities of O. brumata. 5. In light of previous life-table and experimental studies on O. brumata and T. viridana, the current analyses suggest that pupal predation (data not presented) and interspecific competition are significant determinants of O. brumata population growth rates in Wytham Woods. In contrast, T. viridana population growth rates appear to be dominated by rapid negative feedback consistent with intraspecific competition.     

Responses of generalist invertebrate predators to pupal densities of autumnal and winter moths under field conditions

• 1 Generalist natural enemies are usually not considered as being capable of causing population cycles in forest insects, but they may influence the population dynamics of their prey in the low density cycle phase when specialist enemies are largely absent.
• 2 In the present field study, the total response of the generalist invertebrate predator community to experimentally established pupal densities of the closely related autumnal (Epirrita autumnata) and winter moths (Operophtera brumata) was analysed.
• 3 Due to the high amount of variation in the dataset, the exact shape of the response curve could not be convincingly estimated. Nevertheless, two important conclusions can be drawn from the analyses.
• 4 Firstly, the natural invertebrate predator community seems to become saturated at rather low densities of both autumnal and winter moth pupae. Secondly, the predator community seems to become saturated at much lower densities of autumnal than of winter moth pupae.
• 5 Furthermore, pupal mass was significantly negatively correlated with invertebrate predation probability in autumnal moth pupae.
• 6 These results indicate that differences in the predator assemblage being able to consume pupae of the two moth species, as well as different handling times, could be responsible for the substantially higher predation rates in winter than in autumnal moth pupae.
• 7 As a consequence, the population dynamics of autumnal moths might be less affected by generalist invertebrate predators than those of winter moths, as autumnal moths seem able to escape from the regulating influence of generalist predators at much lower population densities than winter moths.

     

Local outbreaks of Operophtera brumata and Operophtera fagata cannot be explained by low vulnerability to pupal predation

• 1 One of the unresolved questions in studies on population dynamics of forest Lepidoptera is why some populations at times reach outbreak densities, whereas others never do. Resolving this question is especially challenging if populations of the same species in different areas or of closely‐related species in the same area are considered.
• 2 The present study focused on three closely‐related geometrid moth species, autumnal Epirrita autumnata, winter Operophtera brumata and northern winter moths Operophtera fagata, in southern Finland. There, winter and northern winter moth populations can reach outbreak densities, whereas autumnal moth densities stay relatively low.
• 3 We tested the hypothesis that a lower vulnerability to pupal predation may explain the observed differences in population dynamics. The results obtained do not support this hypothesis because pupal predation probabilities were not significantly different between the two genera within or without the Operophtera outbreak area or in years with or without a current Operophtera outbreak.
• 4 Overall, pupal predation was even higher in winter and northern winter moths than in autumnal moths. Differences in larval predation and parasitism, as well as in the reproductive capacities of the species, might be other candidates.

     

Key-factor analysis of immature stages of Aedes scapularis (Diptera: Culicidae) populations in southeastern Brazil

Development times and mortality factors were investigated for populations of Aedes scapularis (Rondani) occurring in natural temporary ground pools in southeastern Brazil. Analysis of life tables by the key-factor method showed that mortality caused by desiccation of breeding sites was the principal factor accounting for fluctuations in population sizes throughout the study period. For cohorts that completed development, mortality attributed to predation by aquatic insects ranged from 68 to 96% and was the most important cause of death. In these cohorts, final population size was determined by the high mortality rates that occurred during the fourth larval instar and pupal stages; however, the key-factor best accounting for the population fluctuations was predation of fourth instar larvae. The short mean time of 7.1 days for synchronous development of the larvae and the lack of evidence of density-dependent mortality suggested that the A. scapularis populations were not strongly regulated and can attain relatively high densities.     

THE ECOLOGY AND BEHAVIOUR OF THE LONG-TAILED TIT

The behaviour and ecology of a population of Long-tailed Tits in Wytham Great Wood, Oxford, were studied between October 1970 and June 1971. Flocks were found to have a stable composition in autumn, and to occupy fairly discrete territories which they defended against other flocks.
In spring the population was reduced by about 50% and in the pre-nesting period flocks formed only in cold weather or just before going to roost. These flocks occupied the same territories as the autumn flocks and the individuals composing them eventually nested within the flock territory.
Predation of nests was heavy, increasing to a peak in the first half of May. Only nine pairs out of 36 managed to rear young successfully. At six out of these nine nests, supernumerary birds helped to feed the young, but it seems unlikely that this increased their success, since all young hatched fledged successfully. The presence of supernumerary birds at the nest was accompanied by frequent hover-displays by all the adults.
The distribution of nests in relation to habitat shows that there is a preference for scrub rather than mature woodland, although the latter is used extensively for feeding during the winter. There appears to have been a radical change in the height of nest-sites in Wytham Great Wood between 1955–57 and 1971, and this is probably due to predation.     

The role of predation in the decline and extirpation of woodland caribou

To select appropriate recovery strategies for endangered populations, we must understand the dynamics of small populations and distinguish between the possible causes that drive such populations to low numbers. It has been suggested that the pattern of population decline may be inversely density-dependent with population growth rates decreasing as populations become very small; however, empirical evidence of such accelerated declines at low densities is rare. Here we analyzed the pattern of decline of a threatened population of woodland caribou (Rangifer tarandus caribou) in British Columbia, Canada. Using information on the instantaneous rate of increase relative to caribou density in suitable winter foraging habitat, as well as on pregnancy rates and on causes and temporal distribution of mortalities from a sample of 349 radiocollared animals from 15 subpopulations, we tested 3 hypothesized causes of decline: (a) food regulation caused by loss of suitable winter foraging habitat, (b) predation-sensitive foraging caused by loss of suitable winter foraging habitat and (c) predation with caribou being secondary prey. Population sizes of caribou subpopulations ranged from <5 to >500 individuals. Our results showed that the rates of increase of these subpopulations varied from −0.1871 to 0.0496 with smaller subpopulations declining faster than larger subpopulations. Rates of increase were positively related to the density of caribou in suitable winter foraging habitat. Pregnancy rates averaged 92.4% ±2.24 and did not differ among subpopulations. In addition, we found predation to be the primary cause of mortality in 11 of 13 subpopulations with known causes of mortality and predation predominantly occurred during summer. These results are consistent with predictions that caribou subpopulations are declining as a consequence of increased predation. Recovery of these woodland caribou will thus require a multispecies perspective and an appreciation for the influence of inverse density dependence on population trajectories.     

Type 3 functional response of mice to gypsy moth pupae: is it stabilizing?

We conducted field experiments in 2002 and 2003 to determine whether the functional response of white-footed mice ( Peromyscus leucopus ) to gypsy moth ( Lymantria dispar ) pupae is decelerating (e.g. type 2) or accelerating (e.g. type 3) at low pupal density. In both experiments, live gypsy moth pupae were deployed in June (prior to the appearance of natural pupae) at densities of approximately 1, 8, and 35 pupae per mouse home range in oak-forest grids in upstate New York and monitored over 10 days for signs of predation. Pupae were deployed 1.5-m high on tree boles in 2002, whereas in 2003 the three density treatments were crossed with a height treatment: ground level vs 1.5-m high. The relationship between daily predation rate (proportion of pupae eaten/day) and pupal density was significantly positive in both years, indicating an accelerating functional response. Daily predation rates on ground-level pupae were substantial in the lowest density treatment, suggesting that dense mouse populations could drive gypsy moths to extinction despite an accelerating functional response. Daily predation rates on elevated pupae increased over several days in the medium and high density treatments, suggesting a lagged shift from ground- to tree-level foraging by mice. Within the high-density treatments, predation rates on pupae showed no apparent relationship with the number of pupae on a tree. Our results disagree qualitatively with simple models of type 3 functional response, in which predation rate of prey approaches zero as prey approach extinction, and support the contention that an accelerating functional response alone may be insufficient to prevent prey extinction.     

Comparative mortalities of the brown-tail moth, Euproctis chrysorrhoea (L.) (Lepidoptera: Lymantriidae), in south-east England

STERLING, P. H. & SPEIGHT, M. R., 1989. Comparative mortalities of the brown-tail moth, Euproctis chrysorrhoea (L.) (Lepidoptera: Lymantriidae), in south-east England . The mortalities within natural infestations of the brown-tail moth were investigated from 1982–1984 in four sites in south-east England. Mortality agents were quantified from field and laboratory assessments. Microsporidial disease was the most important factor, causing 91.9% mortality in large larvae in one population; cytoplasmic and nuclear polyhedrosis viruses were present, but not significant. Many species of parasitoid were recorded, particularly at the end of the larval and in the pupal stages; however, combined parasitoid mortality was low. Predators were also of little importance, although cuckoos were found to take up to 19.3% of pupae at one site. Changes in density of larvae in the spring were more strongly influenced by dispersal following defoliation than by any particular mortality agent in most cases. In terms of biological control, diseases offer the greatest potential for development.     

Shifting body weight-fecundity relationship in a capital breeder: maternal effects on egg numbers of the autumnal moth under field conditions

In the literature, various environmental factors are described as being capable of influencing the reproductive output of insect females irrespective of their body size. Still, female body size or weight is widely used as a proxy for fecundity. In the present study, a seven-year data set on the autumnal moth, Epirrita autumnata (Borkhausen) (Lepidoptera: Geometridae), was used to analyze whether the body weight-fecundity relationship in this capital breeding, cyclic forest defoliating lepidopteran is constant across years. Ambient temperature conditions and density of conspecifics during larval development, the length of the pupal period, as well as moth densities in the parent generation were examined as factors capable of modifying the body weight-fecundity relationship. While the regression slope of potential fecundity (total egg numbers per female) on pupal mass was constant across years, the mean total egg number per given body weight (the regression intercept) was significantly different between years. This residual variance in egg numbers after controlling for the effect of pupal mass was best explained by the pooled geometrid density (autumnal and winter moths) in the parent generation. The total egg number per given body weight decreased with increasing density of geometrid moths in the parent generation. Thus, maternal density effects on offspring fecundity were found in this system. Their rather weak nature suggests, however, that this maternal effect alone does not have the potential of causing cyclic population dynamics in the autumnal moth.     

Woodland small mammal population ecology in Britain: a preliminary review of the Mammal Society survey of Wood Mice Apodemus sylvaticus and Bank Voles Clethrionomys glareolus, 1982–87

The Mammal Society has co-ordinated a population survey of Wood Mice Apodemus sylvaticus and Bank Voles Clethrionomys glareolus in 13 0.81 -hectare sites in Britain. Numbers of mice and voles live-trapped using standard methods were collated every May/June and November/December from 1982 to 1987. The data were analysed with results from four independent studies in England and the corresponding assessments of tree seed crop size. Wood Mouse numbers are usually higher in winter than in summer but Bank Vole fluctuations are less regular. In deciduous woodland, Wood Mouse mean relative densities are significantly greater in the winter and the following summer after a good seed crop than after a poor one; rates of population change from summer to winter are significantly higher when a good seed crop falls. Bank Vole relative densities are significantly greater in the summer following a good seed crop than after a poor one, and rates of change from winter to summer are significantly higher. In Wood Mouse populations, tests for density dependence suggest that it is strong from summer to winter but absent from winter to summer; in Bank Voles weaker density dependence is present in both halves of the year. Thus, Wood Mouse numbers are regulated in autumn but are also influenced by seed crop size in winter and the following summer; Bank Vole numbers are less strongly regulated during both autumn and spring and are influenced by seed crop size in the following summer. Evidence is presented suggesting that populations of each species in deciduous woodlands are synchronized over the country in summer and that Wood Mice are also synchronized in winter; highs and lows tend to coincide between different sites. The yield of tree seed is shown to vary significantly from year to year and may be the cause of the synchrony, but weather effects may also be involved.     

Resources and dispersal as factors limiting a population of the tussock moth (Orgyia vetusta), a flightless defoliator

The western tussock moth (Orgyia vetusta) is very abundant on one stand of bush lupine (Lupinus arboreus) at the Bodega Marine Reserve (Sonoma Co, Calif., USA), but is sparse or absent on nearby stands. To determine what controls its abundance, both within the outbreak area and more globally, I performed experimental manipulations of resource availability and dispersal. To measure resource limitation, I inoculated 30 caged and 30 uncaged bushes with a realistic range of numbers of tussock moth eggs. On caged bushes, starvation led to density-dependent reductions in survival, pupal weight and fecundity. Larvae on uncaged bushes experienced density-independent ant predation on early instars and density-dependent emigration by late instars. From the results of this experiment, I predicted the density of a resource-limited tussock moth population. The predictions agreed fairly well with data from the outbreak area in 1992. To measure dispersal by the moth, which has flightless adult females, I released 30,000 tussock moth eggs at a central point in each of two uninfested lupine stands, and censused larvae weekly in a circle of radius 16 m until pupation. Median displacement over one entire generation was only 2 m, or about 2 bush radii. Rearing experiments indicated that bushes outside the out-break area are fully nutritionally suitable for the moth. I conclude that two major factors limiting the population are resources (within the outbreak area) and inefficient dispersal (more globally).     

Genetic evidence for extra-pair fertilizations in a monogamous passerine, the Great Tit Parus major

The incidence of extra-pair paternity in a Great Tit Parus major population at Wytham Wood, Oxford, in 1985–1987 was determined using two polymorphic allozymes. In 831 nestlings from 94 broods, 27 genetic exclusions were detected in 25 (3%) nestlings from 16 broods. Seven (44%) of these broods contained offspring that excluded the putative male parent from being the genetic parent. The distribution of exclusion types indicated that excluded offspring were the result of fertilizations by extra-pair males and not of egg-dumping. The true frequency of extra-pair paternity was estimated as 14% of offspring. These results suggest a mixed reproductive strategy for males in which they breed mo-nogamously whilst simultaneously seeking extra-pair matings with females of other pairs.     

Integrated control of apple pests in New Zealand 10. Population dynamics of codling moth in Nelson

Abstract

Sampling methods are described for estimating the population density, mortality, and natality of a univoltine population of codling moth attacking mature apple trees (cv. ‘Delicious’) at Nelson, New Zealand. These methods were used to construct life tables for the species over eight generations (1967–68 to 1974–75) on trees variously sprayed and not sprayed with ryania in an integrated control programme. Bait traps provided a sensitive measure of seasonal adult population density. Analysis of the life tables shows that migration of adults was the main key factor and that overwintering larval mortality (particularly that due to bird predation), fecundity, and ryania also made a major contribution to variation in generation mortality. In the absence of ryania the resident population usually increased between generations, whereas it usually decreased when ryania sprays were applied. The density dependence of overwintering larval mortality was due to bird predation, and the inverse density dependence of larval mortality from ryania was due to changes in the site of fruit entry with larval population density. Fecundity was density independent, and inconclusive evidence was obtained on the density dependence of migration. The wide variation in fecundity is attributed primarily to weather conditions. The impact on control strategy of the above key factors, density dependence, and total natural mortality is discussed. Ryania is found to be uneconomic, whereas the granulosis virus of codling moth and male removal with pheromone traps show promise as future control methods. The need to eliminate reservoirs of codling moth close to orchards under integrated pest control is emphasised. Regulation of codling moth populations at Nelson on neglected, unsprayed trees appears to result from intraspecific competition for fruits and cocooning sites, and weakly density-dependent mortality of mature larvae when seeking cocooning sites and while overwintering in their cocoons. Variation in fecundity also cohtributes to fluctuations in abundance of the species. In contrast, at low density in an integrated control programme no intraspecific competition was evident; migration, winter mortality, and fecundity were the main determinants of abundance. This illustrates the need to study pest populations at densities similar to those tolerable commercially.     

Maintenance of the phenology of the winter moth (Lepidoptera: Geometridae)

Adult winter moths ( Operophtera brumata (L.)) are active in late autumn or early winter. The eggs overwinter in the canopy of trees and hatch simultaneously with the bursting of host tree buds. Many young larvae disperse on the wind on silk strands. Larvae are polyphagus and feed until late spring when they pupate in soil or leaf litter. The duration of the egg and pupal stages is genetically determined and varies with latitude. The egg stage is long in the north and short in the south, while the pupal stage is short in the north and long in the south.
The literature on the ecology and physiology of winter moth is reviewed. The factors maintaining the unusual phenology are discussed. It is concluded that the larval stage is early because mature leaves of many host trees are unsuitable as food, because parasitism against later larvae is more intense, and because summer temperatures may be injurious to larvae. The adult period is late in the year so that the final stages of pupal development occur in cool conditions and so that adults emerge after most insect predators have ceased activity. Throughout most of the range retarding the adult emergence period would cause activity to be impeded by severe winter weather; in the south this is not so and it is suggested that eggs must be on the trees for a minimum period to ensure synchronization of egg hatch with bud burst. The protracted adult emergence period may be an adaptation reducing predation by birds.     

Outbreaks of the winter moth on Sitka Spruce in Scotland are not influenced by nutrient deficiencies of trees,tree budburst,or pupal predation

Summary Since the early 1980s, the winter moth, Operophtera brumata L. (Lepidoptera: Geometridae) has emerged as a serious pest of Sitka Spruce, Picea sitchensis Bong. plantations in southern Scotland. Outbreaks are characterised by susceptible sites within plantations which can occur immediately adjacent to resistant sites. We investigated the level of some nutrients in the trees, the date of budburst of the trees, and the numbers of some potential predators of winter moth pupae. None could satisfactorily explain outbreak patterns. Although foliage analysis demonstrated that many trees were marginal or deficient in phosphorus, nitrogen and potassium, these deficiencies were not related to the susceptibility of a site. Within sites, the numbers and weights of O. brumata were positively related to phosphorus content and negatively related to calcium content of foliage. Other evidence suggests, however, that these correlations may not represent direct effects of phosphorus and calcium on larval growth and survival. Date of budburst, which commonly determines susceptibility of deciduous hosts to O. brumata, was unrelated to density, and pupal predators were more, not less, abundant in susceptible sites. Although it is difficult to distinguish between factors that initiate outbreaks and those that maintain them, these data suggest that nutrient deficiencies of trees, budburst date, and the distribution of pupal predators of the winter moth cannot explain patterns of outbreak of the winter moth on spruce.