The fine-scale population dynamics of spruce budworm: survival of early instars related to forest condition

Abstract.  1. A lagged, density-dependent relationship between survival of early instars and host-tree condition is revealed during outbreaks of spruce budworm, Choristoneura fumiferana Clem. Persistent damage to hosts leads to deterioration of the stand.
2. Resource limitation affects survival during early-instar dispersal of spruce budworm. Impediments to distinguishing these events with estimates of survival were overcome with a simple model that describes the dispersal and survival processes. The model was used to analyse a recent 15-year population series from Black Sturgeon Lake and two historical datasets from Green River, in Canada.
3. Defoliation-induced damage to the trees resulted in increased losses of spring-emerging larvae that are dispersing in search of feeding sites. Losses were further exacerbated by biotic factors such as maternal fecundity, rates of infection by the pathogen, Nosema fumiferanae , and by weather-related effects on the foraging period.
4. Survival of early-stage budworm larvae in persistent outbreaks declined and the likelihood of other density-related factors such as rate of mortality from natural enemies increased. These results may reconcile outstanding differences in interpretation of the role of the forest resource in spruce budworm population dynamics and point to a common process linking the dynamics of other well-known budworm species.     

Risk of dispersal in western spruce budworm

1 Western spruce budworm Choristoneura occidentalis Free. larvae emerge in the spring before buds have expanded and spend a variable period of time foraging on branches and mining needles.
2 Losses of dispersing budworms during this needle-mining period are related directly to the severity of defoliation in previous years and inversely related to foliage biomass in the study plot and to temperature and rainfall during the needle-mining period.
3 Losses can be interpreted in terms of risk of dispersal, which is the product of the propensity of early-stage budworms to disperse in search of resources and the consequences of this behaviour for survival under variable ecological conditions.
4 A comparison of the species-specific nature of risk of dispersal in three conifer-feeding budworm systems of North America may elucidate the common nature but variable features of their respective population dynamics.     

Interguild interactions between a non-native phloem feeder and a native outbreaking defoliator

1. Competitive and synergistic interactions directly or indirectly drive community dynamics of herbivorous insects. Novel interactions between non-native and native insects are unpredictable and not fully understood. 2. We used manipulative experiments on mature red spruce trees to test interactions between a non-native phloem feeding insect, the brown spruce longhorn beetle (BSLB), and an outbreaking native defoliator, the spruce budworm. We subjected treatment trees to defoliation by three densities of spruce budworm larvae. Treatment trees were: stressed by (i) girdling (to mimic beetle feeding) or (ii) girdling + BSLB before spruce budworm larvae were introduced on branches in sleeve cages. Budworm larvae then fed on foliage and developed to pupation. We assessed all branches for budworm performance, defoliation, shoot production and shoot growth. 3. Shoot length did not differ in response to stress from girdling or BSLB infestation. Neither stress from girdling, nor interactions with BSLB feeding affected spruce budworm performance or defoliation. Intraspecific impacts on performance and defoliation in relation to budworm density were stronger than the effects of tree stress. 4. Prior infestation of red spruce by BSLB in our experimental set-up did not influence spruce budworm performance. BSLB is a successful invader that has blended into its novel ecological niche because of ecological and phylogenetic similarities with a native congener, Tetropium cinnamopterum. 5. Outbreaks by BSLB will not likely impede or facilitate spruce budworm outbreaks if they co-occur. It would be useful to evaluate the reverse scenario of BSLB success after defoliation stress by spruce budworm.     

Identifying spatial relationships at multiple scales: principal coordinates of neighbour matrices (PCNM) and geostatistical approaches

We compared two methodological approaches – principal coordinate analysis of neighbour matrices (PCNM) and geostatistics – that both aim at extracting several spatial scales in order to identify spatial relationships between organisms and environmental variables at multiple scales. From a statistical point of view, PCNM analysis and geostatistics come from "two different worlds"– PCNM is based on classical "data analysis" while geostatistical modelling is developed in a probabilistic context. These two methods were used to investigate the spatial relationships between defoliation caused by spruce budworm Choristoneura fumiferana and bioclimatic conditions in Ontario since 1941 through a wide range of scales. On the one hand, PCNM variables related to defoliation frequency were partitioned into four spatial submodels representing respectively four spatial scales: very broad scale (ca>300 km), broad scale (ca 180 km), fine (ca 100 km), and very fine (<80 km). On the other hand, nested variogram modelling was used to identify the relevant scales. The nested variogram model was composed of four variograms with different characteristic scales close to those of the PCNM spatial submodels. Maps of PCNM submodels and kriging components revealed similar spatial patterns of defoliation frequency at very broad and broad scales while spatial patterns at fine and very fine scales looked quite different. Both methods showed that defoliation by spruce budworm occurs at the broader spatial scales but may be explained by fluctuations at the smaller scales. Finally, results based on geostatistics using a Linear Model of Coregionalisation suggested that climatic conditions can be considered to act at the level of outbreak dynamics while the tree community of spruce budworm's principal hosts controls local population dynamics.     

Variation in budburst phenology of Douglas-fir related to western spruce budworm (Lepidoptera: Tortricidae) fitness

Variation in budburst phenology among individual trees of interior Douglas-fir (Pseudotsuga menziesii var. glauca [Beissn.] Franco) may influence their susceptibility to western spruce budworm (Choristoneura occidentalis Freeman) defoliation. We tested the hypothesis that phenological asynchrony between Douglas-fir and the western spruce budworm is a mechanism of resistance using clones derived from parent trees that showed resistance versus susceptibility to C. occidentalis defoliation in the field. Susceptible clones had earlier budburst phenology compared with resistant clones when they were grown in a common greenhouse environment, demonstrating a genetic basis for parallel phenological differences exhibited by the parent trees. We tested the importance of phenological asynchrony as a factor influencing fitness of C. occidentalis using two different greenhouse bioassay experiments. One experiment compared western spruce budworm performance on equivalent phenological stages of susceptible and resistant clones by matching larval feeding to the columnar (fourth) bud development stage of each clone. Larvae reared on resistant clones had greater realized fitness (i.e., number of F1 offspring produced) than those reared on susceptible clones when the influence of variation in budburst phenology was minimized. In the other experiment, western spruce budworm larvae were placed on all trees on the same date when approximately 50% of all terminal buds in the population were in the yellow (second) budburst stage. Larvae reared on susceptible clones had greater realized fitness than those reared on resistant clones when the influence of phenological asynchrony was expressed. Our results suggest that resistant phenotypes of Douglas-fir have negative effects on survival and reproduction of C. occidentalis under the natural conditions that insects and trees experience in the field. Genetic variation among trees in budburst phenology has an important influence on interactions between the western spruce budworm and Douglas-fir.     

Spatial scale and the detection of density dependence in spruce budworm outbreaks in eastern North America

Using two tests for direct density dependence and standard techniques of time series analysis, we identified density dependence in defoliation time series of the spruce budworm across its outbreak range in eastern North America over the years 1945–1988. We carried out analyses for the entire region and for grid cells of defoliation maps at five spatial scales created by aggregating the smallest grid cells. The rate of detection of direct density dependence, as assessed by two previously published methods, decreased with increasing spatial scale. Using both methods, density dependence was detected more frequently at the periphery of the outbreak range, where defoliation rate was lower. This result suggested that density-dependent regulation may be stronger in those areas. The first order autoregressive process was the basic model for defoliation dynamics overall and the most common model across spatial scales. Second-order processes were encountered much less frequently, and those commonly identified as resulting from delayed density dependence generally occurred across spatial scales at a rate expected by chance alone. Our results were similar to those of other published studies, which have found the detection of density dependence to decrease at larger spatial scales. The results also reinforced the importance of considering spatial scale when diagnosing population processes using time series of abundance for single species. Received: 26 December 1999 / Accepted: 17 March 2000     

A lagged, density-dependent relationship between jack pine budworm Choristoneura pinus pinus and its host tree Pinus banksiana

Abstract. 1. Survival of newly emerged jack pine budworm Choristoneura pinus pinus is related to the density of available pollen cones (microsporangiate strobili) produced by its host tree, jack pine Pinus banksiana. 2. A 7‐year time series of observations from a plot network in Ontario, Canada, compared the propensity of jack pine to produce pollen cones, τ, on trees that were either defoliated or undisturbed by the jack pine budworm. 3. Non‐defoliated jack pine trees have a high propensity to produce pollen cones. More than one‐third of these trees produced pollen cones in every year of the series. Propensity varied significantly among plots and trees. Temporal patterns in propensity were also highly variable but within a plot propensity was often autocorrelated in time. 4. Defoliation by the jack pine budworm was associated with forest plots composed of the oldest and the largest trees and with the fewest trees per hectare. Within a plot, outbreaks lasted 3 or 4 years although individual trees were only defoliated in 1 or 2 years. 5. The propensity to produce pollen cones in jack pine was reduced in the years after defoliation. The most pronounced reductions in propensity occurred where defoliation was most severe. 6. The reduction in propensity to produce pollen cones resulting from previous defoliation, coupled with the dependence of jack pine budworm survival on the availability of pollen cones, induces a lagged, negative feedback between the density of the consumer and that of its resource. 7. The lagged, density‐dependent relationship between jack pine budworm and its jack pine host contributes to oscillatory dynamics of the jack pine budworm. Comparison of the outbreak behaviour of jack pine budworm with that of the closely related eastern spruce budworm C. fumiferana suggests that differences in the strength of the host‐plant interaction may account for differences in the relative frequency of outbreaks in the respective systems.     

How does synchrony with host plant affect the performance of an outbreaking insect defoliator?

Phenological mismatch has been proposed as a key mechanism by which climate change can increase the severity of insect outbreaks. Spruce budworm (Choristoneura fumiferana) is a serious defoliator of North American conifers that feeds on buds in the early spring. Black spruce (Picea mariana) has traditionally been considered a poor-quality host plant since its buds open later than those of the preferred host, balsam fir (Abies balsamea). We hypothesize that advancing black spruce budbreak phenology under a warmer climate would improve its phenological synchrony with budworm and hence increase both its suitability as a host plant and resulting defoliation damage. We evaluated the relationship between tree phenology and both budworm performance and tree defoliation by placing seven cohorts of budworm larvae on black spruce and balsam fir branches at different lags with tree budburst. Our results show that on both host plants, spruce budworm survival and pupal mass decrease sharply when budbreak occurs prior to larval emergence. By contrast, emergence before budbreak decreases survival, but does not negatively impact growth or reproductive output. We also document phytochemical changes that occur as needles mature and define a window of opportunity for the budworm. Finally, larvae that emerged in synchrony with budbreak had the greatest defoliating effect on black spruce. Our results suggest that in the event of advanced black spruce phenology due to climate warming, this host species will support better budworm survival and suffer increased defoliation.     

Anatomy of an outbreak of jack pine budworm, <Emphasis Type="Italic">Choristoneura pinus pinus</Emphasis> (Lepidoptera: Tortricidae)

We report data collected over the entire course of an outbreak of jack pine budworm, Choristoneura pinus pinus Freeman (Lepidoptera: Tortricidae), between 1984 and 1988 at 12 plots in Manitoba. The positive relationship between the level of defoliation, used as a proxy for population density, and the abundance of eggs suggests local reproduction by females. The density of pollen cones apparently reduced larval fitness in 1986, when flowers were least abundant, but had limited impact in other years when pollen cones were abundant; this suggests that the relative abundance of pollen cones is more likely to influence the termination of an outbreak than its onset. Considering the conditions that prevail at the onset of the outbreak (low defoliation combined with a high abundance of eggs), a predictive tool may be developed to anticipate outbreaks of jack pine budworm based on environmental conditions that are conducive to high survival and/or fecundity of females.     

Modeling Spruce Budworm Population Revisited: Impact of Physiological Structure on Outbreak Control

Understanding the dynamics of spruce budworm population is very important for the protection of spruce and balsam fir trees of North American forests, and a full understanding of the dynamics requires careful consideration of the individual physiological structures that is essential for outbreak control. A model as a delay differential equation is derived from structured population system, and is validated by comparing simulation results with real data from the Green River area of New Brunswick (Canada) and with the periodic outbreaks widely observed. Analysis of the equilibrium stability and examination of the amplitudes and frequencies of periodic oscillations are conducted, and the effect of budworm control strategies such as mature population control, immature population control and predation by birds are assessed. Analysis and simulation results suggest that killing only budworm larvae might not be enough for the long-term control of the budworm population. Since the time required for development during the inactive stage (from egg to second instar caterpillar) causes periodic outbreak, a strategy of reducing budworms in the inactive stage, such as removing egg biomass, should also be implemented for successful control.     

Larval feeding behaviour affects the impact of staminate flower production on the suitability of balsam fir trees for spruce budworm

Laboratory rearing of spruce budworm, Choristoneura fumiferana, in conjunction with field rearing indicated that the feeding behaviour of the larvae, which is affected by the insect population density, significantly influenced the impact of balsam fir, Abies balsamea, staminate flowering on spruce budworm biology. At low budworm density, the production of pollen in the midcrown of host trees reduced the insect development time by 5 days without affecting pupal weight, fecundity and survival. However, at high budworm density, the small amount of current-year foliage produced by flowering branches forced old larvae (sixth instar) either to feed on 1-year-old foliage (backfeeding) or to move from the midcrown to the lower crown section where staminate flowers are absent and more current-year foliage is available. When old larvae fed on old foliage, they exhibited reduced pupal weight and fecundity without losing the advange in development time that they obtained from feeding on pollen during their early stages of development. On the other hand, when old larvae moved to the lower crown section, they avoided the negative effects of backfeeding but lost the advantage in development time that was gained from feeding on pollen. Results from this study indicated that the production of staminate flowers by balsam fir trees could have opposite effects on spruce budworm population dynamics depending upon the insect population density when flowering occurs.     

Forest composition, host-population density, and parasitism of spruce budworm Choristoneura fumiferana eggs by Trichogramma minutum

The frequency of egg parasitism in the spruce budworm, Choristoneura fumiferana (Clem.) (Lepidoptera: Tortricidae) by Trichogramma minutum Riley (Hymenoptera: Trichogrammatidae) was recorded in several host populations during the rising phase of an outbreak in the Gatineau Valley (Quebec, Canada) in 1999 and 2000. Some of the observed spruce budworm populations were subjected to experimental aerial applications of Bacillus thuringiensis var. kurstaki during the course of the study. Untreated spruce budworm population densities at the onset of the study were directly related to stand composition (the proportion of non‐host basal area) but unrelated to plant species diversity. Much of the variation in budworm egg mass parasitism rates was accounted for (77.1% of total variance) by three variables: year, spruce budworm egg population density, and proportion of non‐host basal area in the stand. The exploitation of eggs within parasitised egg masses was related to year, B. thuringiensis treatment, egg mass size, and the proportion of egg masses parasitised in the host population (76.9% of variance). The yearly variation in parasitism rate was considerable, and was consistent with climatic conditions, namely heat accumulation and precipitation during the spruce budworm oviposition period. Parasitism by T. minutum on eggs of C. fumiferana was recorded in 16 additional sites in 2000, to confirm the general form of the inverse density dependence of parasitism. Observations recorded by Neilson in the Green River area of New Brunswick (Canada) between 1948 and 1957 are discussed in the context of the present analysis. These results indicate that parasitism by T. minutum may be more predictable than previously believed and may be an important and consistent source of mortality in the several years preceding an outbreak of C. fumiferana, especially in mixed‐wood stands and in warmer, drier portions of the spruce budworm's range.     

Misleading correlations: The case of the Canada warbler and spruce budworm

In a recently published article Sleep et al. ( 2009 ) suggested that 30-yr declines in Canada warbler (1975–2005) based on Breeding Bird Survey (BBS) data could be attributed to 30-yr declines in spruce budworm through a series of analyses based principally on correlations. We demonstrate that the relationship does not hold when a longer-term data set (1968–2008) is used for the analysis. Sleep et al. ( 2009 ) also demonstrated a positive relationship between correlations of Canada warbler abundance and time and spruce budworm defoliation and time using provincial data sets. We examined the underlying BBS data and found that there were insufficient observations of Canada warbler in the western provinces (Alberta and Manitoba) to support the conclusion because most Canada warbler observations occurred far (>100 km) from any budworm defoliation. As well, we used a density-dependent stochastic population growth model as proposed by Sleep et al. ( 2009 ) and found only 2 significant relationships (Nova Scotia, Alberta) between Canada warbler population growth rate and budworm defoliation at the provincial scale. We conclude that little analytical evidence exists to support the idea that Canada warbler decline is a function of spruce budworm decline. © 2011 The Wildlife Society.     

Interspecific variation in resistance of two host tree species to spruce budworm

Woody plants regularly sustain biomass losses to herbivorous insects. Consequently, they have developed various resistance mechanisms to cope with insect attack. However, these mechanisms of defense and how they are affected by resource availability are not well understood. The present study aimed at evaluating and comparing the natural resistance (antibiosis and tolerance) of balsam fir (Abies balsamea [L.] Mill.) and white spruce (Picea glauca [Moench) Voss] to spruce budworm, Choristoneura fumiferana (Clem.), and how drainage site quality as a component of resource availability affects the expression of resistance over time (6 years). Our results showed that there are differences in natural resistance between the two tree species to spruce budworm, but it was not significantly affected by drainage quality. Balsam fir exhibited higher foliar toxic secondary compounds concentrations than white spruce in all drainage classes, resulting in lower male pupal mass, survival and longer male developmental time. This, however, did not prevent spruce budworm from consuming more foliage in balsam fir than in white spruce. This response suggests that either natural levels of measured secondary compounds do not provide sufficient toxicity to reduce defoliation, or spruce budworm has developed compensatory mechanisms, which allow it to utilize food resources more efficiently or minimize the toxic effects that are produced by its host's defensive compounds. Larvae exhibited lower pupal mass and higher mortality in rapidly drained and subhygric sites. Drainage class also affected the amount of foliage destroyed but its impact varied over the years and was probably influenced by climatic variables. These results demonstrate the complexity of predicting the effect of resource availability on tree defenses, especially when other confounding environmental factors can affect tree resource allocation and utilization.     

Three years of aerial field experiments with Bacillus thuringiensis plus chitinase formulation against the spruce budworm

From 1971 to 1973 several Bacillus thuringiensis formulations were tested in the field against larvae of the spruce budworm under various conditions of population and tree defoliation. The results showed B. thuringiensis treatments can be a weapon in the control of spruce budworm outbreaks and the beneficial effect of B. thuringiensis treatments appear to be prolonged over 1 or 2 years. A new compact formulation was developed making B. thuringiensis treatments more economical and competitive with chemical insecticides.     

Modelling Spatial Interactions Among Fire,Spruce Budworm,and Logging in the Boreal Forest

In the boreal forest, fire, insects, and logging all affect spatial patterns in forest age and species composition. In turn, spatial legacies in age and composition can facilitate or constrain further disturbances and have important consequences for forest spatial structure and sustainability. However, the complex three-way interactions among fire, insects, and logging and their combined effects on forest spatial structure have seldom been investigated. We used a spatially explicit landscape simulation model to examine these interactions. Specifically, we investigated how the amount and the spatial scale of logging (cutblock size) in combination with succession, fire, and spruce budworm outbreaks affect area burned and area defoliated. Simulations included 30 replicates of 300 years for each of 19 different disturbance scenarios. More disturbances increased both the fragmentation and the proportion of coniferous species and imposed additional constraints on the extent of each disturbance. We also found that harvesting legacies affect fire and budworm differently due to differences in forest types consumed by each disturbance. Contrary to expectation, budworm defoliation did not affect area burned at the temporal scales studied and neither amount of logging nor cutblock size influenced defoliation extent. Logging increased fire size through conversion of more of the landscape to early seral, highly flammable forest types. Although logging increased the amount of budworm host species, spruce budworm caused mortality was reduced due to reductions in forest age. In general, we found that spatial legacies do not influence all disturbances equally and the duration of a spatial legacy is limited when multiple disturbances are present. Further information on post-disturbance succession is still needed to refine our understanding of long-term disturbance interactions.     

Effects of artificial and western spruce budworm (Lepidoptera: Tortricidae) defoliation on growth and biomass allocation of Douglas-fir seedlings

Artificial defoliation has been used commonly to simulate defoliation by insect herbivores in experiments, in spite of the fact that obvious differences exist between clipping foliage and natural defoliation due to insect feeding. We used a greenhouse experiment to compare the effects of artificial and western spruce budworm (Choristoneura occidentalis Freeman) defoliation on the growth and biomass allocation of 3-yr old half-sib seedlings from mature Douglas-fir [Pseudotsuga menziesii (Mirb.) Franco variety glauca] trees that showed phenotypic resistance versus susceptibility to budworm defoliation in the forest. Artificial clipping of buds mimicked the effects of budworm feeding on total seedling biomass when 50% of the terminal buds were damaged. However, artificial defoliation decreased seedling height, relative growth rate of height, and shoot: root ratio more than budworm defoliation, whereas budworm defoliation decreased stem diameter relative growth rate more than artificial defoliation. Half-sib seedling progeny from resistant maternal tree phenotypes had greater height, diameter, biomass, and shoot: root ratio than seedlings from susceptible phenotypes. We concluded that careful artificial defoliation could generally simulate effects of budworm defoliation on total biomass of Douglas-fir seedlings, but that the two defoliation types did not have equal effects on biomass allocation between shoot and root. Further, an inherently higher growth rate and a greater allocation of biomass to shoot versus root are associated with resistance of Douglas-fir trees to western spruce budworm defoliation.     

Warming counteracts defoliation‐induced mismatch by increasing herbivore‐plant phenological synchrony

Climate change is altering phenology; however, the magnitude of this change varies among taxa. Compared with phenological mismatch between plants and herbivores, synchronization due to climate has been less explored, despite its potential implications for trophic interactions. The earlier budburst induced by defoliation is a phenological strategy for plants against herbivores. Here, we tested whether warming can counteract defoliation‐induced mismatch by increasing herbivore‐plant phenological synchrony. We compared the larval phenology of spruce budworm and budburst in balsam fir, black spruce, and white spruce saplings subjected to defoliation in a controlled environment at temperatures of 12, 17, and 22°C. Budburst in defoliated saplings occurred 6–24 days earlier than in the controls, thus mismatching needle development from larval feeding. This mismatch decreased to only 3–7 days, however, when temperatures warmed by 5 and 10°C, leading to a resynchronization of the host with spruce budworm larvae. The increasing synchrony under warming counteracts the defoliation‐induced mismatch, disrupting trophic interactions and energy flow between forest ecosystem and insect populations. Our results suggest that the predicted warming may improve food quality and provide better growth conditions for larval development, thus promoting longer or more intense insect outbreaks in the future.     

Spruce budworm (Choristoneura spp.) biotype reactions to forest and climate characteristics

The spruce budworm (Choristoneura fumiferana) is the most destructive insect defoliator of forests in North America. Climatic influences on this species' life history are considered a major factor in restricting the extent and intensity of outbreaks. We examine the life history traits of the spruce budworm and related Choristoneura populations with respect to forecasting the conifer‐feeding responses of these insects in changing environments. Analysis of the evolutionary relationships between Choristoneura entities, including their hybridization, genetic distances, and their degree of sympatry leads us to distinguish 15 possible Choristoneura‘biotypes’. Population trend has been associated with recruitment to the feeding stage, and two indicators of recruitment, egg weights and phenological development, are both ‘biotype’ and climate dependent. Among Abietoid feeding ‘biotypes’ and among spruce budworm populations, those from locations with extreme winters tend to have heavier eggs than those from the more benign environments. In spruce budworm, this genetically based adaptation allows populations to increase their potential recruitment substantially where winters are mild. All biotypes feed on the newly developed shoots of their host trees in spring, and are thus vulnerable to the uncertain timing of budbreak. Genetic control of spring emergence is weak so larvae from a single family typically exit from hibernacula over a prolonged period. This guarantees some synchronization with budburst. However, hybrid populations have high heritabilities. This allows rapid adaptation to new conditions (e.g. mixed host‐species stands). Geographic variation in phenological development after establishing feeding sites is largely genetically controlled. The importance of variation in these traits is examined with respect to competing population dynamics theories to evaluate their utility in forecasting future trends in defoliation. We finish with a plea for jointly using alternative approaches in forecasting spatiotemporal patterns of defoliation.     

Host-plant influence on the population ecology of the jack pine budworm, Choristoneura pinus (Lepidoptera: Tortricidae)

Abstract.

• 1 Newly-emerged, second-instar jack pine budworm (Choristoneura pinus Freeman) establish spring feeding sites preferentially in the pollen cones of their host tree, Pinus banksiana Lamb.
• 2 Laboratory studies showed that the rate of establishment and survival of jack pine budworm on pollen cones was high throughout the entire spring emergence period of the insect.
• 3 In contrast, the rate of establishment and survival of jack pine budworm on vegetative buds was very poor early in the spring. Vegetative buds were only acceptable as feeding sites to the jack pine budworm for a relatively brief period in late spring.
• 4 Field studies showed that the change in population density of jack pine budworm during the spring emergence stage, as expressed by k-values, was a function of the abundance of pollen cones in the stand. Population reduction was greatest in those stands with the fewest pollen cones.
• 5 Direct measurement of spring dispersal by jack pine budworm showed that dispersal and consequent losses to the budworm population were greatest in stands with the fewest pollen cones.
• 6 We conclude that changes in the density of jack pine budworm are strongly influenced by production of pollen cones in the host stand. Because pollen cone production is related to previous years of defoliation by the jack pine budworm, we propose that pollen cones act as a density-dependent factor governing the density of early-stage jack pine budworm.
• 7 The resulting dynamics are compared to those of other budworm species and used to explain observed regional and temporal patterns of jack pine budworm outbreaks.