Crowding-induced responses in a geometrid moth revisited: a field experiment

In Lepidoptera, effects of larval crowding on life history traits may manifest themselves as changes in growth rate and duration of larval period. For the outbreaking geometrid moth, Epirrita autumnata , impacts of crowding have been shown in earlier laboratory studies, and these responses were modified by diet quality. In this study, the aim was to find out the specific nature of crowding effects of E. autumnata in the field. E. autumnata larvae were reared individually and in groups in mesh bags on mountain birch, both on good and poor quality foliage. Three field experiments were carried out with different densities in the crowded treatment (12, 21 and 45 larvae per 100 short shoots of mountain birch), simulating intermediate to severe outbreaks. The study revealed the density interval (12–21) within which impacts of crowding become evident. Significant effects were mostly found with the two highest densities in the crowded treatment combined with poor foliage quality. In those treatments, crowding resulted in 8–12% and 24–33% decrease in pupal mass and larval survival, respectively. The results of the present field experiments, however, did not corroborate the results of earlier laboratory studies: duration of larval period was not substantially affected, net effects of crowding were negative and interactions between crowding and foliage quality were contrary to those obtained in earlier laboratory studies. In many other Lepidoptera as well, the knowledge on responses to crowding may be mostly qualitative since most crowding experiments have been conducted in laboratory.     

Spatial responses of two herbivore groups to a geometrid larva on mountain birch

Direct or plant-mediated interactions between herbivores may modify their spatial distribution among and within plants. In this study, we examined the effect of a leaf-chewing geometrid, the autumnal moth (Epirrita autumnata), on two different herbivore groups, leaf rolling Deporaus betulae weevils and Eriocrania spp. leafminers, both feeding on mountain birch (Betula pubescens ssp. czerepanovii). The exact locations of herbivores within tree canopies were mapped during three successive summers. In the first 2 years, some trees were artificially colonized by eggs of the autumnal moth to induce both rapid and delayed resistance in the foliage. The natural infection levels of the pathogenic rust fungus (Melampsoridium betulinum), potentially involved in species interactions, were also recorded. At the level of the whole tree, the density of D. betulae leaf rolls was lower in trees infested by the autumnal moth in the same year. However, the feeding locations within trees were partly segregated: D. betulae favoured shadier branches, while E. autumnata preferred the sunny parts of the canopy. The autumnal moth did not affect current- or following-year density of leafminers at the tree or branch level. Trees infected by rust had fewer leafminers in the same summer than noninfected trees. There were no interaction effects between defoliation by the autumnal moth and rust infection, and no delayed effects on the abundance of other herbivores the following year. Taken together, these findings suggest that the autumnal moth has a negative, partially plant-mediated impact on D. betulae, and can reduce the extent of current-year defoliation caused by D. betulae. This may be beneficial for the mountain birch, since the greater part of D. betulae damage occurs around or after the end of the larval period of the autumnal moth, which may be a critical time for tree recovery after moth outbreaks.     

Stand age-structure influence in a low population peak of Epirrita autumnata in a mountain birch forest

Population densities of Epirrita autumnata caterpillars were recorded from 1984 to 1990 m four mountain birch stands of two age categories, young and old Caterpillar populations peaked m 1986–87, and peak densities were higher in the old than in the young stands Recorded peak densities were low compared with synchronous populations with outbreak densities in neighbouring areas Differences in caterpillar densities between young and old stands could not be explained by differences in survival of early instar caterpillars or in parasitoid-caused larval mortality Instead, tree-age related differences in amounts of suitable oviposition sites and possible changes in food quality due to ageing and lower vigour of mature trees are suggested to explain why E autumnata populations tend to be higher in old stands     

Determinants of mountain birch growth in situ: effects of temperature and herbivory

Variation in in situ growth performance of the mountain birch as indicated by the widths of annual rings was analysed and related mainly to temperature and herbivory using ring width series from five heath forest sites in the Lake Torneträsk area, northern Sweden. Climate explained 48–64% of the variation in age-corrected mean ring width series. In general, the effect of current year July followed by June temperature was most important at all sites. A warm May resulted in wider rings due to an earlier budburst. Short-term (inter-annual) responses to increased temperature were in most cases not reflected into long-term responses (decades). A large proportion of the variation in stem mean ring width was due to variation among stems within trees (81%) in these polycormic trees, while variation among sites was marginal (0.4%). Within trees, main stems grew faster and were more responsive to climate variation than subordinate stems. No effect of insect herbivory on ring width was found at low defoliation levels (≤12%). At a defoliation level of ca 84% a one-year reduction in stem growth was observed while the growth reduction (ca 50% reduction in ring width) lasted for 4 yr after ca 93% defoliation. After outbreaks resulting in complete defoliation and some stem mortality, ring widths of surviving stems mainly responded with increased growth. Basal sprouts, emerging just after a severe insect outbreak with a high mortality of old stems, grew faster than sprouts occurring during other periods. It is concluded that the mountain birch is well adapted to recover from Epirrita outbreaks; the ability to produce basal sprouts, that can benefit from an existing root system for fast initial growth, is one important mechanism for this.     

Movement and disappearance of mountain birch defoliators are influenced by the interactive effects of plant architecture and induced resistance

Abstract. 1. Changes in herbivore movement and feeding behaviour may determine the efficacy of induced plant resistance by affecting the location of damage within the foliage and by modifying the vulnerability of herbivores to predators. 2. Observations of larval feeding sites were used to test whether induced resistance increased the movement of free‐living Epirrita autumnata Borkh. (Lepidoptera, Geometridae) larvae feeding on mountain birch [Betula pubescens ssp. czerepanovii (Orlova) Hämet‐Ahti]. The amount of defoliation at different canopy parts was measured to test the associated changes in the spread of damage within the foliage. 3. The architectural complexity of trees was measured to test its association with the disappearance of larvae from their hosts. The underlying hypothesis was that the architectural traits of the host plant could affect disappearance by influencing the frequency of herbivores encountering predators. 4. Distance between the consecutive feeding positions, the number of leaves damaged, and consumption of long shoot leaves all increased in trees with induced resistance. 5. Disappearance of larvae depended on the architectural complexity of trees. The effect of complexity differed between defoliation treatments, and may depend on the activity and number of predators in relation to the canopy size. 6. Accordingly, this study suggested that the interactive effects of plant architecture, induced resistance, and herbivore behaviour can determine the performance of herbivores on their host plant.     

Uncaged larvae elicit a combination of local and integrated growth responses within mountain birch crown

The degree of autonomy shown by plant structures in response to natural herbivory remains uncertain. Previous studies have usually been based on simulated herbivory or the use of caged herbivores. In this study, experimental trees were subjected to long-lasting defoliation by uncaged insect larvae. The exact 3D architecture of the whole shoot system and the amount of local damage elicited by the larvae in individual tree shoots were measured to investigate herbivore-elicited changes in growth at all levels of branching hierarchy within the crowns of mountain birch ( Betula pubescens ssp. czerepanovii ). Multiple explanatory variables were derived from the data. Defoliation by uncaged larvae elicited a combination of local and integrated growth responses. Especially the integrated responses were clear and consistent during the two years of study and resulted in reduced elongation growth. The localised component most often included the effect of defoliation at several branching orders, including also changes in self-shading. This suggests the involvement of source–sink interactions or other changes in the control of resource allocation within a larger scale. Therefore, the assumption of autonomous modules may offer an overly limited view of plant responses to localised herbivory, which appear to involve both localised and integrated components.     

Ellagitannins: defences of Betula nana against Epirrita autumnata folivory?

• 1 The induced resistance of the subarctic mountain birch Betula pubescens ssp. czerepanovii is a well‐characterized phenomenon, whereas the induced responses of Betula nana L., one of the parental species of mountain birch, have not yet been characterized. Betula nana is more resistant to several classes of insectivorous herbivores than the mountain birch, although the mechanisms responsible for the better ability to resist herbivores are not known.
• 2 The present study aimed to determine the metabolic changes that are induced by early season herbivory in B. nana leaves and to study the effects of rapidly induced resistance on the growth of Epirrita autumnata larvae.
• 3 Defoliation of B. nana was accomplished by E. autumnata larvae and leaf samples for chemical analyses were collected when the defoliating larvae were at their third and fifth instar. At the same time, laboratory assays for the growth and consumption rates of E. autumnata larvae were conducted.
• 4 The wounding of leaves by E. autumna larvae induced the production of ellagitannins (ETs) in B. nana. Intriguingly, the concentrations of protein‐bound amino acids were also induced by herbivory; however, an increase in proteins was not mirrored in the growth rate of larvae, which was less on the induced foliage. The decreased growth rate of larvae was apparently linked to the increased concentrations of oxidatively‐active ETs and the high concentration of ETs may explain the better resistance of this parental species compared with the hybrid mountain birch with its lower levels of ETs.

     

Delayed inducible resistance against a leaf-chewing insect in four deciduous tree species

Summary Both mechanical damage to mountain birch foliage and rearing of moth larvae on the trees reduced the growth of Epirrita autumnata larvae reared on these trees in the following year. The effects of physical damage and some other cues from insects were additive. On bird cherry the performance of Epirrita larvae was equal on untreated trees and on trees artificially defoliated in the previous year, but larval growth was reduced on previously insect-damaged branches. With mountain ash just physical damage per se reduced the performance of Epirrita larvae. On Salix phylicifolia there were no significant differences in the growth or survival of Epirrita on untreated control bushes and on bushes with partial larval damage during the previous year. Among untreated control trees the growth and survivorship of Epirrita were higher on fast-growing willow and bird cherry than on the slow-growing mountain birch. Mountain birch and mountain ash, the two deciduous tree species adapted to nutrient-poor soils, showed delayed inducible resistance triggered by defoliation (artificial or insect-made). This supports the hypothesis that delayed inducible resistance may be a passive response due to nutrient-stress caused by defoliation. On the other hand, the additional increase in the resistance of mountain birch triggered by specific cues from insects suggests that this response may be an evolved defense against leaf-eating insects.     

Delayed induced changes in the biochemical composition of host plant leaves during an insect outbreak

In birch, Betula pubescens, herbivore-induced delayed induced resistance (DIR) of defoliated trees may cause a strong reduction in the potential fecundity of a geometrid folivore Epirrita autumnata. In this study, we examined the biochemical basis of DIR in birch leaves during a natural outbreak of E. autumnata. A set of experimental trees was defoliated at four sites by wild larvae in the peak year of the outbreak, whereas control trees were protected from defoliation by spraying with an insecticide. The biochemical composition of leaves was analysed in the following year and, although the DIR response was weak during this outbreak, causing less than a 20% reduction in the potential fecundity of E. autumnata, some consistent relationships between defoliation, biochemistry and pupal mass of E. autumnata suggested a general biochemical basis for the defoliation-induced responses in birch leaves. Total concentrations of nitrogen, sugars and acetone-insoluble residue (e.g. cell wall polysaccharides, cell-wall-bound phenolics, protein, starch, lignin and hemicellulose) were consistently lower, and total concentrations of phenolics, especially of gallotannins and soluble proanthocyanidins, were higher in the leaves of trees defoliated in the previous year than in those protected from defoliation. The capacity of tannins to precipitate proteins correlated with contents of gallotannins, and was highest in defoliated trees. The pupal mass of E. autumnata showed a strong, positive correlation with concentrations of nitrogen and sugars, and a negative correlation with the acetone-insoluble residue and gallotannins in foliage. Correlations with other measured biochemical traits were weak. The correlation coefficients between biochemical traits and pupal mass consistently had similar signs for both defoliated and insecticide–sprayed trees, suggesting that variation in leaf quality due to defoliation in the previous year was based on similar biochemical traits as variation for other reasons. We suggest that DIR is associated with reduced growth activity of leaves, and may be seen as a delay in the biochemical maturation of leaves in defoliated trees. This explains the high concentration of gallotannins in defoliated trees, a characteristic feature of young leaves. However, the lower content of nitrogen and the higher content of soluble proanthocyanidins in defoliated trees are traits usually characterising mature, not young, leaves, indicating defoliation-induced changes in chemistry in addition to modified leaf age. Our results emphasise the importance of understanding the natural changes in chemistry during leaf maturation when interpreting defoliation-induced changes in leaf biochemistry. Received: 26 January 1998 / Accepted: 10 April 1998     

Opposing effects of spring defoliation on late season oak caterpillars

ABSTRACT. 1. The pedunculate oak, Quercus rohur L., suffers high annual levels of spring defoliation in Wytham Woods. near Oxford.
2. This spring defoliation affects late season caterpillars through a variety of damage-induced changes in the leaves.
3. Diurnea fagella (D. & S.), one of the commonest late season caterpillars, shows reduced larval survival and pupal weight on regrowth foliage when compared to undamaged primary foliage.
4. D. fagella also suffer higher larval mortality on naturally damaged primary foliage than they do on undamaged foliage.
5. Despite this, the three commonest late season caterpillar species are more abundant on damaged trees than undamaged ones. and their distributions are biased towards damaged leaves within the canopy.
6. Other factors that may be more important than leaf damage in determining the distribution and abundance of late season caterpillars are discussed. D.fagella larvae spin their larval refuges more rapidly on damaged than undamaged foliage, and this may reduce mortality by natural enemies, or ameliorate adverse effects of weather.     

Effects of host-plant shift on immune and other key life-history traits of an eruptive Geometrid, Epirrita autumnata (Borkhausen)

Abstract.  1. Population density of Epirrita autumnata (Lepidoptera: Geometridae) reaches outbreak densities regularly in northernmost Scandinavia. During these outbreak years, the most abundant host species, the mountain birch ( Betula pubescens ssp. czerepanovii ), is regularly exhausted, although larvae may rescue themselves from starvation by using alternative host species.
2. In this paper, the effects of the shift of host species on the immune defence and other life-history traits of E. autumnata were investigated, and possible consequences for population dynamics were briefly discussed. Moth larvae were reared on the leaves of the main host, mountain birch, until larvae reached their third instar. After this, larvae were allocated randomly to five treatments: larvae were either allowed to finish larval stage on the mountain birch or were shifted onto four alternative host species that are typical species for the area.
3. As expected, the host species had a major effect on fitness traits: body weight, development, and survival rate of the moths. The pupal weight was lower and development rates slower on the three alternative host species, Salix myrsinifolia Salisb., Vaccinium uliginosum L., and Betula nana L., than on the main host, mountain birch.
4. The immunity was, however, the same or better on the alternative hosts than on the main host. The immunity and pupal weights were negatively related, suggesting a trade-off between body size and immunocompetence.
5. The decreased body size and fecundity of E. autumnata during outbreak years may be partly due to the shift to alternative host species whereas the host-plant species probably does not affect markedly the rate of parasitism.     

The impact of defoliation on the foliar chemistry of southern rātā (

Brushtail possums (Trichosurus vulpecula) tend to eat young canopy foliage in southern rätä (Metrosideros umbellata), and browsing tends to be concentrated on only a few trees. Samples collected as part of an artificial defoliation experiment were analysed for NPK (nitrogen, phosphorus, and potassium), carbohydrate, and polyphenolic concentrations to determine whether changes in foliar chemistry associated with defoliation provide an explanation for these patterns of browsing. In non-defoliated trees, NPK concentrations were highest in young leaves and declined with age, while concentrations of carbohydrates and polyphenolics were independent of leaf age. Nitrogen, phosphorus and polyphenolic concentrations were consistently higher in canopy (sun) versus subcanopy (shade) foliage regardless of leaf age, a trend that was reversed for potassium. Partial (50%) defoliation had little effect on foliar chemistry, regardless of its timing. Total (100%) defoliation stimulated NPK concentrations and depressed condensed tannin concentrations of new foliage produced by the surviving shoots. These results suggest that brushtail possums may focus their feeding on only a few trees because of nutritional changes to leaves as a result of browsing.     

Within-tree variability of mountain birch leaves causes variation in performance for Epirrita autumnata larvae

Within-tree variability in leaf characteristics of the mountain birch (Betula pubescens ssp. tortuosa) was bioassayed for the autumnal moth (Epirrita autumnata) by rearing larvae on birch leaves in a laboratory and measuring their growth, consumption rate, approximate digestibility, efficiency of conversion of ingested food and efficiency of conversion of digested food. Only short shoot leaves, i.e. leaves of the same age, were used. The highest hierarchical level, which included trees and ramets within trees, accounted for most of the total variance in almost all the measured traits. Short shoots (within branches) accounted for more of the variance than branches (within trees/ramets) in most of the traits. The results suggest that differences in leaf quality were reflected in larval growth mainly by differences in food utilization efficiencies (postingestive effects) and less by differences in consumption rate (preingestive effects). The observed within-tree variation is probably a consequence of the modular structure, sectoriality and partial functional independence of tree parts.     

Host tree architecture mediates the effect of predators on herbivore survival

Abstract.  1. Vegetation structural complexity is an important factor influencing ecological interactions between different trophic levels. In order to investigate relationships between the architecture of trees, the presence of arthropod predators, and survival and parasitism of the autumnal moth Epirrita autumnata Borkhausen, two sets of experiments were conducted.
2. In one experiment, the architectural complexity of mountain birch was manipulated to separate the effects of plant structure and age. In the other experiment the trees were left intact, but chosen to represent varying degrees of natural complexity. Young autumnal moth larvae were placed on the trees and their survival was monitored during the larval period.
3. The larvae survived longer in more complex trees if predation by ants was prevented with a glue ring, whereas in control trees smaller canopy size improved survival times in one experiment. The density of ants observed in the trees was not affected by canopy size but spider density was higher on smaller trees. The effect of canopy structure on larval parasitism was weak; larger canopy size decreased parasitism only in one year. Until the fourth instar the larvae travelled shorter distances in trees with reduced branchiness than in trees with reduced foliage or control treatments. Canopy structure manipulation by pruning did not alter the quality of leaves as food for larvae.
4. The effect of canopy structure on herbivore survival may depend on natural enemy abundance and foraging strategy. In complex canopies herbivores are probably better able to escape predation by ambushing spiders but not by actively searching ants.     

Recovery of a Betula pubescens forest in northern Sweden after severe defoliation by Epirrita autumnata

Abstract. Mountain birch (Betula pubescens ssp. czerepanovii) forest in the Abisko valley of northern Sweden was completely defoliated by Epirrita autumnata caterpillars during an outbreak in 1954–1955. The defoliation resulted in an 80–90% mortality of the leaf‐carrying shoots of birches in 1956 and triggered a rejuvenation of stands. The subsequent regrowth of foliage was studied in two damaged birch stands and in one unattacked stand. The number of leaves approximately doubled in the damaged stands between 1961 and 1987, while the number on the reference plot fluctuated without significant increase. Regrowth started with increased production of long shoots from surviving shoots and basal sprouts. Basal sprouts were a substantial source of new shoots in the recovery of the foliage, especially on the most damaged plot. Trees of seed origin constituted a minor fraction of the regrowth. Initial rapid growth of foliage reduced gradually and the annual leaf production in 1986/1987 was 75% of that of the reference plot. Comparison between the recovery curve and data from the reference plot indicates that the shoot population of the damaged forest will, after more than 30 years, need many more years to reach the assumed size of a mature forest. The degree of rejuvenation varied between stands, with different consequences for future dynamics of E. autumnata populations.     

Delayed local responses of downy birch to damage by leafminers and leafrollers

Recent findings suggest that impacts of endemic herbivory on forest ecosystems over the long term may exceed impacts of herbivore outbreaks. However, responses of trees to minor and local damage imposed by small arthropod herbivores, especially by those mining or skeletonising individual leaves, remain poorly understood. We studied the delayed effects of injuries by several leafmining and leafrolling insects on the performance of downy birch shoots. Insect feeding did not affect survival of shoots or survival of individual axillary buds in long shoots. In the year following the damage, shoots produced an average of 13.8% more biomass than undamaged shoots of the same tree. The magnitude of this effect increased with an increase in the leaf area injured during the previous year, but it did not differ among four localities in subarctic and boreo‐nemoral forests, between herbivore feeding guilds, or among herbivores imposing damage in early, mid and late summer. We also found that herbivores attacked the next‐year foliage produced by damaged shoots less frequently than they attacked the next‐year foliage produced by undamaged shoots of the same tree. Thus, our study demonstrated delayed local compensatory growth and increased antiherbivore defence in downy birch shoots following local damage by insect feeding. We suggest that this pattern reflects evolutionary adaptations of plants to permanently acting minor, dispersed and spatially unpredictable damage imposed by endemic herbivory. Local responses are less costly and represent a more sustainable strategy to maintain plant fitness under low levels of herbivory than constitutive resistance or systemic responses.     

Foliage value,apparency and defence investment in birch seedlings and trees

Summary Two factors determining plant anti-herbivore defence investment fitness loss due to herbivory and the probability of herbivory occurring in the field were quantified for birch seedlings and trees. Fitness loss due to defoliation (assumed to be related to loss of growth increment compared to controls) appeared to be greater in seedlings compared to trees, but the result was equivocal. In contrast, seedling foliage at the field site — a typical habitat for birch — suffered much less natural defoliation than tree foliage, suggesting that seedlings are markedly less apparent to most birch herbivores than trees. This low apparency should result in lower investment in anti-herbivore defences by seedlings compared to trees — and being a strong effect, should outweigh the possibly greater growth loss suffered by seedlings, which in isolation would tend to increase their optimum defence investment compared to trees. This prediction was tested using palatability trials with a wide range of common birch herbivores and by direct quantification of anti-herbivore defences. Problems and assumptions inherent in these approaches are discussed, but it seems that birch seedlings are genuinely unapparent to herbivores, and consequently do not need the degree of defence investment required by trees.     

Previous herbivore attack of red alder may improve food quality for fall webworm larvae

Summary Three hypotheses of insect-plant interactions were tested by rearing fall webworm larvae in the laboratory on foliage from red alder trees with different histories of western tent caterpillar herbivory. Fall webworm larvae raised on foliage from trees which had been attacked previously for two summers by moderate densities of western tent caterpillars grew faster and attained heavier pupal weights than did those fed foliage from unattacked trees. This contradicts the hypothesis that moderate levels of previous herbivory induces the production of plant defensive chemicals in red alders. Growth of webworms, when fed foliage from unattacked trees adjacent to alders that were attacked by fall webworm larvae, was the same as when fed foliage from trees isolated by distance from attacked trees. This contradicts the hypothesis that attacked trees stimulate the production of defensive chemicals in neigh-boring trees. Young and mature alder foliage was equally good for fall webworm growth and survival, and foliage from trees heavily attacked by both fall webworm and western tent caterpillars for three years produced slow growth rates and small pupal sizes. This supports the hypothesis that continued heavy insect attack can cause the deterioration of the food quality of attacked trees.     

Phenolic and phenolic-related factors as determinants of suitability of mountain birch leaves to an herbivorous insect

We investigated the role of phenolic and phenolic-related traits of the leaves of mountain birch (Betula pubescens ssp. czerepanovii) as determinants of their suitability for the growth of larvae of the geometrid Epirrita autumnata. As parameters of leaf suitability, we determined the contents of total phenolics, gallotannins, soluble and cell-wall-bound proanthocyanidins (PAS and PAB, respectively), lignin, protein precipitation capacity of tannins (PPC), and leaf toughness. In addition, we examined concentrations of soluble carbohydrates and protein-bound amino acids as background variables describing the nutritive value of leaves. The correlation of the leaf traits of our 40 study trees with the tree-specific relative growth rate (RGR) of E. autumnata showed that the only significant correlation with RGR was that of PAS - the largest fraction of total phenolics - and even that explained only 15% of the variation in E. autumnata growth. The nonlinear estimation of the relationship between RGR and PAS by piecewise linear regression divided the 40 study trees into two groups: (i) 19 trees with good leaves for E. autumnata (RGR ranging from 0.301 to 0.390), and (ii) 21 trees with poor leaves (RGR ranging from 0.196 to 0.296). The suitability of leaves within these two groups of trees was determined by different phenolic traits. Within the good group, the suitability of leaves for larvae was determined by the PPC of extracts, which strongly correlated with gallotannins, and by the total content of gallotannins. In contrast, the leaves of poor trees had significantly higher contents of both PAS and PAB, but leaf toughness correlated only negatively with the RGR of E. autumnata larvae. We also discuss the causes of variation in the phenolic and phenolic-related factors that determine the suitability of leaves for E. autumnata larvae in different groups of trees.     

Spruce budworm growth,development and food utilization on young and old balsam fir trees

Laboratory rearing of spruce budworm, Choristoneura fumiferana, in conjunction with field rearing, gravimetric analyses, a transfer experiment, and foliage chemical analyses at six dates during the period of budworm feeding activity indicated that the age of balsam fir, Abies balsamea, trees (70-year-old mature trees or 30-year-old juvenile trees) affected tree suitability for the spruce budworm via the chemical profile of the foliage. Insects reared on old trees had greater survival and pupal weight, shorter development times, and caused more defoliation than those reared on young trees. Young trees were more suitable for the development of young larvae (instars 2–5), while old trees were more suitable for the development of older, sixth-instar larvae. These results were confirmed by the laboratory transfer experiment. Young larvae fed foliage from young trees had higher relative growth rates (RGR), digestibility (AD), and efficiency of conversion of ingested foliage (ECI) than those fed foliage from old trees. These differences appeared to be related to the high N:tannins ratio, and the high contents of P present in young trees during the development of the young larvae. Old larvae fed foliage from old trees had higher relative growth rates, relative consumption rates (RCR), and digestibility of the foliage than those fed foliage from young trees. The high digestibility of the foliage of old trees was compensated for by a lower efficiency of conversion of digested food (ECD), which in turn resulted in no significant effect of tree age on the efficiency of conversion of ingested foliage by old larvae. The low relative consumption rate of old larvae fed foliage from young trees appeared to be related to the low N:tannins ratio, and the high contents of bornyl acetate, terpinolene, and °-3-carene present in young trees during the budworm sixth instar. Variations in these compounds in relation to tree age may serve as mechanisms of balsam fir resistance to spruce budworm by reducing the feeding rate of sixth instar larvae.