Defoliating insect immune defense interacts with induced plant defense during a population outbreak.

During population outbreaks, top-down and bottom-up factors are unable to control defoliator numbers. To our knowledge, details of biotic interactions leading to increased population density have not been studied during real population outbreaks. We experimentally assessed the strength of plant defenses and of insect immunocompetence, assumed to contribute to active insect resistance against parasitoids and pathogens, in the geometrid Epirrita autumnata during a steep increase in population density. We demonstrated rapid (same-season) induced resistance in the foliage of its host, mountain birch. The response was systemic, spreading throughout the tree, and retarded larval growth rate by approximately 10%. On the other hand, no direct delayed carry-over effects were found in the next season in larval growth rate, mortality, or pupal mass. Larval damage to a tree during the previous year, however, significantly (by approximately 13%) accelerated the advance of the immune response (measured as melanization of an implant inserted into the pupal hemocoel). The encapsulation rate correlated positively with larval mortality in trees in which larvae had been introduced the previous year, but not in control trees. Both of these observations suggest that induced plant defense was associated with an increased insect immunocompetence during the population increase.

Rapid wound-induced resistance in white birch (Betula pubescens) foliage to the geometrid Epirrita autumnata: a comparison of trees and moths within and outside the outbreak range of the moth.

Two strains of a geometrid defoliator, Epirrita autumnata, were used in bioassays to test existence and relative efficacy of rapid, wound-induced foliage resistance in two provenances of the white birch. One birch and one moth strain originated in the outbreak range of the moth and another outside it. Both birch provenances responded to manual leaf damage by changes in foliage quality which significantly retarded growth of the insects, reducing their pupal weights and protracting larval periods. Leaves which were previously damaged were lower quality as Epirrita food than adjacent intact leaves. Both of them were lower quality than intact leaves without damaged leaves nearby. Because of variance between years in the efficacy of the response, and because of different transfer distances of the provenances to the common garden where the experiments were performed, we could not ascertain whether there is any overall difference in the efficacy of rapid inducible responses between the provenances. Both moth strains were affected by wound-induced deterioration in foliage quality. There were no differences in how the moth strains experienced inducible resistance in the two birch provenances. Moths achieved relatively higher pupal weights on the birch provenance matching their origin. Moths from the outbreak range completed their larval period in a shorter time and pupated in a smaller size and, due to dependence of fecundity on size, had a lower potential rate of increase than insects outside the outbreak range.

Life-history evolution in Anodonta piscinalis (Mollusca, Pelecypoda) : Correlation of parameters.

The correlations between certain life-history parameters (reproductive effort, reproductive life-span, age of first reproduction, general growth index, variation in juvenile survival, availability of resources) were studied in 13 populations of the mussel Anodonta piscinalis in south-western Finland in 1975 and 1976Reproductive life-span correlated positively (r s=0,823, P<0.001) with variation in juvenile survival. The average availability of resources correlated negatively both with the reproductive life-span (r s=-0.841, P<0.001 after the variation by juvenile survival had been deleted) and variation in juvenile survival (r s=-0.676, P<0.05).The reproductive effort for female mussels at each site was computed by comparing body weight of reproductive females with body weights of non-reproductive individuals. Availability of resources was much higher in 1976 than in 1975. Consequently, the reproductive effort, an index of the strain of reproduction, was higher in 1975 than in 1976. In 1975 there was a significant correlation between reproductive effort and the length of the reproductive life-span (-0.727, -0.806) and also with the reproductive effort and the variation in juvenile survival (-0.718, -0.758) in females of the length of 60 mm and 70 mm respectively. In 1976, when availability of resources was better, such correlations were not found.Spatial and temporal change in the intake of resources complicates applicability of the principle of resource allocation in the theory of life-history evolution. Studying the mere allocation is not enough if the intake of resources varies in the groups studied. The ratio ovary weight/body weight is a dubious measure of reproductive effort in comparative studies when the input of resources can vary, and this possibility can be ruled out only exceptionally.Correlation between growth and reproductive effort was positive, obviously because both are important components in creating high reproductive capacity. In 1975 reproductive effort increased with size (age). Change in reproductive effort correlated with reproductive life-span (r s=-0.633, P<0.05).The following parameters occurred together: short reproductive life-span, low age of first reproduction, high reproductive effort, rapid growth and high clutch size. They were realized at sites where the availability of resources was good and variation in juvenile survival was low-i.e. the environment was stable. The results conflict with the prediction of the theory of r and K-selection.

Birch leaves as a resource for herbivores: Seasonal occurrence of increased resistance in foliage after mechanical damage of adjacent leaves.

Seasonal occurrence of such wound-induced reaction in birch foliage which deteriorates the quality of nearby leaves for herbivores was tested by means of bioassays. Length of the larval period was protracted in two early and mid-summer (larval period!) lepidopteran species as well as in two mid-summer hymenopteran species when larvae were reared on birch leaves whose adjacent leaves had earlier been damaged mechanically. This response was not found for two late-summer hymenopteran species. In a lepidopteran species whose larval period lasts through the whole season, retardation in growth was significant in the beginning of August but notl later. Hence such response of leaves, interpreted as defensive on the part of the birch, was not efficient after leaves had gained their final size. The potential consequences of wound-induced responses of leaves for herbivores are discussed.

Long-term inducible resistance in birch foliage: triggering cues and efficacy on a defoliator.

Heavy damage of the mountain birch foliage, as well as application of small amounts of insect frass to the soil beneath the trees, reduced growth of Epirrita autumnata larvae reared in these trees in the following year. Foliage damage in the previous year decreased larval survival, too. Both foliage damage and insect frass in the soil decreased a fecundity index which combined the effects of size and survival. Because application of small amounts of fertilizers had an effect indistinguishable from that of insect frass, the effect of the frass may base on responses of trees to an increase in soil nutrient concentration in mid-summer. In previously untreated control trees, all performance indices (growth, survival, and egg production) of Epirrita correlated positively with the distance of the birch from the closest birch defoliated in the previous year, indicating "communication" between adjacent trees. Epirrita egg production in trees that had been both defoliated and treated with frass in the previous summer was at least 70% lower than in previously unmanipulated control trees.

Foliage phenols and nitrogen in relation to growth, insect damage, and ability to recover after defoliation, in the mountain birch Betula pubescens ssp tortuosa.

We studied growth of the mountain birch, and the role of foliage phenols, nitrogen, and variance in the timing of bud burst, as potential defensive characters, in Finnish Lapland in 1975-1979. Annual and local variation both in phenol and nitrogen concentration of foliage were significant. Individual trees retained their position in the foliage and nitrogen distribution of the population in successive years, as well as in the order of leaf flush in spring. Growth of twigs, mature leaf size, and ability of trees to recover in the year following artificial defoliation correlated positively with the sum of degree days in the previous growing season. Foliage nitrogen correlated negatively with foliage phenols in within-site comparisons. Twig growth correlated negatively with foliage phenols, particularly in growing seasons following cool summers, but did not correlate with foliage nitrogen. Birches flushing early did not grow more than birches flushing late. Between-site differences in foliage phenol content were mainly determined by abiotic conditions, like temperature and nutrient availability. In a between-site comparison insect chewing marks in leaves correlated positively with foliage phenols as well as with nitrogen; intensity of invertebrate predation presumably explained variable herbivory between the sites. In a within-site comparison trees with the highest foliage phenol content had few herbivores only at the site with the highest average phenol level.

Nutrient stress: an explanation for plant anti-herbivore responses to defoliation.

A hypothesis is put forward that the long-lasting inducible responses of trees to herbivores, particularly lepidopteran defoliators, may not be active defensive responses, but a by-product of mechanisms which rearrange the plant carbon/nutrient balance in response to nutrient stress caused by defoliation. When defoliation removes the foliage nutrients of trees growing in nutrient-poor soils, it increases nutrient stress wich in turn results in a high production of carbon-based allelochemicals. The excess of carbon that cannot be diverted to growth due to nutrient stress is diverted to the production of plant secondary metabolites. The level of carbon-based secondary substances decays gradually depending on the rate at which nutrient stress is relaxed after defoliation. In nutrient-poor soils and in plant species with slow compensatory nutrient uptake rates the responses induced by defoliation can have relaxation times of several years. The changes in leaf nitrogen and phenolic content of mountain birch support this nutrient stress hypothesis. Defoliation reduces leaf nitrogen content while phenolic content increases. These responses of mountain birch to defoliation are relaxed within 3-4 years.

Consequences of herbivory in the mountain birch (Betula pubescens ssp tortuosa): importance of the functional organization of the tree.

Three types of experiments indicate that the functional organization of the mountain birch may influence the ways in which the tree responds to simulated or natural herbivory. The first experiment showed that herbivory to both short and long shoot leaves affects plant development but, because growth largely proceeds by resources of the previous year, is manifested only in the year following the damage. The second experiment showed that even partial damage to a single long shoot leaf caused the axillary bud of that leaf to produce a shorter shoot the next year. Therefore, the value of a leaf depends also on the organ which it is subtending. In the third experiment we manipulated the apical dominance of shoots in ramets and caused improvement to leaf quality in extant shoots. Ramets within a tree responded individually, probably mediated by disturbance of the hormonal control because removal of apical buds elicited the response although removal of the same number of basal buds did not. Induced amelioration is a different response to induced resistance. The two responses are triggered by different cues and may occur in the same plant. By altering hormonal balance of shoots it is potentially possible for herbivores to induce amelioration of food quality. The ways in which herbivory is simulated may explain variability of results obtained when herbivory-induced responses in plants have been studied.

Delayed induced resistance and increase in leaf fluctuating asymmetry as responses of Salix borealis to insect herbivory.

An outbreak of leaf beetle Melasoma lapponica in two localities around the Severonikel smelter in Kola Peninsula, north-west Russia, resulted in severe defoliation of Salix borealis, observed for the first time in August 1993 and then again in 1994 and 1995. Before the first severe defoliation, in July 1993, performance of M. lapponica larvae in plots with a high beetle density was either better or the same as in low-density plots. However, in 1994 and 1995, the years following severe willow defoliation in high-density plots, M. lapponica performance (in terms of survival, developmental time and beetle weight) decreased with increasing beetle density. Retarded larval growth in high-density plots was related to a decreased consumption rate, whereas the efficiency of the conversion of ingested food was similar in high- and low-density plots. These results indicate that defoliation triggered delayed induced resistance in S. borealis. Leaf fluctuating asymmetry (FA, a non-specific stress indicator) of this willow species in 1992 was similar in low- and high-density plots, but it increased in high-density plots in 1994, at the same time that detrimental effects on beetle performance were recorded at these sites. Plot-specific indices of beetle performance and FA were negatively correlated both in 1994 and 1995, suggesting that plants stressed by defoliation the previous-year were less favourable for leaf beetles.

Crowding-triggered phenotypic responses alleviate consequences of crowding inEpirrita autumnata (Lep., Geometridae).

CrowdedEpirrita larvae had shorter larval periods than, and similar pupal masses to, their solitary siblings when reared on low quality diets. When fed on high quality diets, pupal masses of crowded larvae were lower than in singletons, and there was no difference in larval period. Because changes in food availability (absolute shortage, induced resistance in foliage) are caused by high larval densities in the field, crowding-triggered phenotypic changes may helpEpirrita to overcome detrimental consequences of high larval density. Pupal period was longer in crowded larvae than in singletons and crowded adults emerged later than their solitary siblings. Eggs of late emerging moths eclosed late in the ensuing spring, which coincides with delayed leaf flush in the year after defoliation. The reason for the faster growth of crowded individuals on poor diets was higher intake albeit less thorough processing of food in crowded, but not in solitary, larvae. On good diets solitary individuals tended to consume more than crowded larvae but there was no difference in processing. Predicted differences of host plant use between stealthy and opportunistic types of herbivores (sensu Rhoades 1985) were generally found between solitary and aggregated larvae on poor but not on good diets. The group response could not be explained by benefits to the group although the assumptions of Wilson's model of group selection were satisfied.

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.

Seasonal changes in birch leaf chemistry: are there trade-offs between leaf growth and accumulation of phenolics?

Several plant-herbivore hypotheses are based on the assumption that plants cannot simultaneously allocate resources to growth and defence. We studied seasonal patterns in allocation to growth and putatively defensive compounds by monitoring several chemical and physical traits in the leaves of mountain birch from early June (budburst) to late September (leaf senescence). We found significant seasonal changes in all measured characteristics, both in terms of concentrations (mg g-1) and amounts (mg leaf-1). Changes were very rapid in the spring, slow in the middle of the season, and there was another period of fast changes in the senescing leaves. Co-occurring changes in physical leaf traits and concentrations of several compounds indicated a seasonal decline in foliage suitability for herbivores. Concentrations of protein and free amino acids declined through the growing season whereas individual sugars showed variable seasonal patterns. The seasonal trends of phenolic groups differed drastically: concentrations of soluble proanthocyanidins increased through the season, whereas cell wall-bound proanthocyanidins, gallotannins and flavonoid glycosides declined after an initial increase in young leaves. We failed to find proof that the seasonal accumulation of phenolics would have been seriously compromised by leaf or shoot growth, as assumed by the growth/differentiation balance hypothesis and the protein competition model hypothesis. On the contrary, there was a steady increase in the total amount of phenolics per leaf even during the most active leaf growth.

Multiplicity of biochemical factors determining quality of growing birch leaves.

Due to rapidly changing physical and biochemical characteristics of growing leaves, correlations between traits of foliage biochemistry and the performance indices of flush feeding herbivores may vary considerably following relatively minor changes in experimental conditions. We examined the effects of the seasonal and inter-tree variation of a comprehensive array of biochemical compounds on the success of an early season geometrid, Epirrita autumnata, feeding on maturing foliage of mountain birch, Betula pubescens ssp. czerepanovii. We monitored the concentrations of individual phenolics, sugars, total nitrogen, nitrogen of proteins, and nitrogen of soluble compounds, water and acetone-insoluble residue. Simultaneously we recorded larval consumption, physiological performance, growth, and pupal mass of E. autumnata. We found significant phenological changes in almost all leaf traits measured. In bioassays with half-grown leaves, leaf gallotannin concentrations showed a nonlinear effect: in trees with high foliar gallotannin concentrations (over 10 mg g-1), physiological performance was strongly reduced by high gallotannin concentrations. In trees with lower gallotannin concentrations, on the other hand, larval growth was reduced by soluble proanthocyanidins, not gallotannins. Differences between high and low gallotannin trees largely depended on phenology, i.e., on the age of leaves. However, not all the differences in leaf traits between late (with high gallotannin concentrations at the time of the bioassay) and early flushing trees disappeared with leaf maturation, indicating that there is also phenology-independent variance in the tree population. In the full-grown leaves of all the study trees, low concentrations of water and of nitrogen of proteins (but not nitrogen of soluble compounds) were the main factors reducing pupal masses of E. autumnata, while neither gallotannin nor proanthocyanidins now played a significant role. The observed change in the factors underlying leaf quality (from gallotannins and proanthocyanidins to nitrogen and water) relate to the activity of the shikimate pathway and the formation of cell walls: gallotannins and proanthocyanidins are both produced in the pathway, and these tannins are assumed to contribute - via binding into cell walls - to tough and durable cell walls. Interestingly, low quality of leaves did not automatically translate into low foliar consumption (i.e., benefits to the tree). On the trees with young, high gallotannin leaves, larvae actually increased consumption on low quality foliage. In the group of trees with slightly more developed, low gallotannin leaves, the quality of leaves did not clearly modify amounts consumed. In full-grown leaves, low leaf quality strongly reduced leaf consumption. These results emphasize the strong influence of tree phenology on the relationships between biochemical compounds and the herbivore.

Metabolic modifications of birch leaf phenolics by an herbivorous insect: detoxification of flavonoid aglycones via glycosylation.

The metabolic modifications of birch (Betula pubescens Ehrh.) leaf phenolics in the digestive tract of its major defoliator, larvae of the autumnal moth Epirrita autumnata, were studied. The main phenolic acids of birch, i.e. chlorogenic and p-coumaroylquinic acids, were isomerised in the alkaline digestive tract. Moreover, only 16 to 92% of the ingested amounts of chlorogenic acid were found in the faeces of individual larvae; the missing portion is possibly being used in the formation of reactive o-quinones. Water-soluble flavonoid glycosides were mostly excreted unaltered. In contrast, lipophilic flavonoid aglycones were not excreted as such, but as glycosides after being detoxified by E. autumnata via glycosylation. When the larvae were fed with leaf-painted acacetin and kaempferide, i.e. two naturally occurring birch leaf flavonoid aglycones, acacetin-7-O-glucoside and kaempferide-3-O-glucoside appeared in larval faeces as major metabolites. However, the efficiency of aglycone glycosylation varied-, ranging from 17 to 33%, depending on the aglycone and its dietary level. There was also large variation in the efficiency of glycosylation--from 2 to 57%--among individual larvae. These results demonstrate a compound-specific metabolism of phenolic compounds, leading to different phenolic profiles in the insect gut compared to its leaf diet.

Foliar oxidases as mediators of the rapidly induced resistance of mountain birch against Epirrita autumnata.

Induced resistance of the mountain birch against its main defoliator Epirrita autumnata is a well-characterized phenomenon. The causal mechanism for this induced deterioration, however, has not been unequivocally explained, and no individual compound or group of traditional defensive compounds has been shown to explain the phenomenon. Phenolic compounds are the main secondary metabolites in mountain birch leaves, and the biological activity of phenolics usually depends on their oxidation. In this study, we found that the activity of polyphenoloxidases (PPOs), enzymes that oxidize o-diphenols to o-diquinones, was induced in trees with introduced larvae, and bioassays showed that both growth and consumption rates of larvae were reduced in damaged trees. PPO activity was negatively associated with both larval growth and consumption rates in trees with bagged larvae, but not in control trees. Our results suggest that the oxidation of phenolics by PPOs may be a causal explanation for the rapidly induced resistance of mountain birch against E. autumnata. This finding also helps to explain why correlations between insect performance and phenolics (without measuring indices explaining their oxidation) may not produce consistent results.

Foliar phenolics are differently associated with Epirrita autumnata growth and immunocompetence.

The quality of available food may affect insect herbivores directly (via growth and survivorship) and/or indirectly (by modifying insect vulnerability to parasitoids and pathogens). We examined the relationship between different phenolic compounds, belonging to various phenolic groups, in Betula pubescens spp. czerepanovii (mountain birch) foliage and the larval performance of the geometrid Epirrita autumnata (autumnal moth). Direct effects on insect performance were described by pupal weight, developmental rate, and survivorship; indirect effects were described by the encapsulation rate of an implant inserted into the insect hemocoel, a commonly used way to describe insect immune defense. We found profound differences in the effects of different phenolic categories: several individual hydrolyzable tannins were associated positively with larval performance but negatively with level of immune defense, whereas flavonoid glycosides were inversely related to larval survival but showed no association with the larvae immune defense.

Quantitative analysis of polymeric proanthocyanidins in birch leaves with normal-phase HPLC.

The proanthocyanidin composition and content in the leaves of nine birch species (Betula albosinensis, B. ermanii B. maximowicziana, B. nana, B. papyrifera, B. pendula, B. platyphylla, B. pubescens, and B. pubescens ssp. czerepanovii) were studied with different methods including colorimetric assay, HPLC coupled with PAD or ESI/MS and NMR. Total proanthocyanidin content was determined using the acid butanol assay. A normal phase-HPLC method was applied for the analysis of polymeric proanthocyanidins. The content of polymeric proanthocyanidins was estimated from a late eluting peak in the chromatogram. With this HPLC method, quantitative analysis of polymeric proanthocyanidins could be performed directly from leaf extracts: no additional purification or preparation steps were required. It was shown that birch leaves contained mainly polymeric proanthocyanidins with a degree of polymerisation greater than 10. Total proanthocyanidin content (expressed as dry weight) was found to vary from 44mg/g (B. papyrifera) to 145mg/g (B. nana), and polymeric proanthocyanidin content from 39mg/g (B. pendula) to 119 mg/g (B. nana).

Delayed induced responses of birch glandular trichomes and leaf surface lipophilic compounds to mechanical defoliation and simulated winter browsing.

Changes in morphology and chemistry of leaf surface in response to herbivore damage may increase plant resistance to subsequent herbivore attack; however, there is lack of studies on induced responses of glandular trichomes and their exudates in woody plants and on effects of these changes on herbivores. We studied delayed induced responses in leaf surface traits of five clones of silver birch (Betula pendula Roth) subjected to various types of mechanical defoliation and simulated winter browsing. Glandular trichome density and concentrations of the majority of surface lipophilic compounds increased in trees defoliated during the previous summer. This induced response was systemic, since control branches in branch defoliated trees responded to the treatments similarly to defoliated branches, but differently from control trees. In contrast to defoliation treatments, simulated winter browsing reduced glandular trichome density on the following summer and had fewer effects on individual surface lipophilic compounds. Moreover, constitutive density of glandular trichomes was negatively correlated with induced total amount of lipophilic compounds per trichome, indicating a trade-off between constitutive and induced resistance in silver birch. Induced changes in leaf surface traits had no significant effect on leaf damage by chewers, miners and gall mites, but increased susceptibility of birch trees to aphids. However, leaf damage by chewers, miners and gall mites in defoliated (but not in control) trees was correlated with concentrations of some fatty acids and triterpenoids, although the direction of relationships varied among herbivore species. This indicates that induction of surface lipophilic compounds may influence birch resistance to herbivores. Our study thus demonstrated both specificity of elicitation of induced responses of birch leaf surface traits by different types of damage and specificity of the effects of these responses on different types of herbivores.

Putting the insect into the birch-insect interaction.

Leaf maturation in mountain birch (Betula pubescens ssp. czerepanovii) is characterized by rapid shifts in the types of dominant phenolics: from carbon-economic flavonoids aglycons in flushing leaves, via hydrolysable tannins and flavonoid glycosides, to carbon-rich proanthocyanidins (condensed tannins) in mature foliage. This shift accords with the suggested trade-offs between carbon allocation to plant defense and growth, but may also relate to the simultaneous decline in nutritive leaf traits, such as water, proteins and sugars, which potentially limit insect growth. To elucidate how birch leaf quality translates into insect growth, I introduce a simple model that takes into account defensive compounds but also acknowledges insect demand for nutritive compounds. The effects of defensive compounds on insect growth depend strongly on background variation in nutritive leaf traits: compensatory feeding on low nutritive diets increases the intake of defensive compounds, and the availability of growth-limiting nutritive compounds may modify the effects of defenses. The ratio of consumption to larval growth (both in dry mass) increases very rapidly with leaf maturation: from 2.9 to 9.8 over 2 weeks in June-July, and to 15 by August. High concentrations in mature birch leaves of "quantitative" defenses, such as proanthocyanidins (15-20% of dry mass), presumably prevent further consumption. If the same compounds had also protected half-grown leaves (which supported the same larval growth with only one third of the dry matter consumption of older leaves), the same intake of proanthocyanidins would have demanded improbably high concentrations (close to 50%) in young leaves. The model thus suggests an adaptive explanation for the high levels of "quantitative" defenses, such as proanthocyanidins, in low-nutritive but not in high-nutritive leaves because of the behavioral responses of insect feeding to leaf nutritive levels.