Interactive effects of leaf maturation and phenolics on consumption and growth of a geometrid moth

Phenolic compounds are commonly regarded as the main chemical defenses of deciduous woody plants against insects. To examine how indices of leaf maturation (water content, toughness, and sugar/protein ratio) modified larval consumption and growth relative to phenolics and phenolic-related leaf traits, we measured consumption and growth of fourth-instar Epirrita autumnata (Bkh.) (Lepidoptera: Geometridae) larvae on three different days on young, normal, and mature leaves, respectively, from the same mountain birch (Betula pubescens ssp. czerepanovii (Orlova) Hämet-Ahti) trees. The larvae achieved the same growth rates on young and normal leaves, but had to consume 40% more on the latter. On more mature leaves, larval growth was poorer and was positively correlated with sugar/protein ratios (although the ratio peaked at that time). Indices of leaf maturation correlated with several phenolics in data pooled over the three study days, but poorly in any individual day. Similarly, in the pooled data, larval consumption and growth correlated with several leaf traits, but correlations between leaf and insect traits were few on any of the three days, and no trait was significant on each of the three days.We next examined whether variation in the maturation indices modified the associations of phenolics with insect consumption and growth. When interactions between phenolics and leaf maturation indices were taken into account, the number of phenolic compounds displaying significant associations with insect traits more than doubled. The relative importance of interactive versus direct associations increased with leaf maturation: on young leaves five phenolics showed direct and eleven interactive associations with insect traits, while in mature leaves we found two phenolics to display direct and thirteen phenolics interactive associations. Leaf water content, either alone or together with toughness and sugar/protein ratio, generally explained more of the variance in Epirrita growth (up to 59%) than any phenolic or phenolic-related trait alone (highest value 20%). Including interactive effects between phenolics and indices of leaf maturation in the model increased the proportion explained of variance in larval growth between 49 and 73%. Maturation indices explained 0 to 23% of variance in consumption, and the phenolic compound with the highest (positive!) correlation alone up to 28%, but taking into account interactions between phenolics and maturation indices raised the degree of explanation much (namely, 32 to 53%) over that explained by indices of leaf maturation alone. This indicates strong interactive effects on consumption between phenolics and indices of leaf maturation.