Oviposition of a leaf-miner on Erythroxylum tortuosum (Erythroxylaceae) leaves: hierarchical variation of physical leaf traits

Abstract

Plants do not offer homogeneous supplies of the resources required by herbivorous insects as many resource traits show considerable variation both within and between plants. The distribution of variation among host-plant attributes determines the optimal spatial resolution level for insect females to select the best resource patches for oviposition. In this study, we examine whether variation in fluctuating asymmetry and size of Erythroxylum tortuosum Mart. (Erythroxylaceae) leaves influence oviposition of the specialist leaf-miner Agnippe Chambers (Lepidoptera: Gelechiidae). By partitioning the variance across five hierarchical levels, the oviposition pattern was investigated at the spatial resolution level where these leaf traits varied the most. We confirm that the largest variation in both these leaf attributes occurs at the leaf level. We hypothesise that leaf-miner females will respond to this variation by selecting the best leaves (resources) on which to lay their eggs. We find that the probability of oviposition is not significantly related to fluctuating asymmetry or to leaf size (oviposition preference test), suggesting that these two physical traits are not relevant to leaf-miners as indicators of resource patch quality. Therefore, although we show that females laid significantly more eggs on larger leaves, this behaviour appears not to be a result of active selection of leaves. Our results suggest that Agnippe females probably adjust their oviposition proportionally to leaf area because the relationship between egg density and leaf area was not statistically significant.