Geographic mosaic in host-plant use: Regional variability in host preference and tri-trophic performance of the fall webworm

Background/Question/Methods Host plant choice by phytophagous insects may be driven by factors including host abundance, host quality, and the degree to which host plants provide enemy free space. These factors may also present herbivores with fitness trade-offs among hosts, promoting dietary generalism and preferences that maximize different fitness components on different hosts. Because selection pressures vary across a species' range, patterns of host use may also vary. The fall webworm (Hyphantria cunea) is a dietary generalist that feeds and constructs webs on a broad array of trees across its large geographic range. Here, we explore patterns of host tree use between Connecticut and Maryland fall webworm populations. Vegetation surveys adjacent to fall webworm webs were carried out at several sites in each state to assess host use, and caterpillars collected from these areas were reared in the laboratory to measure host-associated variation in performance and parasitism frequency. Results/Conclusions For both localities it was found that host use was non-random with respect to tree identity, and that among potential hosts, the main predictor of host use was abundance. Because there was little overlap between host trees between CT and MD sites, it was not possible to evaluate differences in host preference between the two areas. Significant variation in caterpillar growth and development on selected hosts was observed in both Maryland and Connecticut, but growth performance on a given host plant was not associated with its frequency of use by the fall webworm. In Connecticut, parasitism frequency was also measured on various host plants, revealing that host plants vary in the degree of enemy free space they afford caterpillars. However, parasitism frequency also failed to predict host use. No tradeoff between host quality and enemy free space was identified in Connecticut, nor was a tradeoff between host plant quality and availability identified in Maryland. However, the Connecticut population shows a trade off between host quality and host availability. Possible causes for this disparity amid otherwise similar host use patterns are discussed.