Phenotypic variation in cold tolerance of an invasive insect (Epiphyas postvittana Walker): Implications for forecasting risk

Abstract

Predicting the spread and potential impacts of biological invasions is vital to management and policy decision-making. For much of North America, cold temperatures dictate portions of the geographic boundaries of ectotherm survival and are often included as a key component of risk models and maps for invasive insects. However, current risk assessments typically assume an invading species' response to cold and other environmental stressors is immutable over short to medium time horizons, in part due to a paucity of relevant biological data. Such assumptions may lead to mis-estimated probabilities of establishment and spread of invasive insects, such as a recent invader to the U.S.A, the light brown apple moth (Epiphyas postvittana Walker). My research examined multiple sources of potential variation in cold tolerance of E. postvittana, a species that is predominantly freeze avoidant but also shows some amount of partial freeze tolerance in its overwintering stage. First, I evaluated the effects of larval host plant on direct tolerance to acute partial-freezing in late instars. There was a nearly ten-fold change in survival after partial freezing based on larval host, as well as changes in mean supercooling points. Importantly, the degree of cold tolerance could be predicted by the developmental suitability of a given host. Second, I compared the fitness tradeoffs following exposure to two subzero temperatures (-10°C or the supercooling point) in late instars and pupae. Both lifestages could survive brief exposure to -10°C without significant fitness cost, and displayed small proportions of partial freeze tolerance. However, individuals that survived partial-freezing did so with a reproductive fitness cost. Lastly, I conducted a long-term laboratory selection experiment to evaluate the potential for adaptation to partial freezing in late instars. While there did not appear to be sufficient additive genetic variance to see a consistent response to selection, there was considerable variability through time in survival after partial-freezing. Importantly, this variability cycled in a regular pattern (~six generations), driven by a yet unknown mechanism. The variation in E. postvittana cold response seen in this study resulted in different forecasts of geographic distribution, which can have important management and regulatory implications.