For studies of plant-herbivore interactions, hybrid zones represent natural experiments in which genetic-based defenses have become scrambled through backcrossing. Studies of insect pests of cottonwoods (Populus) show that pests are concentrated in hybrid zones and, within the zone, on backcrossed trees. Pests from river systems without hybrid zones perform better on the parent tree species than do pests from areas with hybrid zones. These data suggest that highly susceptible hosts (i.e., naturally occurring complex hybrids) not only support most of the pest population, their presence may result in pest species either losing or never evolving the ability to deal with the greater resistance of the parental tree species. An implication of this work is that susceptible hybrids can be used to manage pest dynamics in forestry and agriculture. Because of the basic and applied implications of this topic, this project will examine the generality of this phenomenon (the Sink Hypothesis) in a second system, Eucalyptus, in which hybridization is common. The researchers will collaborate with eucalypt systematists, geneticists, and ecologists to study pest dynamics in several well known natural hybrid systems. The following questions will be addressed: 1. Are hybrid eucalypts sinks for herbivores such that a small proportion of potential hosts support most of the pest population? 2. If complex hybrids are more susceptible to one pest, are they also more susceptible to very different insect guilds? 3. Does the presence or absence of hybrid zones affect pest densities in the pure host zone? These questions will be addressed in a natural system, common gardens with trees of known genotypes, and by documenting insect attacks on herbarium specimens of parents and hybrids collected independent of the study. Answers to these questions have not been examined for any system besides cottonwoods and should be fundamentally important in understanding the ecological and evolutionary interactions of plants and pests.
