GEOGRAPHIC VARIATION AMONG HERBIVORE POPULATIONS IN TOLERANCE FOR A CHEMICALLY RICH SEAWEED

Previous investigations have shown that the sedentary amphipod Ampithoe longimana escapes consumers by selectively living on and eating chemically defended seaweeds in the genus Dictyota . However, A. longimana and Dictyota overlap only in the southern portion of the amphipod's range; Dictyota is not available to amphipods from more northerly regions. We used this disjunction in distribution to test the hypothesis that A. longimana populations co‐occurring with Dictyota would have greater tolerance for the seaweed's chemical defenses and would display higher feeding preference for, and fitness on, the seaweed than would more northerly populations. We also evaluated the genetic vs. phenotypic basis of these patterns and attempted to detect trade‐offs between tolerance for Dictyota and ability to use other plants as foods. Such geographic studies of herbivory have been conducted using terrestrial insects, but few studies have focused on other herbivores, and this is especially true for marine systems. In multiple‐choice feeding assays with both field‐collected and laboratory‐reared adults, a North Carolina population of A. longimana sympatric with Dictyota more readily fed on Dictyota and was more resistant to Dictyota 's deterrent chemistry than was a Connecticut population from outside of Dictyota's geographic range. When raised on Dictyota menstrualis and D. ciliolata , A. longimana juveniles from North Carolina grew faster, matured faster, and produced more reproductive females than did Connecticut juveniles. The differential tolerance for Dictyota has a genetic basis, as it was maintained through two generations grown to maturity in a common environment. When several northern and southern populations were assayed, they displayed similar regional differences in feeding preferences. Though southern juveniles had higher fitness on Dictyota than northern juveniles, southern juveniles performed as well as northern juveniles when raised on seven other seaweeds, including seaweeds (e.g., Fucus vesiculosus and Sargassum filipendula ) that produce secondary metabolites in a different class from those found in Dictyota . Thus, tolerating Dictyota did not incur detectable performance trade‐offs when juveniles were confined to feeding on alternative seaweeds. Our results suggest that the evolution of host preferences may depend more on the host value as a refuge from enemies than on minimizing the costs of tolerating plant secondary metabolites.