Evolutionary ecology of host use in the silver-spotted skipper (Epargyreus clarus)

Tuesday, November 18, 2014: 3:18 PM
D135 (Oregon Convention Center)
John T. Lill , Biological Sciences, George Washington University, Washington, DC
Martha Weiss , Biology, Georgetown University, Washington, DC
Eric Lind , Ecology, Evolution, and Behavior, University of Minnesota, Saint Paul, MN
Developing our understanding of the ecological factors shaping patterns of host plant use in herbivorous insects is a major goal of insect ecologists.  Introductions of exotic plants for human use in agriculture and horticulture present unique opportunities to investigate the dynamics of host range expansion in native insects.  In this talk, we report early results of an ongoing project investigating the evolutionary and behavioral ecology of host selection in a common butterfly, the silver-spotted skipper (Epargyreus clarus).  As caterpillars, these skippers oviposit and feed on a variety of leguminous hosts, which include both native (black locust, hog-peanut, and tick-trefoil) and non-native (Chinese wiseria, soybean, and kudzu)  plants.  Oviposition trials with this skipper indicated differential host selection by adult females and no-choice feeding assays in the laboratory revealed strong host plant effects on multiple measures of caterpillar performance (survival, development time, pupal mass).  Similarly, larval feeding assays revealed consistent preferences of some hosts over in both early and late instars, and some evidence for induced feeding preferences in early instars.  Field trials using bagged caterpillars that were protected from natural enemies were generally consistent with lab performance findings, whereas exposed larvae subject to predation and parasitism indicated some evidence for ecological tradeoffs between bitrophic and tritrophic measures of larval fitness.  The consistency of these tradeoffs among generations and years will help determine whether host use patterns in this butterfly support optimality theory or instead support bet-hedging models of host use in a temporally unpredictable landscape.