BERKE, S. K.; MILLER, M.; WOODIN, S. A.; University of South Carolina; University of South Carolina; University of South Carolina: Modeling Behavior through Reproductive Value

Historically, behavioral modeling has been largely limited to simulations and dynamic programming models, perhaps because it is easier to write �rules� for individual behavior than to write continuous equations describing it. One well-known exception is optimal foraging, which has significantly advanced the field by producing generally applicable predictions (Schoener 1971, Charnov 1976). Models involving continuous equations are frequently more generalizable than IBMs; hence, they are perhaps more likely to generate broad theoretical advances (Grimm, 1999). Here, we present a new continuous technique for modeling behavior using reproductive value. Using a phyletically widespread behavior�self-decorating�as a model system, we describe the costs and benefits associated with this behavior in terms of energetics and mortality. These quantities are related using an equation derived from Fisher�s (1958) reproductive value. The result is a simple, continuous model which captures the fitness consequences of a behavior. This model can generate predictions about situations in which the behavior might be selectively advantageous and helps explain observed variation in the behavior across taxonomic groups. Because reproductive value can be linked to population-level fecundity and survivorship parameters, the potential exists for this technique to bridge the gap between individual behavior and populations.