Use of numerical, computer-intensive methods to solve optimization problems in evolutionary ecology

• Philip H. Crowley

Abstract
While at NCEAS, Dr. Philip Crowley will be working on the following two projects: (1) Classifier-system models of reciprocal altruism and of behavioral complementarity. Classifier systems amount to a genetic-algorithm-like technique for evolving solutions to complex problems, and the problems I'm addressing are (a) finding strategies consistent with mutual cooperation and (b) finding strategies that permit complementary behavior such as simultaneous hermaphroditism. But the same basic approach can be used to solve a wide range of numerical optimization problems including many other ecological ones. (2) Theory on gender-specific life-history strategies. I am attempting to develop empirically testable models of optimal time and survival to maturation and size at maturation, where males and females may differ. My students and I are continuing some empirical studies with fish and dragonflies in replicated tubs and ponds in hopes that the models and experiments will prove mutually stimulatory.