Resolving within- and between-population variation in feeding ecology with a biomechanical model.

Studies of phenotypic plasticity have emphasized the effect of the environment on the phenotype, but plasticity can also be used as a tool to study the functional significance of key traits. By inducing variation in phenotypes and testing quantitative models that predict performance based on biological mechanisms, we can develop functionally general models of performance. Pumpkinseed sunfish from lakes with high snail availability have large levator posterior muscles (which are used to crush snail shells), whereas fish from lakes with few snails have relatively small muscles. Here we: (1) quantify differences in the feeding ability of an ontogenetic series of pumpkinseed from two populations; and (2) evaluate whether a biomechanical model can resolve the observed ontogenetic and between-population variation in feeding ecology. Mass, but not length, of the levator posterior muscle in fish from Three Lakes (a lake rich in snails) was greater than for comparably sized fish from Wintergreen Lake (a lake with few snails). Handling times were shorter, crushing strengths were 71% greater, and foraging rate (snail tissue mass consumed per time) and the fraction of thick-shelled snails in the diet were approximately 100% greater for fish from Three Lakes compared to comparably sized fish from Wintergreen. These between-lake differences were not significant after adjusting for variation in pharyngeal morphology, suggesting that the biomechanical model of snail crushing resolved observed ontogenetic and population-level variation in the feeding ecology of pumpkinseed.