QUANTIFYING NATURAL SELECTION ON BODY SIZE FROM FIELD DATA: WINTER MORTALITY IN MENIDIA MENIDIA

Abstract

Analyses of changes in size distributions over time frequently suggest that mortality rates depend on body size. Such observations, however, are probably confounded with changes in size due to growth. We describe a parametric method by which the size dependence of both mortality and growth may simultaneously be inferred from pairs of size distributions collected at different times. The method is tested in a Monte Carlo study and found to have sampling properties similar to those of other methods that require more data. Survival of the first winter of life appears to be size dependent in a diverse array of taxa, although few prior studies have accounted for growth. Analysis of sizes of Menidia menidia (Atlantic silverside) from three different latitudes revealed that changes in size distributions through winter resulted from growth in southern populations and mortality in the north. Winter mortality was better described by a power function of size than an exponential. Allometric exponents increased with latitude and were greater than predicted from metabolic- or starvation-based models. The steepness of the estimated survivorship curves imply that winter mortality in M. menidia is more consistent with a threshold effect perhaps attributable to offshore migration.