Pinniped diets inferred from scats: analysis of biases in prey occurrence

Abstract

The diets of pinnipeds have often been used to study their ecology and resource dynamics and in ecosystem monitoring. Scat analysis is now the most widely used method of inferring the diets of pinnipeds. Using a mathematical model the present study explores the expected biases in prey occurrence that are related to prey mass, proportion of loss of remains, predator assimilation efficiency, size of the scats collected, and meal size. With the exception of meal size, variation in parameters implied variation in the probability of biases in prey occurrence. Biases were asymmetric among prey sizes, but in contrast with previous ideas, the results indicate that small prey have smaller biases than large ones do. Furthermore, the number of scats required to detect a large prey was notably larger than that needed to detect a small prey with the same occurrence. The present study is not limited to a particular pinniped or prey species, so it has the potential to represent a general framework for interpreting the results of scat analysis in pinnipeds. The model complements empirical studies, advancing our comprehension of biases associated with prey occurrence in dietary studies of pinnipeds.