As the prey thickens: rainbow trout select prey based upon width not length

Abstract

Drift-feeding fish are typically considered size-selective predators. Yet, few studies have explicitly tested which aspect of prey “size” best explains size selection by drift-foraging fish. Here, we develop a Bayesian discrete choice model to evaluate how attributes of both prey and predator simultaneously influence size-selective foraging. We apply the model to a large dataset of paired invertebrate drift (n = 784) and rainbow trout (Oncorhynchus mykiss) diets (n = 1028). We characterized prey “size” using six metrics (length, width, area, hemispherical area, volume, mass) and used pseudo-R2 to determine which metric best explained observed prey selection across seven taxa. We found that rainbow trout are positively size-selective, they are selecting prey based upon differences in prey width, and size-selectivity increases with fish length. Rainbow trout demonstrated strong selection for the adult and pupae stages of aquatic insects relative to their larval stages. Our study provides strong empirical evidence for size-selective foraging in rainbow trout and demonstrates prey selection is based primarily upon width, not length or area as has been widely reported.