RESUMEN
We used acoustic telemetry and acceleration sensors to compare population-specific measures of the metabolic costs of an apex fish predator living in four separate lakes. We chose our study species and populations to provide a strong test of recent theoretical predictions that optimal foraging by an apex fish predator in a typical aquatic environment would be consistent with feeding to satiation rather than continuous feeding. We chose four populations where the primary prey type differed along a body size gradient (from small invertebrates to large planktivorous fish) and along a thermal accessibility gradient (from easily accessible cold-water pelagic prey to less accessible warm-water epilimnetic and littoral prey). We expected that these gradients in prey type would evoke distinctly different activity gradients depending on whether predators fed to satiation (e.g., less frequent "rest" detections where primary prey are smaller/less accessible) or fed continuously (e.g., fixed level of "rest" detections under all prey conditions). Our study organism was a fall spawning, cold-water visual apex predator (lake trout). Therefore, we focused our study on diel (early night, dawn, day, dusk, late night) changes in metabolic costs associated with summer feeding behaviour. The duration (~20 days) and fine temporal scale (~30 min) of our behavioural data provided a uniquely detailed picture of intra- and inter-population differences in activity patterns over a critical period in the annual growing season. In all populations, diel shifts in activity were qualitatively consistent with that expected of a visual predator (e.g., resting state detections were most frequent at night). Between-lake differences in daytime thermal experience were qualitatively consistent with between-lake differences in the location of primary prey (e.g., excursions to warm habitats were common in lakes with epilimnetic/littoral fish as primary prey and relatively rare in lakes with pelagic cold-water invertebrates/fish as primary prey). Daytime activity patterns were more consistent with the feeding pattern expected from feeding to satiation rather than continuous feeding: (a) individuals in all four populations exhibited clearly delineated bouts of resting behaviour and active behaviour; (b) the frequency of resting bouts and the resultant overall cost of daily activity were strongly associated with the size and accessibility of prey-in lakes with smaller and/or less accessible prey, predators rested less frequently, exhibited marginally higher costs when active and had higher overall daytime activity costs. Within each lake, similar changes in activity occurred concurrently with diel changes in prey accessibility/relative density.