Peak performance is repeatable and captures large individual differences in ruby-throated hummingbirds

Locomotor performance often determines the outcome of interactions with competitors, predators, and prey. In flying animals, the asymptotic load-lifting assay measures maximal muscle power output in vertical flight. Previous studies of small birds have shown that load-lifting performance is linked to flight maneuverability and the outcome of competitive species interactions. Here, we quantify sources of performance variation within a species, namely repeatability, and determine the number of trials that accurately capture individual differences. We conducted 124 load-lifting trials on 13 wild-caught male ruby-throated hummingbirds (Archilochus colubris), testing each individual repeatedly over a 3-day period. We report large individual differences in peak performance, with 70% of the total variation in lifted mass attributed to differences among males. Notably, these differences in muscle performance are independent of body mass and size. An additional 23% of the variation in lifted mass was due to short-term fluctuations, wherein a given male's performance varied across trials and days. We find no systematic effects of experience or time on load-lifting performance. Using simulation, we test the effect of different sampling protocols for measuring individual performance, and show that single-trial protocols yield the highest repeatabilities, but are less suitable for capturing the true underlying differences among individuals. We discuss recommendations for future studies that aim to measure maximum performance. Overall, our results show that the asymptotic load-lifting assay reveals large individual differences and can closely reflect individuals' true maximum capacity.