Multidimensional analyses of physical performance reveal a size‐dependent trade‐off between suites of traits

1. Animal movement is multidimensional and complex, and to understand the motor system of wild animals in the context of their natural ecology, we must analyse how suites of performance traits both mutualistically and antagonistically affect function-a necessity highlighted by previous work on performance trade-offs. 2. Evidence from some studies of human athletes using multidimensional analyses of performance suggests that overall quality among individuals can mask functional trade-offs within them, yet no studies have tested this idea using wild animals. In this study, we investigated the possible mutualistic and antagonistic associations among eight different whole-animal performance traits in male and female northern quolls (Dasyurus hallucatus). 3. We detected trade-offs between pairs of performance traits when conducted on raw standardized data using both Pearson product moment correlations and partial correlation analyses. For example, grasp strength was negatively associated with beam-running speed using both analyses, suggesting that morphological designs that enhance grasp strength simultaneously compromise an animal's motor control or stability on a narrow beam. 4. In addition, we detected a trade-off between two distinct sets of performance traits; grasp strength, bite force and maximum oxygen consumption were negatively associated with jump acceleration and beam-running speed. This trade-off between sets of performance traits accounted for around one-third of the total variance in performance among individuals and was primarily driven by the effects of body size on both groups of traits. Larger body sizes improved grasp strength, bite forces and maximum oxygen consumption rates but decreased jump accelerations and beam-running speeds. 5. Because the first component of a principal component analysis based on all eight performance traits (PC(P)1) did not load in the same direction for all traits, PC(P)1 did not represent an overall metric of motor performance-which differs from previous multivariate analyses of human physical performance. Our study highlights the importance of studying suites of traits when exploring the functional phenotype of organisms rather than just one or two dimensions of performance.