Summary1. Flight is a key innovation in the evolutionary success of insects and essential to dispersal, territoriality, courtship and oviposition. Wing shape influences flight performance and selection likely acts to maximize performance for conducting essential behaviours that in turn results in the evolution of wing shape. As wing shape also contributes to fitness, optimal shapes for particular flight behaviours can be assessed with aerodynamic predictions and placed in an ecomorphological context. 2. Butterflies in the tribe Haeterini (Nymphalidae) are conspicuous members of understorey faunas in lowland Neotropical forests. Field observations indicate that the five genera in this clade differ in flight height and behaviour: four use gliding flight at the forest floor level, and one utilizes flapping flight above the forest floor. Nonetheless, the association of ground level gliding flight behaviour and wing shape has never been investigated in this or any other butterfly group. 3. We used landmark-based geometric morphometrics to test whether wing shapes in Haeterini and their close relatives reflected observed flight behaviours. Four genera of Haeterini and some distantly related Satyrinae showed significant correspondence between wing shape and theoretical expectations in performance trade-offs that we attribute to selection for gliding in ground effect. Forewing shape differed between sexes for all taxa, and male wing shapes were aerodynamically more efficient for gliding flight than corresponding females. This suggests selection acts differentially on male and female wing shapes, reinforcing the idea that sex-specific flight behaviours contribute to the evolution of sexual dimorphism. 4. Our study indicates that wing shapes in Haeterini butterflies evolved in response to habitat-specific flight behaviours, namely gliding in ground effect along the forest floor, resulting in ecomorphological partitions of taxa in morphospace. The convergent flight behaviour and wing morphology between tribes of Satyrinae suggest that the flight environment may offset phylogenetic constraints. Overall, this study provides a basis for exploring similar patterns of wing shape evolution in other taxa that glide in ground effect.

Trade‐offs in performance expression occur because animals must perform multiple whole‐organism performance tasks that place conflicting demands on shared underlying morphology. Although not always detectable within populations, such trade‐offs may be apparent when analyzed at the level of the individual, particularly when all of the available data are taken into account as opposed to only maximum values. Detection of performance trade‐offs is further complicated in species where sexual dimorphism drives performance differences between males and females, leading potentially to differing patterns of trade‐offs within each sex. We tested for within‐ and between‐individual trade‐offs among three whole‐organism performance traits (sprint speed, endurance, and bite force) in adult male and female Anolis carolinensis lizards using all of the measured performance data. Sprinting and endurance did not trade‐off among individuals in either sex, but we found a significant negative among‐individual relationship between sprint speed and bite force in females only, likely driven by the mechanical burden of larger than optimal heads imposed on females through intralocus sexual conflict. We also found evidence for marked within‐individual plasticity in male bite force, but no within‐individual trade‐offs between any traits in either sex. These data offer new insight into the sex‐specific nature of performance trade‐offs and plasticity and, ultimately, into the constraints on multivariate performance evolution.

Maximal whole-organism performance traits measured in the laboratory and the levels of performance expressed in the field often exhibit a mismatch, complicating our understanding of the selection pressures influencing the evolution of performance traits. To better understand the evolution of locomotor performance, we built an individual-based simulation to test hypotheses about selection on locomotor performance. Starting with a population of individuals with two correlated but variable performance traits, we simulated these individuals surviving and reproducing in a complex environment, presenting each individual with successive ecological challenges requiring specific performance capabilities over their lifespan. While most challenges require sub-maximal speeds, intermittent bouts requiring increased performance, such as escape from predators, introduce strong, but infrequent, selection for maximal performance. By comparing the results of simulations run with individuals that only perform at their maximum levels versus those that adjust this effort, we show that intra-individual variation in speed confers a selective advantage, regardless of the extent of that variation. We also show that the direction and strength of the correlation between the two performance traits affects the evolutionary trajectory of phenotypic change. Ultimately, this model allows us to simulate the evolution of movement speeds over a range of selective contexts, offering insight into the factors affecting the evolutionary relationship between optimal and maximal performance.

Animals rely on their ability to perform certain tasks sufficiently well to survive, secure mates, and reproduce. Performance traits depend on morphology, and so morphological traits should predict performance, yet this relationship is often confounded by multiple competing performance demands. Males and females experience different selection pressures on performance, and the consequent sexual conflict over performance expression can either constrain performance evolution or drive sexual dimorphism in both size and shape. Furthermore, change in a single morphological trait may benefit some performance traits at the expense of others, resulting in functional trade-offs. Identifying general or sex-specific relationships between morphology and performance at the organismal level thus requires a multivariate approach, as individuals are products both of an integrated phenotype and the ecological environment in which they have developed and evolved. We estimated the multivariate morphology→performance gradient in wild-caught, green anoles (Anolis carolinensis) by measuring external morphology and fore- and hindlimb musculature, and mapping these morphological traits to seven measured performance traits that cover the broad range of ecological challenges faced by these animals (sprint speed, endurance, exertion distance, climbing power, jump power, cling force, and bite force). We demonstrate that males and females differ in their multivariate mapping of traits on performance, indicating that sex-specific ecological demands likely shape these relationships, but do not differ in performance integration.

Performance traits are energetically costly, and their expression and use can drive trade‐offs with other energetically costly life‐history traits. However, different performance traits incur distinct costs and may be sensitive to both resource limitation and to the types of resources that are accrued. Protein is likely to be especially important for supporting burst performance traits such as sprint speed, but the effect of varying diet composition on sprint training in lizards, an emerging model system for exercise training, is unknown. We tested the hypothesis that the response to sprint training is sensitive to both the type and amount of resources in Anolis carolinensis. We also measured bite force across all treatments as a control whole‐organism performance trait that should be unaffected by locomotor training. Both mass and bite force are reduced by dietary restriction over the course of 9 weeks of sprint training, but sprint speed is unaffected by either training or dietary restriction relative to controls. Furthermore, protein supplementation does not rescue a decline in either mass or bite force in trained, diet‐restricted males. These results contrast with those for endurance training, and suggest that sprint speed is more canalized than either endurance or bite force in green anoles.