Energetic trade-offs and feedbacks between behavior and metabolism influence correlations between pace-of-life attributes

Correlations between behavioral, physiological, and morphological traits linked to life history' have been given the label "pace-of-life syndrome" (POLS), hypothesized to arise through variation in the resolution of a trade-off between present and future reproduction. However, other trade-offs over energy allocation may also have effects and influence the present-future trade-off. We analyzed an optimality model of basal metabolic rate (BMR) across variation in food availability and two types of mortality. The model contained three major features: (1) feedback between activity and energy acquisition, (2) links between BMR and the use of energy for other traits, and (3) allocation trade-offs between BMR and all other traits, between activity and defense, and between defense against activity-related risk and activity-independent risk. The model produced an intermediate optimal BMR that was usually highest at an intermediate level of food availability. Food availability and both types of mortality risk interacted to influence the exact value of optimal BMR. Trait correlations expected in the POLS existed under some environmental conditions, but these correlations flipped sign under different conditions and were not always strong. Our model reproduces trait correlations consistent with the POLS, but also generated a "sloppy" syndrome with considerable non-POLS-like variation. In addition, among-individual, non-adaptive variation in BMR produced adjustments of the other traits. These fit a best-of-a-had job strategy, and the adjustments further weakened trait correlations. The results emphasize that variation in resources and mortality risk creates a diversity of correlation structures. This complexity means the POLS is likely to be a variable construct. Significance statement Many attributes important for reproduction and survival are associated. Such associations may arise through common physiological processes and con-elated selection. We modeled metabolic rate within a system in which foraging behavior both depended on and mediated the acquisition of resources necessary for metabolism, while energy was allocated among multiple attributes. Variation in several environmental variables (food availability and two types of mortality risk) influenced basal metabolic rate, activity, and defenses against mortality risk. This variation affected the correlations between the traits in complex ways. When basal metabolic rate was non-optimal, evolution of the allocation of energy to other traits partially compensated, but this further eroded consistent trait correlations. Our results indicate that complexity in how energy is acquired and used can potentially disrupt trait correlations normally associated with the pace-of-life syndrome.