Does metabolic rate predict risk-taking behaviour? A field experiment in a wild passerine bird

Individuals often show consistent differences in risk-taking behaviours; behaviours that increase resource acquisition at the expense of an increased risk of mortality. Recently, basal metabolic rate (BMR) has been suggested as a potentially important state variable underlying adaptive individual differences in a range of behaviours, including risk-taking. We tested the relationship between BMR and risk-taking in free-living great tits (Parus major) using experimental manipulations of perceived predation risk. We compared the latency of individuals to return to feeders following control (human) and predator (model sparrowhawk, Accipiter nisus) disturbances at fixed feeder locations. We predicted that if variation in risk-taking is shaped by energetic constraints, high BMR individuals should return to feeders sooner following both disturbance types and show smaller changes in risk-taking as a function of predation danger. Higher BMR tended to be associated with lower risk-taking following control disturbances but greater risk-taking following predator disturbances, resulting in a significant interaction between BMR and treatment. Within-individual changes in risk-taking as a function of ambient temperature (a proxy for within-individual changes in energetic constraints) mirrored these results. Lower temperatures tended to be associated with lower risk-taking following control disturbances, but greater risk-taking following predator disturbances, resulting in a significant interaction between temperature and treatment. The effects of BMR and temperature on variation in risk-taking as a function of perceived predation danger were qualitatively similar, suggesting that energetic constraints play a role in shaping risk-taking. However, the hypothesized mechanism (energetic requirements directly influence the optimal expression of risk-taking behaviour) is insufficient to account for the observed negative relationship between energetic constraint and risk-taking following control disturbances. We conclude that variation in risk-taking is associated with differences in energetic constraints, including BMR and temperature, but that the relationship is context-specific, here high vs. low perceived predation risk. Further studies are needed to elucidate potential mechanisms that could generate context-specific relationships between energetic constraints and risk-taking.