The role of lipid-store body condition in the behaviour and conservation of sperm whales

Abstract

Lipid store body condition influences, and is influenced by, individual fitness and behaviour. This makes it a potentially important factor in understanding the impact of disturbance and other stressors, which are of increasing conservation concern in a rapidly changing environment. However, body condition is challenging to measure in cetaceans. In Chapter 1, I review the cetacean body condition and behaviour literature, identifying key knowledge gaps and methodological limitations. The remaining chapters then aim to develop improved ways to quantify condition in large cetaceans, investigate its influence on the behaviour of sperm whales (Physeter macrocephalus) in natural and disturbed contexts, and explore the effects of condition-mediated behaviours on vital rates. These aims are pursued using agent-based modelling (Chapter 2) combined with biologging and unoccupied aerial vehicle (UAV) photogrammetry data from the Norwegian sub-arctic (Chapter 3) and Azores archipelago (Chapter 4). In Chapter 2, I develop a formal model of alternative biologically plausible relationships between individual body condition and risk tolerance and use model simulations to explore their potential effects on population consequences of disturbance. I find that such consequences may be very sensitive to condition-dependent response. In Chapter 3, I test for condition-dependent response empirically using biologging data from behavioural response studies (BRS) which exposed tagged sperm whales to naval sonar. Sperm whales in poorer condition exhibited greater reductions in foraging performance for a given sonar exposure relative to better-condition animals. In Chapter 4, I cross-validate UAV volumetrics and tag-derived tissue density as independent measures of body condition. I then extend and combine both methods to develop a novel analysis approach to estimate absolute lipid mass. These results, in combination with future work tying the observed patterns to longitudinal drivers, could help improve prediction, monitoring and forecasting of individual and population consequences of disturbance and multiple stressors.