An Examination of the Social and Physiological Experience of Captive Livingstone’s Fruit Bats (Pteropus livingstonii)

Abstract

This thesis presents a series of empirical studies designed to quantify the complex social and physiological experiences of individuals within the captive population of a critically endangered bat species, the Livingstone’s fruit bat (Pteropus livingstonii). The ultimate goal of this work was to apply leading-edge methodologies and technology to gain novel, evidence-based insights into the behaviour and welfare of this species in captivity. This work has not only informed the captive management of P. livingstonii, but has also developed techniques that can be applied to maximise the captive welfare of a range of other species. A brief summary, explaining how each chapter is embedded within the overall thesis and summarising its major findings, precedes each chapter. Chapter One, the introductory chapter, will start by evaluating our current understanding of the natural history and behavioural ecology of P. livingstonii, with a focus on captive behaviour. The following sections of this chapter will then provide necessary background information to the methodologies applied in subsequent data chapters, including social network analysis, space use assessment, hormone analysis and radio-frequency identification technology. The introductory chapter is intended to provide the reader with sufficient scientific context to interpret the results of each data chapter. Chapter Two outlines the implementation of social network analysis to characterise individual social experiences in the largest captive population of P. livingstonii, housed at Jersey Zoo. Variation in network complexity and significant individual variable-based assortment is identified in this chapter. Chapter Three presents principal component analysis as a novel compliment to traditional enclosure use assessment methodologies, quantifying the individual spatial preferences of P. livingstonii. This chapter also quantifies the relationship between high-value resource location within the enclosure at Jersey Zoo and population-level space use. Chapter Four outlines the validation and application of an enzyme immunoassay to non-invasively measure faecal cortisol in the P. livingstonii population. This chapter examines the relationship between cortisol titres, individual social roles and behavioural frequencies performed by individuals in the captive population. We find that males, older individuals and lactating females have higher concentrations of cortisol than non-lactating females and younger individuals and that social roles related to conspecific affiliation were linked to decreased cortisol titres. Lastly, Chapter Five demonstrates the application of unique radio-frequency identification technology to autonomously collect individual locational data. The resulting dataset was then utilised to interrogate the foraging network displayed by P. livingstonii and Pteropus rodricensis (a cohabitating heterospecific) populations housed at Jersey Zoo, elucidating the relationship between individual bat characteristics and social foraging patterns. Finally, the discussion chapter evaluates the significance of the main findings of each chapter of this thesis, highlighting the wider applicability of results and methodologies to the assessment of captive animal welfare. This section also critically discusses recent challenges to hypothesis testing in the field of animal social network analysis, to evaluate the validity of methods utilised throughout this thesis. Finally, future research directives leading on from this work which may provide further benefit to the management of captive populations are proposed.