The ecology and epidemiology of devil facial tumour disease

Abstract

Emerging infectious diseases are increasingly recognised as a significant threatening process in conservation biology. Empirical studies aimed at understanding the epidemiological and ecological processes underlying disease transmission and its impact on host populations are crucial for designing disease management strategies. The Tasmanian devil (Sarcophilus harrisii), is threatened with extinction by Devil Facial Tumour Disease (DFTD), a novel and fatal transmissible cancer. In this thesis, I use proximity-sensing radio collars to reveal empirical contact networks in a wild devil population and infer the role of seasonal and demographic network structure dynamics in the epidemiology of DFTD. I further use this information to build disease-simulation network models to assess the role of contact heterogeneities in epidemic behaviour. Finally, I use longitudinal data sets that followed the natural progression of DFTD to compare contact patterns, epidemiology and impact of the disease in subpopulations which differ in their genotypic structure and diversity. The results indicate increased frequency and length of male-female contacts during the mating season, compared with the non-mating season. These strong inter-sex contact preferences during the mating season suggest that the transmission dynamics of DFTD might be frequency dependent. There are strongly connected individuals in the network, but their identity changes with season, providing limited scope for targeting disease control actions to particular demographic groups. Incorporating network structure and its seasonal and demographic dynamics in disease simulation models had a modest effect on the epidemic threshold for DFTD compared to traditional compartmental disease models. Quantifying heterogeneities in contact patterns and assessing their role in disease spread do, nonetheless, represent an important step for predicting epidemic behaviour and developing approaches for managing wildlife diseases. This study has provided the first evidence of reduced impact and slow progression of DFTD in a wild population as the disease moves to a genetic subpopulation with major histocompatibility complex genes differing from those of the tumour itself. Biting patterns associated with disease transmission did not significantly differ between the two genetic subpopulations. Therefore, differences in genetic diversity in the host immune system or reduced virulence in the pathogen are proposed as plausible explanations for the epidemiological differences between subpopulations. In addition, I found that most tumours were located inside the oral cavity and devils with fewer bites were more likely to develop DFTD, which suggests that the probability of acquiring infection is higher in devils delivering bites than in those receiving bites. The findings of this study, which examined the natural progression of an emerging disease in an ecological and epidemiological context, have direct implications for designing future conservation strategies for the species, and are broadly applicable to a range of other conservation challenges posed by wildlife diseases.