48 Multi-species assemblages and their disease risk to livestock: a case study of wildlife, livestock and bovine tuberculosis in the UK Monika Böhm, Stijn Bierman*, Glenn Marion*, Michael R. Hutchings† and Piran C.L. White Environment Department, University of York, York, YO10 5DD, UK * Biomathematics and Statistics Scotland (BioSS), Edinburgh EH9 3JZ, UK † Animal Nutrition and Health, Scottish Agricultural College, West Mains Road, Edinburgh EH9 3JG, UK Correspondence to: P.C.L. White, Environment Department, University of York, York YO10 5DD, UK. Email: PCLW1@york.ac.uk ABSTRACT The majority of livestock and human pathogens are generalist in nature and can infect multiple wildlife species. Host species diversity can have positive and negative effects on disease dynamics, by amplifying outbreaks of directly-transmitted pathogens and buffering against those of vector-transmitted diseases. In this study, we assessed the contribution of multi-species assemblages of badger and deer to disease risk in UK livestock at the 10-km square resolution, with particular focus on bovine tuberculosis (bTB) risk. Wildlife assemblages were richest in the southwest and East Anglia. We assumed that increased wildlife species richness poses an increased risk of disease in livestock, and found major high-risk areas for disease in livestock in areas of endemic bTB infection. Using models accounting for spatial autocorrelation, neither deer nor cattle variables had an effect on cattle bTB incidence. Only probability of badger main sett presence per 10-km square was consistently related to cattle bTB incidence in our models. This significant effect was small, however, compared to the strong effect of local spatial processes on cattle bTB. Our results suggest that control of bTB in badgers via non-lethal measures should play a role in future control strategies of the disease in cattle. We discuss the importance of collecting wildlife population and disease data at regional or national