Bats (Order: Chiroptera) are important reservoirs for viral pathogens, a number of which are high impact zoonotic diseases. Previous work has indicated that geographic and life history traits are important drivers of viral diversity in this group. However, these studies have focused on limited numbers of bat species or on particular viral families. We expand on prior analyses by examining how geographic distribution, life history traits (body size, birthing frequency), foraging ecology (diet, foraging mode), group characteristics (size, roosting associations), and traits related to body temperature variation (torpor use) may explain patterns of viral diversity in bats. We assembled a trait database of 19 ecological predictors for 770 species of bats and collected information on the number of viral families hosted by these species. Using boosted regression trees, we assessed the importance of bat ecological traits for predicting patterns of viral diversity. Additionally, we also wanted to understand how the importance of different bat traits might shift when considering morphologically distinct viruses. To accomplish this, we grouped the viral families carried by bats using a cluster analysis of 9 viral traits and, again using boosted regression trees, examined which bat ecological traits predicted diversity within these viral clusters.
Results/Conclusions
Our models identify species’ trait profiles that are in accordance with prior work. Larger, more widely distributed bats that are sympatric with a greater number of species host a greater diversity of viruses. Traits related to body temperature variation are also important, suggesting that species that undergo frequent changes in body temperature host a lower diversity of viruses. Our cluster analysis partitioned viral families into two groups, roughly representing DNA and RNA viruses. The ecological traits that predicted RNA viral cluster diversity were similar to those predicting total viral family diversity. However, for DNA viruses, lifespan had a much greater relative importance in our regression tree models. Our results highlight key ecological traits that distinguish viral carriers and provide support that traits previously shown to be significant are important drivers of viral diversity across the order Chiroptera. We also find that traits related to changes in body temperature, which has been hypothesized to influence viral dynamics in bats, are important correlates of viral diversity. Finally, our models enable us to make predictions about species’ of bats that may be undetected viral carriers. These predictions could help to direct future disease surveillance efforts.