Predicting the global mammalian viral sharing network using phylogeography

Gregory F Albery; Evan A Eskew; Noam Ross; Kevin J Olival

Previous investigations into pairwise determinants of viral sharing have been limited to one 47 or two host orders (e.g., bats 17,18 , primates 19 , ungulates 16 , and carnivores 14,16 ), while 48 sometimes lumping together different types of pathogen (e.g., helminths, viruses, and 49 bacteria). Viruses are sometimes shared across large host phylogenetic distances (e.g., Nipah 50 virus in bats and pigs, among many others 20,21 ), requiring a broader understanding of viral 51 sharing across mammals to predict patterns at different taxonomic and geographic scales. In 52 addition, many mammalian orders have yet to be investigated in these analysesmost 53 notably rodents, which are highly diverse and host important zoonotic viruses 5, 8 . In addition, 54 although phylogenetic and geographic viral sharing effects have been empirically 55 demonstrated, the models have not yet been applied to validate viral sharing predictions using 56 external datasets or make inferences about mammals with no known viral associations. If 57 geographic and phylogenetic effects on viral sharing are as ubiquitous as they seem, these 58 variables alone could provide a useful baseline model of viral sharing applicable across the 59 mammal class.