Social network characteristics and predicted pathogen transmission in summer colonies of female big brown bats (Eptesicus fuscus)

Abstract

Host behavior can affect host-pathogen dynamics, and sociality is predicted to increase risk of pathogen exposure. Many species minimize costs of parasitism by only aggregating seasonally, such as during reproductive periods, but colonial species may still be limited in their potential to evade pathogens. Bats are among the most gregarious mammals and females of many temperate species form maternity colonies in summer where they communally raise pups in both natural and anthropogenic roost structures. Social network structure may differ between natural and anthropogenic roosts in ways that affect pathogen dynamics. We used social network analysis to quantify interactions of big brown bats (Eptesicus fuscus) in a tree-roosting colony, where the colony is divided among multiple trees each day, and a building colony, where most of the colony roosts together each day. We simulated transmission of a pathogen throughout both sets of networks. We tested three hypotheses: (1) network metrics differ between pregnancy and lactation; (2) changing network structure between reproductive stages influences predicted pathogen dynamics; and (3) network metrics and predicted pathogen dynamics differ between colonies of bats in trees versus buildings. Network structure was weaker for bats roosting in trees during pregnancy and lactation compared to bats roosting in a building, and our models showed that a hypothetical pathogen would spread more rapidly for bats in the building colony. Our results are important for understanding variation in social tendencies and pathogen transmission among colonies of bats and have implications for conservation and public health. Host behavior, particularly social behavior, can affect dynamics of wildlife pathogens. Bats are highly social mammals and females of temperate species form colonies in spring and early summer in tree or building roosts. Thermal characteristics of trees and buildings appear to differ in ways that affect roosting behavior and social interactions. We used social network analyses to quantify interactions of big brown bats in tree and building roosts and simulated consequences for pathogen dynamics. Network structure was weaker for bats roosting in trees with more frequent roost switching and relatively diffuse contacts across the network. Our models showed that a hypothetical pathogen could spread up to four times faster in a building colony compared to a colony of bats roosting in trees. Our results are important for understanding how sociality can influence pathogen dynamics in bats and have implications for conservation and public health.