A global-scale dataset of bat viral detection suggests that pregnancy reduces viral shedding.

Understanding viral infection dynamics in wildlife hosts can help forecast zoonotic pathogen spillover and human disease risk. Bats are important reservoirs of zoonotic viruses, and bat metapopulation dynamics, seasonal reproductive patterns and other life-history characteristics might explain temporal variation in the spillover of bat-associated viruses. Here, we analyse reproductive effects on viral dynamics in free-ranging bat hosts, leveraging a multi-year, global-scale viral detection dataset that spans eight viral families and 96 bat species from 14 countries. Bayesian models revealed that pregnancy had a negative effect on viral shedding across multiple data subsets, and this effect was robust to different model formulations. By contrast, lactation had a weaker influence that was inconsistent across models. These results are unusual for mammalian hosts, but given recent findings that bats may have high individual viral loads and population-level prevalence due to dampening of antiviral immunity, we propose that it would be evolutionarily advantageous for pregnancy to either not further reduce immunity or actually increase the immune response, reducing viral load, shedding and risk of fetal infection. This novel hypothesis would be valuable to test given its potential to help monitor, predict and manage viral spillover from bats.