What do we not know – quantifying data gaps and biases in knowledge of bat co-roosting

Conference: 19th International Bat Research Conference
Location: Austin, TX, USA
Date and Time: Aug 11, 2022
Description: Exploring the data available building the knowledge about bat co-roosting and its limitations.

Abstract

Improved understanding of co-habitation of roosts by multiple species of bats is essential for estimating the risks of zoonotic disease transmission. However, ecological data on roosting environments, species richness, bat-bat interactions, viral infections, and other species interactions are scattered throughout the literature, making them difficult to study on a global scale. The research scope for most roost studies has been narrow, focusing on roost type, bat abundance, and locality data while failing to investigate interspecific roosting interactions. To meet this need, we have collaboratively built an open-access dataset of ecological interactions (including co-roosting, trophic, anthropogenic, and parasitic) extracted from the literature to improve our understanding of roost dynamics on a global scale, and to elucidate the role of shared roosts in disease transmission. As of April 2022, > 11,500 interaction records involving > 360 bat species from > 137 countries encompassing a variety of habitats have been extracted from > 175 publications spanning from 1860-2020, all accessible via the Coronavirus-Host community at Zenodo. With this benchmark dataset of open-access digitized interaction data, tools, and workflows, we provide evidence of co-roosting events that we aligned with multiple ontologies (interaction terms, taxonomies, administrative regions) and phylogenies suitable for high-throughput analysis We followed open access and FAIR (Findable, Accessible, Interoperable, and Reusable) data principles for extracting data and choosing methodologies. We identify biases in the coverage of bat interaction records, suggest new tools for biodiversity informatics, and explore obstacles and opportunities in the mining of eco-interactions previously lost in the annals of scientific literature.

More info