Old world fruit bats are important seed dispersers of forest plants as well as of commercial fruit crops. Bats scatter seeds across landscapes and also aggregate (clump) seeds under feeding and roosting trees. In agroforestry landscapes, bat frugivory and seed dispersal can result in simultaneous accrual of socio-economic benefits and costs to humans, which may be further affected by human health risks from zoonotic spillover through human-bat interactions. In our study, we used an integrated approach to assess socio-economic benefits (from seed dispersal) and costs (from frugivory) from bats, in relation to latent zoonotic risk. We carried out the study in five agroforestry landscapes along India’s Western Ghats, for selected commercial fruit crops (cashew, areca, banana, and other pulpy/fleshy fruits). Specifically, we hypothesized that people’s perceptions of benefits from fruit bats would be correlated with measured extents of clumped seed dispersal in fruit plantations. We conducted ecological studies to investigate the effects of fruit crop type and plantation attributes on clump-dispersal, and in turn, on perceived “net benefits” assessed from interviews with plantation-associated people. We then explored associations between perceived socio-economic net benefits and spatial data on disease risk factors for fruit-bat-linked henipavirus transmission. We found that the extent of bat-mediated clump-dispersal was highest for cashew, especially in isolated plantations. People’s perceptions of benefits from fruit bats matched the measured extents of clump-dispersal of cashew and areca. These benefits came with some costs from scatter-dispersal and damage to other fleshy fruits from frugivory by bats. Interestingly, we did not detect tradeoffs between perceived net benefits from bats and disease risk, which is of significance for bat conservation and its implications for human well-being. Overall, our results highlight that bat-mediated seed dispersal needs to be sustained as an important ecosystem service, despite some latent zoonotic risk, in the Anthropocene.

The dataset was collected using a combination of multiple methods, described in detail in our manuscript and supporting information. The methods included field-based ecological studies of fruit bat seed dispersal at fruit plantations, social data collection from interviews with plantation-associated people, and assessments for henipavirus transmission risk in our study area based on secondary information and our field observations.

In the uploaded dataset, we have provided data on 1) plantation-level attributes indicating ecological variables and extents of bat seed dispersal in relation to naturally dropped seeds, 2) data used to assign disease risk levels, at the scale of administrative units in our study area, and 3) disease risk categories assigned by us, and net benefit scores derived from interview responses about fruit bats. All these data were used in our analyses.

For social perception data, the identity of respondents or site-level identifiers or detailed responses were not uploaded due to ethical considerations. Summaries of social perception data have already been provided in the main manuscript and the supporting information.

We have provided detailed metadata on all data files and variables in the uploaded dataset. The metadata are self-explanatory. Any missing values or unavailable data have been denoted by NE or NA. The explanation for all acronyms, units, etc. used can be found in the ReadMe and metadata worksheets of the main data file (Excel workbook).