Strengths and limitations of using participatory science data to characterize a wildlife mass mortality event

Abstract

Large participatory science (i.e., "community science" or "citizen science") platforms are increasingly used at every level of ecological and conservation research, including disease monitoring. Here, we used a comprehensive, ground-truthed mortality dataset to judge how well participatory science data from iNaturalist represented the magnitude, taxonomic, temporal, and spatial patterns of waterbird mortality associated with a mass mortality event following the incursion of Highly Pathogenic Avian Influenza in eastern Canada in 2022. The iNaturalist dataset was effective at identifying species with high mortality (especially Northern Gannets, Morus bassanus), along with the time period and spatial regions with high concentrations of avian deaths. However, iNaturalist data severely underestimated the magnitude, overestimated the taxonomic breadth, and poorly represented the full geographic scope of disease-related deaths. Our results suggest iNaturalist can be used to identify the species, timing, and location of relatively high mortality in situations where no other information is available, and to supplement conventional sources of data. However, iNaturalist alone can neither quantify the magnitude nor pinpoint the mechanisms of mortality and therefore is not a viable substitute for comprehensive mortality assessments.