Harnessing social media data to track species range shifts

Biodiversity monitoring programs and citizen science data remain heavily biased toward the Global North. Especially in megadiverse countries with limited biodiversity records, incorporating social media data can help address existing data gaps. However, whether such data can significantly improve our understanding of range-shifting species is still unknown. We tested whether social media data improved our knowledge of the range dynamics of a rapid range-shifting butterfly, the tawny coster (Acraea terpsicore). We collated locality data from Flickr and Facebook and compared these with occurrence data from the Global Biodiversity Information Facility (GBIF). We used species distribution models (SDMs) and niche assessments, which we calibrated with data from GBIF alone and both sources combined (GBIF and social media data) to analyze range shift dynamics. Social media data increased occurrence records by 35%, and the proportion of social media data was higher in countries poorly represented in GBIF. In addition, we obtained new distributional information from well-represented countries (e.g., Australia and Malaysia). Over time, the SDMs calibrated with GBIF and social media data showed greater expansion rates than SDMs based solely on GBIF data. The niche assessments revealed that GBIF-only data failed to capture regions with relatively low maximum temperature, relatively low precipitation and high elevation. Our results highlight the potential of harnessing social media data to track rapid biodiversity redistribution in response to climate change.