Increasingly, animals utilize anthropogenic landscapes and interact with humans through the sharing of resources, infrastructure, and direct contact. Monitoring the behavior and ecology of nocturnal and cryptic animals in human-dominated habitats remains challenging. Records deposited onto social media websites can provide a resource for gathering knowledge of animal distributions, behavior, and interactions with humans. We capitalize on social media postings of ringtails (Bassariscus astutus), a common mammal species native to North America, to gain knowledge of ringtail ecology in human-modified environments. Ringtails are not currently recognized as a synanthropic species, but are known to use human-provided food and structures in parts of their range. Using image classification of 8,177 media results identified through systematic queries, we detect 473 publicly available videos and photos of wild ringtails from Youtube and Flickr.
Results/Conclusions
With our novel dataset, we infer new occurrences, behavioral activities, and interactions with other species. By comparing human interaction events captured by handheld cameras with animal events captured on trail cameras, we describe a range of ringtail and human behaviors, and predict which environmental contexts promote human-wildlife conflict. For 142 media records, we identify geographic locations, and create a species distribution model that controls for underlying biases in the data set. Using ringtails as a test case, we review the challenges and reliability of social media data, and emphasize the complementary use of social media-derived records with more traditional field methods and citizen science repositories to expand spatial and ecological knowledge of an elusive species. Lastly, we suggest other species that may be candidates for social media assessment.