Using citizen science and artificial intelligence to understand stingray spatial ecology

Conservation of threatened stingrays relies on taxonomic clarity and the accurate knowledge of a species’ spatial ecology. Elasmobranch taxonomy is complex, often relying on external morphologies for classifications. Photo identification (photo-ID) has emerged as a vital research tool, especially when studying species that are difficult to distinguish. Where uncertainties in taxonomy exist, high-quality photographic evidence of external features can be a useful tool to help resolve these issues. Photos can also be used to identify individuals by using their natural markings as "tags", to track movement and residency patterns, thereby helping to resolve spatial ecology uncertainties. Contributions from citizen scientists can facilitate the accumulation of large, long-term image databases, providing valuable scientific information for conservation and management. However, these large databases present challenges, particularly being time-consuming to work with. Technological advancements such as artificial intelligence (AI) hold potential to streamline and expedite research and analysis of large datasets, but currently remains relatively underexplored in rays. 

In this study, we used a combination of photo-ID, citizen science, and AI to clarify the distribution of Taeniura lymma and Taeniura lessoni, examine the effectiveness of citizen science data in answering spatial ecology research questions, and explore the potential of AI in distinguishing between these closely related species. The outcomes of this project have practical implications by contributing new data on taxonomic distribution of closely-related rays, demonstrating the value of citizen science in spatial ecology research, and providing a crucial first step into further applications of innovative technology for stingray spatial ecology.