Drivers of biodiversity change in the Anthropocene are global, dynamic, non-stationary, and pervasive across the tree of life. A predictive understanding of biodiversity responses to these drivers will require a full exploration of the 3-dimensional space-time-taxonomy domain of ecology. Big, unstructured and opportunistic species occurrence data from citizen science and museum specimen collections represent an increasing proportion of all existing biodiversity data and provide information across spatial, temporal, and taxonomic scales difficult to cover using any other data source. Yet, the potential of these data sources to detect spatiotemporal biodiversity change remains largely unrealized, primarily due to the challenges of extracting useful ecological signals from the noise in the data. Despite the challenges, understanding how to make the most out of this huge and fast-growing data stream will be key to detect and understand biodiversity change across large spatial, temporal, and taxonomic scales. In this talk, I will discuss a general approach for extracting signals of biodiversity change using unstructured and opportunistic species occurrence data from citizen science and museum specimen collections.
Results/Conclusions
Four main steps can improve our ability to extract signals of spatiotemporal biodiversity change from unstructured and opportunistic species occurrence data: introducing spatiotemporal structure, borrowing strength across taxa, modeling the observation process, and integrating data sources. Combining some or all of these steps will enable us to unlock the potential held in citizen science and museum specimen databases to detect spatiotemporal biodiversity change and translate these near-constant streams of data into early-warning systems and actionable insights for biodiversity conservation and management.