Hierarchical models for estimating density from DNA mark-recapture studies

Gardner, Royle, and Wegan developed spatially explicit hierarchical models for estimating density when individuals are identified by DNA from non-invasive samples such as hair or scat, conceptually parallel to camera-trap individual identification by coat pattern. By placing the spatial encounter process at fixed sampling locations within a Bayesian framework, the method estimates density directly and handles the spatial structure of detections. Although developed for genetic data, the modeling approach is closely related to and informs camera-trap SECR, and the two data types are sometimes combined. The paper contributed to the maturation of spatial capture-recapture and demonstrated its generality across detection methods. It is cited in both non-invasive genetics and camera-trap density literatures as a key methodological advance.