Estimating the Probability That Historical Populations of Fish Species Are Extirpated

A dilemma fishery managers sometimes face is how to state with reasonable confidence that a rare fish species has been extirpated from a given system. Data on abundance and distribution are typically scarce for rare species and are mostly in the form of incidental observations, such as fisheries bycatch or museum collections. An objective and quantitative method of evaluating the status of a rare fish species in selected systems—one that is based on incidental observations and that could be used to provide a first warning to managers that a species may be extirpated—would be highly desirable. In this paper, we review two methods that use incidental observations for estimating the probability that a species is extirpated, and we demonstrate the methods with available data sets for Atlantic sturgeon Acipenser oxyrinchus and Alabama sturgeon Scaphirhynchus suttkusi. Special considerations should be made when applying the methods to data sets for fish species. We recommend against using or combining data sets that contain both early life stage and adult observations, and we recommend that priority should be given to data sets obtained in preferred habitat areas. Ancillary information that could provide additional insight regarding the status of the species in the system should be used whenever possible. Finally, we offer a protocol for assembling information and estimating the probability that a fish species is likely to have been extirpated from a system.