Population and sex-specific survival estimates for Atlantic sturgeon: addressing detection probability and tag loss

Abstract

Survival estimates of animal populations provide managers with critical information on productivity, population stability, and demography. Telemetry-based survival estimates can be obtained remotely. The Atlantic sturgeon Acipenser oxyrinchus oxyrinchus is a wide-ranging species whose populations overlap along the East Coast of North America, complicating survival estimation. The objective of this study was to estimate apparent annual survival of the York River population using a Cormack-Jolly-Seber model. In this study, 36 males and 24 females were telemetered and monitored between 2013 and 2019. We considered the fit of a variety of models, selecting the best fit using Akaike’s information criterion. The optimal model estimated survival in seasonal increments and detection probability by sex in monthly increments. Five transmitters failed to leave the river and another 3 stopped being detected within 21 mo, but of those, recapturing fish confirmed 2 had been lost and 3 were technological failures (12.8% of 39 recaptured). Apparent adult annual survival was estimated to be 99.2% (95% CL: 97.9-99.7%). Addressing sex-specific detection probability and failed transmitters while including a length covariate for each individual produced higher survival estimates than previously reported studies of Atlantic sturgeon. Four males and one female appear to have died, with the location of last detection for 4 of the suspected mortalities in shipping channels near the mouth of the Chesapeake Bay, suggesting managers should focus on this area of increased risk. Such high survival estimates of the adult stage suggest Atlantic sturgeon survival may be more similar to other long-lived, late-maturing animal species than to most other short-lived fish species.