Effect of Pre‐Harvest Mortality on Harvest Rates and Derived Population Estimates

• Published in: The Journal of Wildlife Management
• Main Authors: Cooch, Evan G., Alisauskas, Ray T., Buderman, Frances E.
• Other Authors: National Institute of Food and Agriculture
• Format: Article in Journal/Newspaper
• Language: English
• Published: Wiley 2021
• Subjects: Nature and Landscape Conservation; Ecology; Ecology, Evolution, Behavior and Systematics; Arctic
• Online Access: http://dx.doi.org/10.1002/jwmg.21986; https://onlinelibrary.wiley.com/doi/pdf/10.1002/jwmg.21986; https://onlinelibrary.wiley.com/doi/full-xml/10.1002/jwmg.21986; https://wildlife.onlinelibrary.wiley.com/doi/pdf/10.1002/jwmg.21986

ABSTRACT Banding waterfowl, in combination with the citizen science provided by hunters that report marks from harvested birds, is a long‐standing, institutionalized practice for estimating probabilities of survival and exploitation (i.e., legal harvest from such populations). Range‐wide population abundance can also be estimated by combining the number of banded individuals with the number harvested from the population. Waterfowl marking with uniquely identifiable bands done during late summer in North America is often referred to as pre‐season banding. For example, mass capture of arctic geese for pre‐season banding is normally done in July (nonbreeders) or August (failed breeders and breeders with young) during flightless molt of respective groups. An important assumption for proper inference about harvest probability provided from such samples is that there is no mortality, natural or otherwise, during the interval between when individuals are marked and when hunting seasons begin. We evaluated the effect of variable mortality that could occur between marking and subsequent hunting seasons on estimates of survival, recovery, and harvest probabilities using simulation pertinent to a typical waterfowl species. We fit a Brownie tag‐recovery model to the simulated data and calculated the estimator bias that resulted from various pre‐harvest mortality scenarios. There was no effect on survival probability during the interval between annual banding in subsequent years, but recovery probability, and thus estimated harvest probability, was directly and inversely related to pre‐harvest mortality of juveniles. The magnitude of negative bias in harvest probability of juveniles increased further as the fraction of the population sampled declined. If the probability of pre‐harvest mortality differs between marked and unmarked individuals, the negative bias in harvest probability results in overestimates of derived abundance that increases as the proportion of marked individuals in the population declines. We used our ...