Estimation of black bear abundance by management zone in Wisconsin

TL;DR

This study employs a Bayesian state-space model with age-at-harvest data to estimate black bear abundance across Wisconsin's management zones, providing a transparent, data-integrated approach for wildlife management and conservation.

Contribution

It introduces a Bayesian state-space modeling framework that combines prior expert knowledge with harvest data to estimate black bear populations at a regional scale.

Findings

Abundance estimates showed decreasing trends in zones A and B.

Stable population trends observed in zones C and D.

Model improved precision of survival estimates using harvest data.

Abstract

Estimates of population abundance are fundamental to wildlife management and conservation, but are difficult to obtain across large geographic scales and for cryptic species. We used a state-space model with age-at-harvest data in a Bayesian framework to model American black bear (Ursus americanus) abundance and demographic parameters in four management zones in Wisconsin from 2011-2017. We had limited demographic data available from the population, and relied upon a) the model, b) age-at-harvest data, and c) informative prior distributions from a literature review. The estimated posterior means and distributions for abundance and demographic parameters from our models were reasonable for each management zone, and indicated a decreasing trend in zones A and B, and a generally stable trend in zones C and D. The age-at-harvest data updated the posterior distribution and means for initial population size, harvest season survival, and non-harvest season survival, with a notable increase in precision for the survival values. A strength of the model for managers is the formalized process for providing biologically supported information as prior distributions to transparently accommodate expert opinion with measures of confidence when estimating wildlife populations, which can then be updated by the age-at-harvest data and model structure. The integration of prior information and age-at-harvest state-space models in a Bayesian framework efficiently leverage all available information for making zone-specific abundance estimates for the management of harvested species. This may create more informative data for decision-makers when setting harvest quotas, and could lead to more effective monitoring, conservation, and management of cryptic carnivore species.