Estimating crippling loss from hunting with multistate models: a case study on northern bobwhites

Hunting as a recreational pursuit provides an important ecosystem service worldwide. Harvest management plays a vital role in regulating wildlife take to ensure long‐term population sustainability and meet value‐based objectives (e.g. hunter satisfaction). However, managers rarely have complete control or observability of harvest mortality. A particularly uncertain component is the number of animals killed but not recovered, known as crippling loss. We examined this issue in a commonly hunted species in the USA, the northern bobwhite Colinus virginianus , which shares life‐history traits with many game birds. Using an extensive dataset spanning three hunting seasons across four sites, we parameterized a multistate survival model to estimate survival, hunter‐recovered mortality, and crippling loss. We also explored field‐observed crippling rates to estimate the difference between realized and perceived rates of crippling. Estimated total mortality from harvest was 0.11 (95% CrI = 0.10–0.13). The probability of a bobwhite being harvested and recovered was 0.08 (95% CrI = 0.07–0.09), while the probability of being shot but not recovered (i.e. crippled) was 0.03 (95% CrI = 0.02–0.05). Crippling represented 29% (95% CrI = 22–38%) of total harvest. For field‐observed crippling, the average observer reported that 25% of total harvest was due to crippling loss, though this was underestimated compared to the multistate model for some sites and years. This may reflect the limited ability of hunters/observers to detect birds that were wounded but died later. Our findings demonstrate that crippling loss can represent a substantial and overlooked component of total harvest mortality. If unaccounted for, this source of mortality may lead to overharvest. The modeling framework presented here offers an approach for estimating crippling loss for species lacking direct estimates. In the absence of species‐specific data, our estimates may serve as provisional values to support conservative harvest management.