Error in trapper-reported sex of lynx (Lynx canadensis) and wolverine (Gulo gulo): implications for analyses of harvest records

Monitoring trends in the abundance of furbearers may be challenging, particularly at the spatio-temporal scales relevant for management. As such, wildlife managers often rely on harvest records to identify broad-scale harvest patterns and trends. Yet, the reliability of harvest records is often unknown. Analyses of harvest records to determine quotas and assess sustainability need to account for error rates when they are > 0. We evaluated the accuracy of trapper-reported sex of lynx (Lynx canadensis; n = 185) and wolverine (Gulo gulo; n = 467)—two meso-carnivores commonly targeted by fur trappers in northwestern Canada and Alaska—by comparing that to sex determined via necropsies of the same carcasses. Overall error rates differed significantly between wolverine (5%) and lynx (13%). Error rates were sex-biased for wolverine, but not lynx. Body size did not affect error rates for either species. Our data demonstrated species- and sex-specific error rates in the sex reported in harvest records. Error rates for wolverine (5%) were likely trivial for determining harvest sustainability because sex-based bias was small, given that overall accuracy was high. While error rates were greater for lynx (13%), there was no sex-based bias in trapper-reported sex. Because accuracy was lower for lynx, managers should exercise caution when using trapper-reported sex to conduct population analyses or assess harvest sustainability. Managers should be particularly interested in error rates of harvested species that exhibit relatively little sexual size dimorphism and lack obvious genitalia, similar to lynx. We recommend assessing error in trapper-reported sex prior to analysis of harvest records, as well as ongoing education with trappers to increase their ability to reliably determine the sex of animals they harvest.