Predicting hunter success rates from elk and hunter abundance, season structure, and habitat

Many factors have been hypothesized to affect hunter success rates, and models have been developed to predict success rates as a function of some of these factors. However, no models have been developed that can accommodate complex elk (Cervus elaphus) hunting season structures such as those in Idaho in which a single population may be targeted by multiple hunts and the proportion of the population residing within a hunt's geographic boundary is unknown and variable over time. We developed such a model by applying generalized linear mixed-effects models to time-series data on population-scale success rates, hunter and elk abundance, hunting season structure, and habitat variables. The average success rate from a particular season structure was fit using data from 1990-1995 unlimited-entry elk hunting seasons throughout Idaho. We tested the model's predictive ability using data from 1996-1997. Results indicated that road density, season structure, elk abundance, and hunter-elk ratios were important predictors of elk hunter success rates in Idaho for hunting seasons similar to those experienced from 1990-1995. The model had poor predictive ability for seasons dissimilar to those from 1990-1995, While providing insight into factors associated with a limited range of circumstances, this first step in developing predictive equations highlighted the importance of model validation and suggested the need for collecting longer time-series data and appropriately scaled and representative climate data.