Evaluating threshold management for willow grouse harvest: tracking open and closed areas during 12 years.

Abstract

Managers must make decisions in the face of uncertainty, especially when available resources are limiting. Identifying thresholds when certain conditions are met or exceeded enable the potential to mitigate risks. In 2005, sustainable harvest levels of willow ptarmigan were identified to avoid harvest efforts exceeding three hunter days km2. Here we evaluate these recommendations by analyzing line transect counts and harvest data from six areas forming three open/closed pairs in a region of state managed willow ptarmigan harvest. We developed three sets of Bayesian hierarchical models, one static distance model, and two dynamics models. One mechanistic hazard model and a Gompertz phenomenological model. Adult and juvenile density showed pronounced year-to-year variation that was largely synchronous across all six sites regardless of hunting status. The harvest effort parameter shows a striking difference between the two models. In the Hazard model, is positive, and excludes zero with near certainty, but in the Gompertz model, the parameter is highly uncertain. However, the two models do not contradict each other but answer complementary questions with different sensitivity to the harvest signal, harvest mortality is additive at the individual level, but this additive mortality is masked at the level of population abundance. The demographic cost of harvest is therefore real and quantifiable through the survival chain, but bounded in the long run by the stabilizing dynamics. A fixed limit anchored to monitored effort and bag is not a crude substitute for adaptive management but the appropriate design under the information commonly at hand. It will be a precautionary instrument grounded in the one relationship this study establishes firmly, the translation of hunter effort into harvest mortality.