Modeling for Decision-Support: Identifying Thresholds and Reducing Uncertainty in the Harvest Management of Barren Ground Caribou

Decision-support systems, scenario analysis, and risk assessment are important tools for resilience-based resource stewardship. In the case of stewardship of caribou herds, decision makers face great uncertainties and significant data gaps, which make policy making particularly challenging. Understanding the effects of human harvest on highly dynamic caribou populations and knowing when to restrict harvesting are part of this process. We developed and used 'the caribou calculator', a generalizable modeling framework, to examine the effects of human harvest levels on caribou herd population dynamics. With the model, we generate a set of middle-range propositions to help guide harvest and monitoring decision-making in this fluctuating ecological system. For example, if the total annual harvest level does not exceed 2,000 cows and 1,000 bulls, then the hunting is not a major driver of herd dynamics unless the herd size drops below around 70,000 animals. We illustrate these general principles using three comparative case studies: the Porcupine Caribou Herd (1985-2009), the Bluenose West Caribou Herd (2000-2009), and the Bathurst Caribou Herd (2003-2009). These herds differ in their management context, their ecological setting, and the amount of monitoring data available. Through simulation modeling, we demonstrate an approach that managers can use to place bounds around the uncertainty in harvest-related decision-making for Human-Rangifer Systems.