The authors describe a conceptual model focused on grizzly bear mortality, which in turn in is essentially a model of human interactions and the natural system. Grizzly bear deaths in the Rocky Mountains continue to be almost exclusively caused by humans, despite factors such as legal protection and controlled hunts. The authors conclude that the greatest threat to grizzly bears in the Southern Rocky Mountains is human-caused mortality. Because of this conclusion , their focus on the problem centers on deterministic (i.e., human-related) causes of mortality rather than stochastic phenomena. Grizzly bear mortality caused by humans is determined by the rate of encounter between the two and the probability that such an encounter will result in a grizzly bear death. These rates can be influenced by several factors such as number of humans in grizzly habitat, human behavior (ability to avoid encounters and conflicts, responses to encounters), amount and dispersion of human access to public and private lands, grizzly behavior (aggressiveness, levels of habitutation or conditioning to humans and human foods), grizzly population size, distribution of food, and distribution of dominant bears or bears otherwise dangerous to the survival of cubs and subordinates. The authors further conclude that human behavior, access and numbers are controlled by individual and cultural factors, mainly political agents (elected officials and policy formulators), agency decision makers, and the public.

They recommend 8 specific measures to change science and management in ways that reduce complexity, uncertainty, and conflict, and thus, promote the survival of the grizzly bear:
"1) Grizzly bear conservation would benefit from continued and more comprehensive mapping of its problems at all relevent scales and domains. . .,
2) grizzly bear conservation would benefit from the integration of research and management not just into an adaptive management framework but also as part of a larger policy-making and management system with increased learning capabilities at all levels. . . ,
3) natural resource management agencies would better accomplish any complex task if they were redesigned, it is virtually certain that agencies would be much more effective at conserving grizzly bears if they were more creative and open to learning from the entire system they were managing. . . ,
4) Management would be more effective if assessment of human impacts emphasized a scale commensurate to policy goals and the limitations of grizzly bear investigations. . . ,
5) grizzly bear conservation could benefit from clarification of the burden of proof on agencies managing grizzly bears and their habitat . . . ,
6) science and management could also benefit from a clear statement of acceptable risk and relevant time frames for management . . . ,
7) following from recomendations one through three, researchers could contribute by their pursuit of reliable means to monitor trends and size of grizzly bear populations. . . ,
8) grizzly bears could benefit by the widspread adoption of management strategies known to be effective but not always implemented . . ."

---