Cooperative Fish and Wildlife Research Units Program: Headquarters
Education, Research and Technical Assistance for Managing Our Natural Resources

Adaptive management of renewable biotic resources accounts explicitly for uncertainties in system responses to management and recognizes the importance of reducing uncertainties while pursuing other management goals. An adaptive approach to the harvest of wildlife populations that are subject to (1) uncontrollable environmental variation, (2) uncertainties about the appropriate characterization of resource dynamics, (3) limitations on the controllability of harvest rates, and (4) uncertainties as to population status, expressed as sampling variation in the monitoring of populations and habitats, is described. Adaptive management is framed in terms of maximizing long-term harvest value against a background of various kinds and degrees of uncertainty, with an emphasis on structural uncertainty. By an appropriate extension of the ''system state,'' adaptive optimization can be defined in terms of Markov decision processes. Solution algorithms are described for systems that are subject to structural uncertainty and are either partially or completely observable. Adaptive optimization is illustrated with an example in waterfowl harvest management that incorporates uncertainty in the relationship between harvest rates and survivorship.Adaptive management of renewable biotic resources accounts explicitly for uncertainties in system responses to management and recognizes the importance of reducing uncertainties while pursuing other management goals. An adaptive approach to the harvest of wildlife populations that are subject to (1) uncontrollable environmental variation, (2) uncertainties about the appropriate characterization of resource dynamics, (3) limitations on the controllability of harvest rates, and (4) uncertainties as to population status, expressed as sampling variation in the monitoring of populations and habitats, is described. Adaptive management is framed in terms of maximizing long-term harvest value against a background of various kinds and degrees of uncertainty, with an emphasis on structural uncertainty. By an appropriate extension of the ''system state,'' adaptive optimization can be defined in terms of Markov decision processes. Solution algorithms are described for systems that are subject to structural uncertainty and are either partially or completely observable. Adaptive optimization is illustrated with an example in waterfowl harvest management that incorporates uncertainty in the relationship between harvest rates and survivorship.