Uncertainty, learning, and the optimal management of wildlife
Environmental and Ecological Statistics
By: B.K. Williams
Links
- More information: Publisher Index Page (via DOI)
- Download citation as: RIS | Dublin Core
Abstract
Wildlife management is limited by uncontrolled and often unrecognized environmental variation, by limited capabilities to observe and control animal populations, and by a lack of understanding about the biological processes driving population dynamics. In this paper I describe a comprehensive framework for management that includes multiple models and likelihood values to account for structural uncertainty, along with stochastic factors to account for environmental variation, random sampling, and partial controllability. Adaptive optimization is developed in terms of the optimal control of incompletely understood populations, with the expected value of perfect information measuring the potential for improving control through learning. The framework for optimal adaptive control is generalized by including partial observability and non-adaptive, sample-based updating of model likelihoods. Passive adaptive management is derived as a special case of constrained adaptive optimization, representing a potentially efficient suboptimal alternative that nonetheless accounts for structural uncertainty.
| Publication type | Article |
|---|---|
| Publication Subtype | Journal Article |
| Title | Uncertainty, learning, and the optimal management of wildlife |
| Series title | Environmental and Ecological Statistics |
| DOI | 10.1023/A:1011395725123 |
| Volume | 8 |
| Issue | 3 |
| Year Published | 2001 |
| Language | English |
| Larger Work Type | Article |
| Larger Work Subtype | Journal Article |
| Larger Work Title | Environmental and Ecological Statistics |
| First page | 269 |
| Last page | 288 |
| Google Analytic Metrics | Metrics page |