thumbnail

Uncertainty, learning, and the optimal management of wildlife

Environmental and Ecological Statistics

By:
DOI: 10.1023/A:1011395725123

Links

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.

Additional Publication Details

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
Number of Pages:
20