Robust Abundance Estimation in Animal Surveys with Imperfect Detection

• Activity Number: 667
• Type: Contributed
• Date/Time: Thursday, August 2, 2012 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Thomas Bohrmann*+ and Mary C Christman
• Companies: EPA and MCC Statistical Consulting LLC
• Keywords: abundance ; estimation ; model-assisted

Abstract

Surveys of animal abundance are central to the conservation and management of living natural resources. However, detection uncertainty complicates the sampling process of many species. One sampling method employed to deal with this problem is depletion (or removal) surveys in which animals are sequentially removed (and not replaced) from a closed subunit of the population. Information obtained in such a survey is translated into estimates of total population abundance via a statistical abundance estimator, of which there are many choices. Abundance estimators are generally either from the class of design-based estimators or model-based estimators, each having strengths and weaknesses. In this talk we present a new hybrid abundance estimator which draws from the strengths of both model-based and design-based estimators. We show that the coherent combination of these two frameworks yields a useful, flexible and yet robust total abundance estimator. We apply the estimator in the context of a simulation study based on annual depletion surveys of Chesapeake Bay blue crab abundance, and we compare the performance of our hybrid estimator with a typical fully model-based estimator.