Abstract:
|
All credible information should be used when building an environmental sampling design for population status. In particular habitat association models can help in solving difficult sampling questions such as stratification and sample allocation. However, these models often lack a term for spatial correlation. Spatial correlation can help decide on questions such as size of sampling unit and sampling frequency in systematic sampling. We propose a Markov habitat association model for spring chinook salmon in the Middle Fork Salmon River in Idaho. Monte-Carlo Markov chain provides estimates for the sampling distribution of statistics, which in turn provide guidance for sampling decisions. In particular, we address the following questions: (1) How long should the sampling units be?; (2) Should we stratify by habitat characteristics and which ones?; (3) Systematic, adaptive, or simple random sampling?; (4) If systematic, what sampling frequency should be used?; (5) If adaptive, how should we decide to terminate adaptive sampling? These methods are briefly compared to the traditional design-based sampling methods.
|