Application of adaptive sampling in fishery part 1: Adaptive cluster sampling and its strip designs

Abstract

Abstract: The precision of conventional sampling designs is not usually satisfactory for estimating parameters of clump and rare populations. Many of fish species live in school and disperse all over a vast area like a sea so that they are rare compare to their habitats. Theory of a class of sampling designs called adaptive sampling designs has rapidly grown during last decade which solved the mentioned problem. These sampling designs are successfully used to asses fish life stocks in number of cases. The first aim of this article is to develop some versions of adaptive sampling which are more useful for fishery study. The author will introduce Adaptive Cluster Sampling (ACS) and its strip version which is more useful for fishery and marine studies. Since the parameter of interest is usually the density in fishery study, estimators of the density are developed in this article. In many biological studies, it has been shown that non-overlap two stage adaptive sampling and strip adaptive sampling are very efficient sampling designs so that we also develop non-overlap two stage adaptive cluster sampling of strips. The second aim of this article is to introduce adaptive cluster sampling to our fishery researchers' community and demonstrate the idea and computation to those who know statistics in some level.