Adaptive Sampling

Abstract

Animal populations are often highly clustered. For example, fish can form large, widely scattered schools with few fish in between. Even rare species of animals may form small groups that are hard to find. Applying standard sampling methods such as simple random sampling of plots to such a population could yield little information, with most of the plots being empty. Adaptive sampling is based on the simple idea that when some animals are located on a sample plot, the neighboring plots (and possibly their neighbors as well) are added to the sample; in this way, the whole cluster can be sampled. The scheme lends itself to a variety of modifications, for example, sampling with and without replacement, sequential and inverse sampling, stratified sampling, two‐stage sampling, and so‐called adaptive allocation, as well as extensions to multivariate data and incomplete detectability.