Abstract
Sampling theory is essential also in fisheries science, where estimates of mean, total and ratio are often required, for example, total spawning, total landings, population density and rate of matured individuals. However, sampling procedures have great variation in practice, and cases for which we cannot find the appropriate estimators and variances in sampling text books are common. This paper treats such cases in estimating population total for recreational catch and commercial landings. Estimators for totals and ratios are derived with respective variances for a single-stage cluster, a two-stage and a three-stage simple random sampling without replacement when the sizes of the sampling units are unequal. Estimators derived here are general and can be applied for various situations. As case studies, data from an ayu recreatioal fishing and a masu salmon stock enhancement program were analyzed and sampling strategies were also discussed.
