Title: Evaluation of the sampling scheme for Chub Mackerel (Scomber japonicas, Houttuyn, 1782) in the inshore fishery in Ghana

Management decisions in fisheries are largely based on estimates from fisheries data collection programmes. The aim of a sampling strategy is to sample a catch to mirror the population of interest. This study was conducted using both an ANOVA (Analysis of Variance) and a block bootstrapped approach to evaluate a preferred sampling scheme to collect length frequency data of chub mackerel (Scomber japonicus) from the inshore fishery in Ghana. An extensive data set on haddock (Melanogrammus aeglefinus) in Icelandic waters was use to evaluate and develop the sampling scheme for chub mackerel. In general the study showed that for a gain in precision in a sample, the number of samples is more important than the number of measurements in each sample. Thus to get a representative sample of the population; the sampling should be spread out, as fish caught in clusters contain less information about the population than an equal number caught at random. It is suggested from the study that due to constraints by logistics and cost of sampling in the fishery, sample size of 30 with 30 individual measurements in a sample should be considered the absolute minimum number of samples in a sampling scheme for chub mackerel annually.