| Abstract: | Abstract. In this article, we compare biomass results for three fish species (Symphodus ocellatus, Serranus scriba and Diplodus annularis) between May 1993 and May 2000 in the Bonifacio Strait Marine Reserve. The data were obtained from seagrass meadows using the visual census method. A randomised procedure was used to test the null hypothesis of biomass equality. To this end, decreasing sample sizes were used for the May 2000 data to determine the sample size necessary to reject the null hypothesis (5 % level). For Symphodus ocellatus, 10 samples (fixed points) were necessary to reject the null hypothesis, whereas 15 samples were required for Diplodus annularis. The null hypothesis was never rejected for Serranus scriba regardless of the sample size considered.These results are discussed, as is the power statistical analysis proposed by Ortiz (2002; this issue, pp. 1 – 9) , to determine the optimal sample size necessary to detect perturbation effects in the marine environment. Although the power analysis is certainly a useful method to determine the sampling design in assessment the effect of marine reserves on fish biomass, the magnitude of the marine reserve effect remains problematic in ecology. |
