A comment on the methodology of stocking experiments

Stocking density can have an important influence on the success of aquaculture operations. However, various aspects of the methodology of published experiments on stocking density appear problematical. For example, size variation in the system under study may have determinant effects on yield, survivorship, and estimates of the optimal stocking density. To document whether this might be a source of concern in scientific advice to aquaculturists, I reviewed all stocking experiments published in Vol. 1 through Vol. 224 of Aquaculture and checked whether information about size variation had been reported in sufficient detail, whether size-sorting had occurred and whether groups of experimental individuals and cultured individuals had similar size variability. This was done for experiments on crustaceans and finfish (the corresponding analysis for bivalves is presented in Frechette et al. (2005) [Frechette, M., Alunno-Bruscia, M., Dumais, J.-F., Daigle, G., Sirois, R., 2005. Incompleteness and statistical uncertainty in competition/stocking experiments. Aquaculture 246, 209-225]). The analysis clearly shows that the issue of size variability of test cage populations has been largely overlooked in a large proportion of experiments. The study raised other sources of concern. For instance, allometry was overlooked in a significant proportion of experiments on crustaceans (55%) and finfish (25%), which caused ration level to be adjusted following inappropriate principles. The issues of replacement/removal of individuals and of the dimensions of the variable used to measure stocking level were also examined. Inappropriate procedures were found in a number of cases. Suggestions are made to deal with these issues in a more acceptable way. These problems, which may lead to serious flaws in advice to growers, were not restricted to papers published in Aquaculture. Crown Copyright (C) 2005 Published by Elsevier B.V. All rights reserved.