Benthic impacts of fish farming: meta-analysis of community and geochemical data.

A number of 41 papers dealing with the benthic effects of fish farming were reviewed and the values of the variables studied were extracted to be used in a meta-analysis of effects. The papers used covered a wide range of farmed species, geographic regions, management practices and specific site characteristics (e.g., depth, exposure, and sediment type). Therefore, the total data-set may not be considered as biased towards a particular set of conditions as is often the case with data collected in a single study. More than 120 biological and geochemical variables were monitored, occasionally using different sampling and analytical protocols for the same variables. The rank correlation analysis between all possible pairs of variables in the data set showed a large number of significant positive or negative correlations, reflecting the response of these variables to benthic organic enrichment. The use of stepwise regression showed that most biological and geochemical variables are determined by a combination of distance from the farm with bottom depth and/or latitude. Results of stepwise regression, repeated separately for each type of sediment, showed that although the general pattern was similar among different types of sediments, the coefficients varied considerably indicating changes of the distance affected by settling particulate organic material for different sediment types. The overall conclusion is that the complicated interactions between variables and the lack of data, such as current speed, induce difficulties in setting common or uniform environmental quality standards for benthic effects of fish farming and these should take into account the existing considerable differences between geographic regions, depth zones and sediment types.