Ecological effects of aquaculture on living and non-living suspended fractions of the water column: a meta-analysis

The effects of aquaculture on the ecology of the water column have been extensively studied in the last two decades. However, to date, it has not been possible to extrapolate homogeneous information from the peer-reviewed literature. In the present study, 68 peer-reviewed articles were analysed and about 1087 study cases were used to test whether worldwide cultivations of aquatic organisms (shrimps, fish, bivalves and polyculture) have a differential effect on living and non-living fractions of the water column (suspended matter, chlorophyll-a, particulate organic carbon, nitrogen and phosphorus, bacteria and plankton). Meta-analysis feasibility depends on obtaining an estimate of the effect size from every study, and the most common measure of effect size (Hedges'd) is the difference between means of controls and impacts standardised by dividing by the pooled standard deviation. Shrimp, fish and bivalve cultivation differentially affected water column dynamics, with a general major impact on bacteria and phytoplankton. In addition, results showed that the water column dynamics are probably affected by organic aquaculture loading but, due to the substantial heterogeneity across studies, the information available on the effects can be considered partially flawed and therefore not sufficient to either support or exclude the notion that different forms of aquaculture affect ecological processes of the water column.

A meta-analysis on the ecological effects of aquaculture on the water column: dissolved nutrients

Environmental effects of aquaculture loadings have often been reviewed descriptively, and thus have not provided quantitative estimates of the overall response in the water column. Meta-analytical reviewing techniques allow the contextualisation of quantitative effects in the domain of current literature. In the present paper, more than 50 peer-reviewed articles were analysed and about 425 study cases used to test whether worldwide cultivations have a differential effect on dissolved nutrient levels. Meta-analysis feasibility depends on obtaining an estimate of the effect size from every study and the most common measure of effect size (Hedges' d) is the difference between means of controls and impacts standardised by dividing by the pooled standard deviation. Across all study cases, irrespective of cultivation and organism type, the cumulative effect size was large and significant (d>0.8) for ammonium, nitrite and nitrate, medium (0.8>d>0.5) for dissolved phosphorus, and not significant (d<0.2) for silicates. Effects were mainly correlated with the degree of openness in water bodies, and ammonium and the other nitrogen forms were the most highly informative descriptors of effects in the area surrounding farms, even though weakness in statistical approach was highlighted. The results partially contradict the common view that effects of aquaculture and associated environmental patterns are well defined throughout the current literature.