Coastal aquaculture is a globally expanding enterprise. Currently, 1200 salmon farms operate in coastal Norway, yet their capacity to aggregate and subsequently modify wild fish distributions is poorly known. Aggregations of wild fish at 9 farms and 9 control locations were counted on 3 separate days in June to August 2007. On each sampling occasion, 6 counts were made at 5 distinct depth-strata at each farm and control location. Wild fish were 1 to 3 orders of magnitude more abundant at farms than at control sites, depending on the location. Gadoid fish (Pollachius virens, Gadus morhua and Melanogrammus aeglefinus) dominated farm-associated assemblages and were present across a wide range of sizes, from juveniles to large adults. Estimated total farmaggregated wild fish biomass averaged 10.2 metric tonnes (t) per farm across the 9 farms (range: 600 kg to 41.6 t). Applied across the geographical range of Norway's 1200 salmon farms, our estimates indicate that salmon farms attract and aggregate over 12 000 t of wild fish into a total of just 750 ha of coastal waters on any given day in summer. Possible consequences of these persistent, substantial aggregations of wild fishes at farms include a heightened potential for the transfer of pathogens from salmon farms to wild fish and among adjacent salmon farms, and altered availability of wild fish to fisheries. Restrictions on fishing in the immediate surrounds of salmon farms may avoid farms acting as ecological traps, particularly for species with depressed populations such as G. morhua, which are highly attracted to farms.

No abstract

Fernandez-Jover, D., Sanchez-Jerez, P., Bayle-Sempere, J. T., Valle, C., and Dempster, T. 2008. Seasonal patterns and diets of wild fish assemblages associated with Mediterranean coastal fish farms. – ICES Journal of Marine Science, 65: 1153–1160. Fish are attracted to floating structures, including coastal cage fish farms, sometimes in dense aggregations. To understand better the influence of aquaculture on wild fish stocks, we carried out seasonal visual censuses around three southwestern Mediterranean farms over 2 years to assess the temporal patterns of the aggregated fish assemblage. In addition, we analysed the diet of the five most abundant species. Aggregations around all farms were large throughout the year, although species composition and abundance differed among farms and seasons. Fish farms are attractive habitats for certain species of wild fish in specific seasons. Adult fish of reproductive size dominated the assemblages, and stomach content analysis revealed that 66–89% of fish of the five most abundant taxa had consumed food pellets lost from the cages. We estimated that wild fish consume up to 10% of the pellets used at farms, indicating that food is a key attractant. Regional monitoring of farm-associated wild fish assemblages could aid management of the interaction of aquaculture and wild fish resources, because changes in feeding behaviour may have consequences for fish populations and local fisheries.

BackgroundEcological traps form when artificial structures are added to natural habitats and induce mismatches between habitat preferences and fitness consequences. Their existence in terrestrial systems has been documented, yet little evidence suggests they occur in marine environments. Coastal fish farms are widespread artificial structures in coastal ecosystems and are highly attractive to wild fish.Methodology/Principal FindingsTo investigate if coastal salmon farms act as ecological traps for wild Atlantic cod (Gadus morhua) and saithe (Pollachius virens), we compared proxy measures of fitness between farm-associated fish and control fish caught distant from farms in nine locations throughout coastal Norway, the largest coastal fish farming industry in the world. Farms modified wild fish diets in both quality and quantity, thereby providing farm-associated wild fish with a strong trophic subsidy. This translated to greater somatic (saithe: 1.06–1.12 times; cod: 1.06–1.11 times) and liver condition indices (saithe: 1.4–1.8 times; cod: 2.0–2.8 times) than control fish caught distant from farms. Parasite loads of farm-associated wild fish were modified from control fish, with increased external and decreased internal parasites, however the strong effect of the trophic subsidy overrode any effects of altered loads upon condition.Conclusions and SignificanceProxy measures of fitness provided no evidence that salmon farms function as ecological traps for wild fish. We suggest fish farms may act as population sources for wild fish, provided they are protected from fishing while resident at farms to allow their increased condition to manifest as greater reproductive output.

Wild fish were counted in 4 depth-related strata (bottom, midwater, cage and surface) around 3 floating sea-cage fish farms (Altea, Campello, Guardamar) along the Spanish coastline and 2 farms (Los Cristianos, San Andrés) in the Canary Islands. Almost 200 000 wild fish belonging to 53 species were seen; representatives of Sparidae (8 species), Carangidae (6 species), Mugilidae (5 species) and Chondrichthyid rays (7 species) were commonly observed. At all 5 farms, 1 to 3 taxa accounted for > 74% of the abundance and > 94% of the biomass of aggregated wild fish. Abundances (5.7 to 162 times) and biomasses (42 to 1728 times) of wild fish were higher in the cage stratum at the 3 Mediterranean farms than at the bottom. In contrast, abundances and biomasses of wild fish at the Canary Island farms were highest at the bottom at San Andrés and highest at the surface at Los Cristianos. Large differences in the sizes of associated fish existed among farms, with low percentages of fish > 20 cm total length at Guardamar, Los Cristianos and San Andrés (11 to 25%) compared to Altea and Campello (81 to 95%). However, the greatest proportions of large fish were present in the cage stratum at each of the 5 farms. This variability suggests that assemblage structure and aggregated biomass of wild fish at farms cannot be predicted prior to their installation, causing uncertainty in modelling of nutrient dispersal. Further, the sedimentation 'footprint' of temperate seacage fish farms may vary depending on (1) the species and biomass of associated wild fish, and (2) where these fish are distributed in the water column. As wild fish consume lost feed and assimilate nutrients, we suggest that coastal managers prohibit fishing of large planktivorous species at farms to fully harness their ability to ameliorate benthic impacts.

Based on the fact that farmed fish experience different environments, stocking densities and feeding regimes compared with wild fish, several techniques have been developed to discriminate the wild or farmed origin of fish. These techniques quantify differences between genetics, chemical characteristics, fatty acid compositions, trace elements, pollutants, stable isotopes, morphology and organoleptic characteristics. Gilthead sea bream and European sea bass are the most important marine fish in Mediterranean aquaculture and are highly appreciated by commercial and recreational fisheries. A total of 60 studies that used techniques to discriminate farmed from wild fish for sea bream and sea bass form the basis of this review. The most common technique used differences in the lipid and fatty acid composition of fish. Many of these studies dealt with food science and product quality, rather than tracing escapees. A wide range of identification tools is useful in determining the correct origin of captures and proper labelling of marketed fish. External appearance and morphometry are useful for rapid assessments and can be achieved with high accuracy and little cost, especially for sea bream. This makes these methods suitable for detecting large and recent escape events, applicable in fisheries studies, and for ensuring that wild and farmed fish are separated in the marketplace. Techniques using differences in chemical or genetic composition are more useful for environmental monitoring, as they have higher accuracy and can detect escapees long after the escape incident. Regulatory bodies should legislate protocols that describe the technique(s) that must be applied in specific circumstances.

No abstract

Coastal aquaculture installations concentrate large numbers of wild ¢sh species of both ecological and economic importance, including schools of bogue, Boops boops (L.), in high abundance and biomass. The aggregated species consume large quantities of the easily available pellets lost from cages. As a consequence, the physical condition of farm-associated wild ¢sh is increased and their physiology is altered. These changes may in£uence local ¢sheries as many of these aggregating species are targeted by ¢shers. We assess whether local ¢shers catch wild ¢sh that have previously aggregated at ¢sh farms by comparing the body condition, trophic indexes, diet overlap and the fatty acid (FA) composition of B. boops obtained from local ¢sh markets and around ¢sh farms. Bogue captured by trammel-nets and around ¢sh farms facilities presented a similar biological condition, as well as high quantities of pellets in the gut, and their FA pro¢le was a¡ected by vegetalderived FAs. In contrast, bogue captured by trawlers were not in£uenced by ¢sh farms, as they consumed natural trophic resources.We conclude that artisanal ¢shers exploit these aggregated ¢sh populations once they have dispersed away from farms, and bene¢t from a 'biomass export' from ¢sh farms at a local scale.