Integrated multi-trophic aquaculture driven by nutrient wastes released from Atlantic salmon (Salmo salar) farming

Abstract

Norwegian aquaculture grew from 151,000 tonnes in 1990 to more than one million tonnes in 2010, becoming the world’s leading producer of farmed salmon. The expansion of aquaculture has raised some concerns about the loadings of dissolved and solid wastes and their effects on the marine environment. Integrated multi-trophic aquaculture (IMTA) has been suggested as a promising, ecologically sound means of providing mutual benefits to the co-cultured species while at the same time mitigating the ecological effects of the wastes discharged from fish farms.

The main objective of this thesis was to study the feasibility of integrating macroalgae (Saccharina latissima) and blue mussels (Mytilus edulis) with salmon farming in Norwegian coastal waters. The first step in this study was estimating the release rate of the nutrient wastes on a national scale using a mass balance model with coefficients mainly taken from the literature. The next step was measuring the chemical compositions of salmon feed, fish tissues and salmon faeces from a single salmon farm. This field data was used to revise the coefficients in the mass balance model in order to obtain more realistic estimates of the release rate of nutrient wastes both at the national and single salmon farm scales. The study also allowed an evaluation of the nutritional values of the nutrient wastes for mussels and macroalgae. The last step involved investigating the assimilation of salmon farm wastes in S. latissima and blue mussels and estimating the potential production of IMTA.

On a national scale, 70 % of feed C, 62 % of feed N and 70% of feed P were lost into the environment. These are equivalent to total annual discharges of about 404 000, 50 600 and 9400 tonnes of C, N and P, respectively. The major C wastes were respired CO2 and the major N wastes were excreted dissolved inorganic nitrogen (DIN), whereas the major P wastes were associated with feed and faeces particles. The recalculated release rate of C, N and P wastes on a national scale using the revised coefficients showed that the total N and the DIN wastes were decreased by 23% and 25%, respectively. The total P wastes were decreased by 16%. These results indicated that the discharge of nutrient wastes was reduced mainly because the composition of the feed was optimised to better retain proteins in the feed.

Estimates on a single salmon farm scale showed that 62 % of feed C, 57 % of feed N and 76% of feed P were lost into the environment. This means 38% of feed C, 43 % of feed N and 24% of feed P were retained in salmon biomass. The C content of faeces was still high whereas the N content of faeces was half that of the feed. The P content of faeces was higher than that of the feed. The lipid, DHA and EPA contents of faeces were far lower than that of the feed, but were comparable to those of some phytoplankton species. These results indicated preliminarily that both salmon feed and faeces can be nutritionally adequate food resources for some filter and deposit-feeder species considering their nutritional values.

The model-predicted N:C ratio for faeces fitted well with the measured value, while there were significant differences in the P:C and N:P ratios between the predicted and measured values. The assumptions for C (80%) and N (85%) digestibility of the feed in the mass balance were fairly robust, while the P (50%) digestibility was uncertain. An exercise revealed that the predicted and measured P:C and N:P ratios of the faeces were almost equal if the input P digestibility to the model was set to 30%, suggesting that the P digestibility of the feed might be as low as 30%.

S. latissima at the salmon farm stations grew faster (p<0.05) than at the reference station. The increased N supply did not result in N accumulation in S. latissima at the salmon farm station due to the dilution caused by increased growth. The N isotope ratio (δ15 N) of S. latissima was higher at the salmon farm station than at the reference station from April to June (p<0.05). These results suggested that the origin of the N in the S. latissima was partly from the salmon farm.

Mussels at the Farm West station (FW) grew faster in the spring (p<0.05), and slower in summer compared to those at the reference station (RS). The soft tissue content of mussels at the salmon farm stations was significantly higher than at the RS during autumn and winter, while it was lower in June (p<0.05) because of spawning. The percentage of 18:1 (n-9) of total fatty acids in mussel tissues increased from June to August and February; the increase in 18:1 (n-9) was more pronounced in February than in August. These results suggested that mussels grown in proximity to salmon cages assimilated a part of the salmon wastes, showing a higher soft tissue content during autumn and winter and higher growth rate in spring.