Fish oil substitution with vegetable oils in diets for greater amberjack (Seriola dumerili): A consequential life cycle assessment approach

Abstract

[EN] This study assessed the environmental impacts of substituting fish oil with vegetable oils in the diets of greater amberjack (Seriola dumerili), based on both attributional and consequential life cycle assessments. The dietary lipid source was either fish oil (FO diet) or a blend of vegetable oils (linseed oil, sunflower oil, and palm oil, 4:3:3; VO diet). From an attributional perspective, the production of 1 kg of the FO diet released 1.10 kg CO2-eq and 1.60 kg CO2-eq as for global warming potential with and without land-use change, and 8.6 g PO4-eq as for the eutrophication potential; the depleted stock fraction index was equivalent to 2.8E-07. Compared with the FO diet, the production of 1 kg of the VO diet increased global warming potential with and without land-use change (by 5% and 4%, respectively) and PO4-eq by 58%, and reduced depleted stock fraction index by 24%. Conse-quential modelling revealed that in a scenario using fish oil and fishmeal derived from fish by-products, the marginal production of 1 kg of the VO diet compared with that of the FO diet produced a lower global warming potential with (-19%) and without land-use change (-12%), a greater eutrophication potential (+33%), and reduced depleted stock fraction index (-23%). In a scenario using fish oil and fishmeal diverted from a pre-existing fish production system, compared with that of the FO diet, the marginal production of 1 kg of the VO diet yielded similar values of global warming potentials and depleted stock fraction index, and a considerably higher eutrophication potential (+44%) value. For both FO and VO diets, fish-derived products (oil and meal) notably contributed to global warming potentials, as determined using both the attributional and consequential approaches. In contrast, the two diets differed with respect to the major contribution to eutrophication potential, which originated from non-fish-derived constituents for the FO diet and from vegetable oils (particularly sun-flower oil) for the VO diet. Sensitivity analysis indicated that the replacement of sunflower oil with an equal quantity of another oil, e.g. palm oil, could reduce the eutrophication potential of the VO diet. In conclusion, the substitution of fish oil in aquafeeds with a mixture of vegetable oils could contribute to reductions in the exploitation of wild fish stocks and greenhouse gas emissions, although potential trade-offs, such as increased eutrophication, should be taken into consideration to enhance the environmental sustainability of aquafeed production based on an integrated approach.