Model uncertainty versus variability in the life cycle assessment of commercial fisheries

Abstract

Results from life cycle assessment (LCA) studies are sensitive to modeling choices and data used in building the underlying model. This is also relevant for the case of fisheries and LCAs of fish products. Fisheries' product systems show both multifunctionality because of the simultaneous co-catch of multiple species and potential constraints to supply due to natural stock limits or socially established limits such as quota systems. The performance of fisheries also varies across seasons, locations, vessels, and target species. In this study, we investigate the combined effect of modeling choices and variability on the uncertainty of LCA results of fish products. We use time series data from official Danish statistics for catch and fuel use of several fisheries disaggregated using a top-down procedure. We apply multiple modeling approaches with different assumptions regarding the type of partitioning, substitution, and constraints. The analysis demonstrates that, in the presence of relevant multifunctionality, the results are substantially affected by the modeling approach chosen. These findings are robust across years and fisheries, indicating that modeling choices contribute to uncertainty more than the variability in fishing conditions. We stress the need for a more careful alignment of research questions and methods for LCA studies of fisheries and recommend a very transparent statement of assumptions, combined with uncertainty and sensitivity analysis. This article met the requirements for a gold-gold data openness badge described at http://jie.click.badges.