Environmental assessment of trout farming in France by life cycle assessment: using bootstrapped principal component analysis to better define system classification. (15th January 2015)

• Record Type: Journal Article
• Authors: Chen, Xiaobo; Samson, Elisabeth; Tocqueville, Aurélien; Aubin, Joël
• Is Part Of: Journal of cleaner production. Volume 87(2015:Jan. 15)
• Page Start: 87
• Page End: 95
• Publication Date: 2015-01-15
• Subjects: Life cycle assessment -- Non-parametric bootstrap -- Principal component analysis -- Trout farming system; Factory and trade waste -- Management -- Periodicals; Manufactures -- Environmental aspects -- Periodicals; Déchets industriels -- Gestion -- Périodiques; Usines -- Aspect de l'environnement -- Périodiques; 628.5
• DOI: 10.1016/j.jclepro.2014.09.021
• Languages: English
• ISSNs: 0959-6526

Abstract

Abstract: Trout farming is the main fish production system in France. This article describes a system to classify trout farms based on environmental impacts calculated by life cycle assessment and technical and economic indicators. Since the number of surveyed farms was too small for a robust assessment, we combined principal component analysis (PCA) with a non-parametric bootstrap technique. French trout farms were surveyed to collect technical and economic indicators. The representativeness of the survey was verified by comparing it to a national inventory. Life cycle assessment was used to estimate environmental impacts of farms and the contribution of each production stage to impacts. PCA was used to evaluate both technical-economic and environmental indicators of the trout farms, which were separated into three groups based on the size of fish produced (pan-size, large and mixed-size, and very large). Non-parametric bootstrap was used to compare the groups and to test the significance of PCA results. Results validated the fish-farm classification system based on the size of fish produced and indicated that farm operations and fish feeding contributed the most to environmental impacts. The PCA method distinguished three groups via their technical indicators, with non-significant differences among the groups in environmental impacts. However, environmental indicators showed strong links with technical and economic indicators. In conclusion, bootstrapped PCA offers the ability to assess groups of trout production system when the sample size is too small and provides more conservative results by considering uncertainty. Future studies should focus on providing reliable data to reduce uncertainty. Graphical abstract: Highlights: Sample size is a major concern to obtain a robust environmental characterization. Bootstrap sampling can estimate uncertainty due to small sample size in PCA result. Bootstrapped PCA was used to build a typology of trout farms. Technical indicators influence more the trout farm typology than the LCA impacts.