From residual biomass to novel feed ingredients for pigs, broilers and fish: a methodology for integrating feed formulation and consequential life cycle assessment

Life cycle assessments (LCAs) of residual biomass for animal feed often highlight environmental 4 benefits from substituting high-impact, protein-rich conventional feed ingredients, but either neglect 5 the full consequences of this substitution by using attributional LCA or treat nutritional constraints and 6 animal requirements as a black box, assuming uniform nutrient needs across species. To address these 7 limitations, the potential environmental benefits of incorporating new ingredients into pigs, sows, 8 broilers and fish compound feeds were evaluated using by integrating market-based feed formulation 9 and consequential LCA. Four new ingredients were assessed and compared to a reference feed: (1) 10 single-cell protein from bacteria fed methane from manure anaerobic digestion (SCP-ADM), ( 2) 11 mycoprotein from fungi grown on wheat straw-derived fermentable sugars (MYCP-WS), (3) green 12 protein concentrates from sugar beet leaves (GPC-SBL), and (4) black soldier fly larvae reared on food 13 waste (BSF-FW). Among the four ingredients investigated, SCP-ADM obtained the highest 14 environmental benefits in all animal species. Three mechanisms explain the higher benefits of SCP-15 ADM compared to the other scenarios: the new ingredient's unit impact, its inclusion rate and the 16 eventual soybean meal reductions. The effect of the latter two, here formalized as the 'displacement 17 effect', was compared with the displacement effect obtained using conventional substitution proxy 18 used in LCA. It showed that the proxy consistently overestimated climate change savings from 14% to 19 38%. Overall, the key methodological strength of the proposed feed formulation-LCA framework lies 20 in estimating the displacement effect consequences in function of animal-specific feed formulation 21 and specific market prices rather than relying on arbitrary allocations or generalized marginal 22 assumptions.