Effect of life cycle inventory choices and nutritional variation on carbon footprint of broiler meat production

The goal of the current study is to identify the effect of life cycle inventory (LCI) specifications on carbon footprint of feed used to produce one kg of broiler meat in multi-treatment trials. A total of 384 one-day Hubbard JA787 male chicks were allocated equally to 48 pens. The pens were distributed randomly to three treatments (16 pens/treatment), high density (crude protein (ME): 23% (12.5 MJ/ kg),20.4% (13 MJ/kg), 19.1% (13.7 MJ/kg) for starter grower and finisher diets, respectively), medium density (crude protein (ME): 22.3% (12.2 MJ/kg), 19.7 (12.6 MJ/kg),18.2% (13 MJ/kg) for starter grower and finisher diets, respectively) and low density (crude protein (ME): 21.6% (11.9 MJ/kg), 19 (12.1 MJ/kg),17.3% (12.4 MJ/ kg) for starter grower and finisher diets, respectively). The experimental birds had free access to water and feed throughout the trial. All experimental diets were in crumb form in starter phase and in pellet form in grower and finisher phase. Two birds/pen were euthanised at the end of the trial (day 42) to obtain meat yield. Feed consumption and bird weight of each pen were recorded weekly. The carbon footprint (CF) of feed consumed by each pen was calculated and normalised for one kg meat yield. The CF of the pens was calculated using 3 open access databases (Agribalyse (ReCiPe (H) impact assessment method and mass allocation), GLFI (ReCiPe 2016 midpoint (H), mass allocation) and ECO-ALIM (CLM impact assessment method and mass allocation). The three CF datasets were combined and analysed to determine the effect of treatment, LCI database and treatment*LCI database on CF of broilers. The CF of each pen was calculated using GLFI database according to three allocation methods, mass, energy and economic. Data of the three allocation methods were combined and analysed to identify the effect of treatment, allocation method and treatment*allocation method on CF. Agribalyse dataset was used to calculated CF of each pen according to 3 impact assessment methods, ReCiPe (i), IPCC 2013 and CLM (with AI as a baseline). The three datasets were combined and analysed to determine the effect of the treatment, impact assessment method of LCA and treatment*assessment method on CF. There was significant effect of treatment, database and treatment*database on CF (P < 0.001). The CF was significantly affected by treatment, impact assessment method and the treatment*impact assessment method (P = 0.037). Treatment and treatment*allocation method significantly affected the CF (P < 0.001). In conclusion, the effect of the dietary treatment on CF depends on the LCI database, impact assessment method and method of allocation. Thus, LCI specifications should be presented in detail in broiler trials aiming at identifying of the effect dietary treatment on CF of broilers.