Greenhouse gas emissions of black soldier fly larvae grown on different feed substrates throughout larval development

Protein derived from black soldier fly larvae (BSFL) is considered a valuable alternative diet component in aquaculture and poultry feeding. Through more efficient nutrient recycling from agrifood waste and mitigating overexploitation of limited marine and terrestrial resources a twofold sustainability increase may be provided. Yet, imperative for overall life cycle assessments, more detailed data on environmental impact criteria of BSFL production, e.g. greenhouse gas emissions, are needed. In open circuit respiration chambers running at 27.5°C and 50% relative humidity 7 days old BSFL were incubated for 14 days during which four feeding events took place. Six identical preconsumer food waste compo-nents were mixed in mirrored relative proportions of 1:3 for individual components in order to obtain two feed substrates representing a high fibre (A) versus a well digestible nutrient-rich feedstuff (B), respectively. Both feed substrates contained equal dry matter contents and were provided in equal amounts in each feed treatment. Three chamber replicates each comprising of triplicated boxes of 10’000 BSFL were monitored per feed treatment, including continuous measurements of CO2 and CH4 every minute, as well as N2O 4–8 times per day. Prior to each feeding event and at harvest larvae were separated from residue material (frass and non-digested feed remains) to assess phase-specific biomass gain. Separations that took place before harvest were temporary only, and larvae were placed back to their respective residue material again. While feed B yielded higher BSFL biomass production and lower feed conversion ratios, larvae of both feeding treatments survived equally and entered the prepupal stage by the majority within 14 days. All gas emission profiles were strikingly affected by feeding events for both feed substrates, resulting in tem-porarily extreme peaks. Also similar for both feed substrates individual developmental phases of BSFL exhibited quite characteristic responses, with strongly diverging patterns of decreasing biomass while maintaining substantial emissions towards the 6th larval instar. Overall, CO2 emission profiles between feeding treatments were similar. Feed A resulted in higher CH4 emissions, while higher N2O emissions were detected for feed B. CH4 and N2O emissions during BSFL rearing may primarily coincide with particular conditions of type of feed, substrate depth and moisture, excess nutrients and associated microbiota. Total CO2equivalents were high and comparable to conventional livestock biomass production. Therefore, context-specific considerations of feed substrates and rearing regimes, and particularly larval growth period until harvest may positively influence outcomes of comprehensive sustainability assessments of a given BSFL feed protein production