A genomic prediction framework for growth traits in black soldier fly (Hermetia illucens) breeding programmes

Genomic prediction for four growth traits in farmed Hermetia illucens is strongest for larval body weight (LBW), with performance varying by diet, model, and SNP density.

• Why it matters: Genomic selection in BSF is not yet established, and feed-source diversity means breeding strategies must work across nutritional environments with reliable marker and model choices.
• What they did: They reared 2,083 larvae from the base population on SYK, BSG, and FVW diets and 536 from the F7 population on chicken starter feed, then phenotyped LBW, LL, LW, and LSA and genotyped 4,680 SNPs. Seven genomic prediction methods were tested under additive (A) and additive-dominance (AD) models, including across- and within-diet effects, G×D, reference vs test generations, and SNP density changes.
• The result: Dominance often improved predictive ability, multivariate G×D models matched or slightly exceeded within-diet predictions, and cross-diet prediction was more variable and generally lower; predictive ability stayed comparable at 3,000 SNPs but dropped below that, and declined across generations.

AI summary of the abstract · may contain errors · not medical advice · read the paper.