Comparison of pedigree genetic evaluation methods for black soldier fly breeding

This study describes comparison of different methods for pedigree genetic evaluation in a black soldier fly (BSF)
breeding programme. Tracking individual insects and their parentage is resource intensive due to their small size
and large number. However, some traits can be phenotyped on a per individual basis and pedigree between expected
full-sib groups can be established, albeit with uncertain parentage. Here we present three alternative methods of
pedigree genetic evaluation in such a setting. In the first method we estimate simplified additive relationship between
groups of individuals – ignoring the within group variation. In the second method we estimate relationship between
individuals and allow for uncertain parentage. In the third method we assume that the pedigree is known. To compare
these methods, we have conducted a life-cycle analysis of the BSF and simulated a breeding programme with the
R package AlphaSimR. The simulation generated individual genomes, breeding values and phenotypes influenced
by breeding value, management, year-season and other environmental effects. Using these data, we compared the
different methods on how accurately they estimate variance components with airemlf90 and breeding values with
blupf90. We further used the estimated breeding values to drive selection decisions and compared in silico achieved
genetic gain and depleted genetic variance over generations. This work will guide the development of state-of-the-art
breeding programmes for BSF.