AbstractBackgroundThe goal of managing genetic diversity is to maintain both the genomic uniqueness and adaptability of a population. This entails both maintaining heterozygosity and avoiding deleterious alleles from drifting to higher frequencies. Thus, genomic management strategies should not systematically change frequencies at neutral loci, as has been observed in current genomic management. We simulated 50 replicates of a population managed with optimal contribution selection (OCS), where inbreeding was controlled using different relationship matrices, where Van Raden method I (GVR), runs-of-homozygosity (GROH), and identity by descent relationships (IBD) only used genomic information and a linkage analysis-based matrix (GFGLA) leveraged both genomic and pedigree information.ResultsAcross 50 simulated replicates, GFGLA achieved the highest correlation with true IBD for both inbreeding and coancestry estimates and provided the most accurate rate of inbreeding (ΔF). GFGLA also delivered the greatest genetic gain per unit of inbreeding and was the only scheme to remain below the inbreeding target while minimizing frequency change at neutral loci. GVR using base population frequencies produced high genetic gain and controlled drift effectively but introduced systematic allele frequency changes towards the closest extreme frequency. Using current allele frequency resulted in high levels of inbreeding compared to the genetic gain achieved. GROH showed strong dependence on the minimum ROH length: short ROH improved ΔF estimation but increased allele frequency changes toward intermediate values, whereas long ROH underestimated inbreeding and reduced genetic gain.ConclusionsIBD-based matrices such as GFGLA offer the most balanced approach for sustainable population management in OCS schemes, combining accurate relationship estimates, effective inbreeding control, and neutrality at neutral loci. In contrast, GVR and GROH require calibration and cause systematic allele frequency changes.