Genetic gain & Inbreeding

Impacts of effective population size and genomic selection on genetic gain and between replicate variance using a simulated swine breeding program

---

Abstract

Effective population size (Ne) is a critical parameter in breeding program design, influencing both the rate of genetic gain and the maintenance of genetic diversity. While larger Ne values help preserve genetic variation and reduce inbreeding, they can reduce selection intensity. Genomic selection (GS) has revolutionized animal breeding by enabling more accurate selection decisions earlier in life, but its interaction with Ne level on genetic gain and variance in response (i.e. variation in TBV means across replicates) requires investigation. The objective of this study is to determine how Ne and GS jointly affect genetic gain and variance of response in a simulated swine breeding program. Stochastic simulations were conducted using AlphaSimR with 50 replicates. Eight scenarios were tested combining four Ne levels (20, 50, 100, 200) with and without GS. Each scenario maintained 300 sows per generation with 6 offspring per litter. Boar numbers were adjusted to achieve target Ne values (5, 13, 27, and 60, respectively). Three generations of random selection were utilized followed by seven generations of index selection (5 traits). Only means from generation 10 were utilized for modeling and inferences. Linear models with Ne, GS, and their interaction were fitted in R. Variance in response across scenarios was evaluated using variance ratio tests (F-tests). Mean index values ranged from 555 to 707, with the largest absolute response observed for Ne=20 with GS. Genomic selection consistently increased genetic gain at each Ne level as expected. The benefit of GS decreased as Ne increased. An Ne x GS interaction indicated that under pedigree-based selection, the greatest response occurred at Ne=50, not Ne=20. Pairwise comparisons of means revealed significant differences (p < 0.05) between most pedigree and GS scenarios, however genetic gain did not significantly decline between Ne=20 and Ne=100 for both GS scenarios. Standard deviation in response decreased with increasing Ne, regardless of GS implementation. At Ne=20, standard deviations across replicates were approximately 72-89, declining to 36-44 at Ne=200. The pairwise F-tests of variance showed significant differences (p < 0.05) between many scenario pairs. However, between pedigree and GS scenarios at each level of Ne, no statistical differences were detected with 50 replicates. Therefore, variance in response is driven by Ne with current parameters. While smaller Ne values combined with GS maximize short-term genetic gain, they also increase variance across replicates and may compromise long-term genetic sustainability. In conclusion, this research supports larger Ne values compared to historic Ne values (e.g. moving from ~50 to ~100) with GS to provide a favorable balance between genetic progress and response variance reduction, supporting more predictable and sustainable breeding outcomes.

Keywords: 2026

How to Cite:

Putz, A., Jamieson, D., Bink, M. & Huisman, A., (2026) “Impacts of effective population size and genomic selection on genetic gain and between replicate variance using a simulated swine breeding program”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2286248. doi: https://doi.org/10.31274/wcgalp.24577

Rights: 1