Reproduction

Genetic control of canalization for twinning and triplet rates in Polypay sheep using double hierarchical generalized linear models

---

Abstract

For sheep flocks that rely on the natural rearing of lambs, the ideal ewe should give birth to and raise twins. However, breeders are often reluctant to select for increased litter size because of associated increase in triplet and higher-order births, which contributes to the persistently low U.S. lambing percentage (106%). To assist breeders in achieving a more desirable reproductive profile, the objective of this research was to develop a strategy to identify rams whose daughters consistently produce twins. This requires identifying sires that produce a consistent phenotype regardless of genetic and environmental conditions, a concept referred to as canalization. To increase twinning rate, the mean number of lambs born (NLB) needs to be maintained at two while simultaneously placing downward pressure on triplet percentage. Data from 1999 to 2025 from the USDA-ARS-U.S. Sheep Experiment Station Polypay population were used for the analyses. The dataset included 6,256 litters from 2,600 mature ewes and 241 rams. Despite selection of twin-born rams as sires for the flock, the phenotypic trend in twins has significantly decreased (P=0.03) over time, confirming previous reports that selecting twin-born rams alone is not effective in increasing the percentage of twins in the flock. When considering sires with daughters totaling at least 15 litters (n=158), the mean NLB ranged from 1.7 to 2.8 while the percentage of twin litters ranged from 7 to 80%, indicating considerable phenotypic variation in the population. A double hierarchical generalized linear model (DHGLM) was used to estimate both the mean and dispersion of the NLB, twinning (TWIN), and triplet (TRIP) traits using the DHGLMF90 package. A multivariate animal model was fitted for NLB (1, 2, or 3), TWIN (1 for singles or triplets and higher order and 2 for twins) and TRIP (2 for singles or twins and 3 for triplets and higher order) and their dispersions (vNLB, vTWIN, and vTRIP) with contemporary group (CG) as a categorical fixed effect and animal (ewe), permanent environment, and the residual as random effects. The CG included lamb birth year and ewe age (2 or 3+) but not flock or season because this dataset consists of a single flock with one reproductive season. The heritability estimates were generally low (0.015 to 0.065) and higher for the dispersion traits than the mean traits. The genetic correlations between NLB, TRIP, and TWIN were high; therefore, selection for only one trait is needed. Similarly, the genetic correlations between vNLB, vTRIP, and vTWIN were high, suggesting selection for one trait will be effective. The negative genetic correlation (-0.255) between TWIN and vTWIN suggests that selecting simultaneously for higher TWIN and lower vTWIN may maintain a twinning rate near two while decreasing the variation associated with twinning and constraining triplet frequency.

Keywords: 2026

How to Cite:

Wilson, C., Taylor, B. & Brito, L., (2026) “Genetic control of canalization for twinning and triplet rates in Polypay sheep using double hierarchical generalized linear models”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2294415. doi: https://doi.org/10.31274/wcgalp.24320

Rights: 1