Dairy cattle

Integration of external data into Uruguayan Holstein evaluation for fertility

---

Abstract

Uruguayan dairy cattle genomic evaluations rely heavily on imported genetics. Consequently, incorporating external information from Interbull evaluations (MACE) and foreign genotypes may improve the accuracy of the national dairy genomic evaluation by enlarging the training population. This study aimed to assess the feasibility of integrating MACE proofs and foreign genotypes into the national dairy genomic evaluation for daughter pregnancy rate (DPR). The dataset consisted of DPR records from first to fifth parities used in the April 2023 Uruguayan Holstein genetic evaluation. A total of 896,757 records from 371,606 cows were available, with calving years from 1998 to 2022. Genomic evaluations were conducted using single-step BLUP (ssGBLUP) under a single-trait repeatability model for DPR. The pedigree file contained 531,234 animals, obtained by tracing back three generations of ancestors of cows with DPR records or genotyped animals. External information consisted of MACE proofs and reliabilities for bulls from the April 2018 MACE evaluation (partial MACE) and the April 2023 MACE evaluation (whole MACE). Several scenarios were defined to evaluate the impact of incorporating external information. The reference scenario (G0) included only national phenotypes and genotypes (4,175 animals, of which 2,950 were cows and 1225 bulls). Additional scenarios incorporated MACE proofs as deregressed proofs (pseudo-phenotypes) weighted by their effective record contribution. Scenario G1 included MACE proof for all bulls with national proof. Scenario G2 further included MACE proof for foreign bulls with no national proof but with genotype available (1482 genotyped bulls). In the three scenarios G0, G1 and G2, genomic predictions (GEBVs) were generated using both a whole dataset (all DPR records up to 2023 and MACE proofs from April 2023 run) and a partial dataset (DPR records up to 2018 and MACE proofs from April 2018 run). Validation of GEBVs were based on the linear regression method, using 41 genotyped focal bulls without MACE proofs but with at least 10 daughters contributing DPR records only to the whole dataset. Correlations between full and partial GEBVs were highest in G2 (0.77-0.89). Using MACE proofs for national evaluated bulls and enlarging the reference population by including both genotypes and MACE proofs for foreign bulls (G2), resulted in a 37% increase in GEBV accuracy compared to the base scenario (G0, i.e., national training population). In conclusion, integrating MACE proofs and foreign genotypes of bulls into the Uruguayan ssGBLUP evaluation substantially improves the reliability of genomic predictions for DPR.

Keywords: 2026

How to Cite:

López-Correa, R., Bermann, M., Legarra, A., Lourenco, D., Misztal, I., Carracelas, B., Ravagnolo, O. & Aguilar, I., (2026) “Integration of external data into Uruguayan Holstein evaluation for fertility”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2287247. doi: https://doi.org/10.31274/wcgalp.24246

Rights: 1