Study of survival of Holstein cattle of Iran using random regression and single-step BLUP

The survival genetic parameters of Iranian Holstein cattle estimated with the random regression (RR) method in best linear unbiased prediction (BLUP) and single-step genomic BLUP (ssGBLUP) models. The dataset included 30 million test-day records from 1,360,000 first-calving cows in 4,370 herds from 2001 to 2021. Also, the genotypes of 2419 bulls, the frozen semen of which were used for artificial insemination were available. The survival variable for each animal was defined as the number of months from first parturition, with a maximum follow-up period of 100 months thereafter. The cows were assigned to three production groups according to their 305-day milk yield. Results showed the average of survival was 36.2 (±21.9) months. The heritability of the survival trait was 0.049, indicating a relatively small genetic contribution to the variation of the trait. Heritability gradually increased from around 0.02 at the beginning of the period to a peak of 0.06 at approximately 60 months after first calving, and then declined slightly up to 100 months. The RR-ssGBLUP model, with genomic information, significantly improved prediction accuracy (up to 0.74) compared to traditional methods. Higher-producing cows tended to survive longer. These findings highlight the importance of genomic data for enhancing genetic evaluations of survival trait.