Genetic Evaluation of Resilience Indicators in Holstein Cows

Abstract

The analysis of resilience indicators was based on daily milk yields recorded from 3347 lactations of 3080 Holstein cows located on 10 farms between 2022 and 2024. Six farms used an automatic milking system. A random regression function with a fourth-degree Legendre polynomial was used to predict the lactation curve. The indicators were the natural log-transformed variance (LnVar), lag-1 autocorrelation (r-auto), and skewness (skew) of daily milk yield (DMY) deviations from the predicted lactation curve, as well as the log-transformed variance of DMY (Var). The single-step genomic prediction method (ssGBLUP) was used for genomic evaluation. A total of 9845 genotyped animals and 36,839 SNPs were included. Heritability estimates were low (0.02-0.13). The strongest genetic correlation (0.87) was found between LnVar and Var. The genetic correlation between r-auto and skew was also strong but negative (-0.73). Resilience indicators showed a negative correlation with milk yield per lactation and a positive correlation with fat and protein contents. The negative correlation between fertility and two resilience indicators may be due to the evaluation period (50th-150th day of lactation) being when cows are most often bred after calving, and a decrease in production may accompany a significant oestrus. The associations between resilience indicators and health traits (clinical mastitis, claw health) were weak but mostly favourable.

Keywords: dairy cattle; genomic selection; resilience; single-step genomic prediction.

Figure 1

Average daily milk yield and number of animals in the first and later parities.

Figure 2

Average accuracies of GEBVs for resilience indicators in particular groups of animals. Resilience indicators are log-transformed variance of daily milk yields (Var), log-transformed variance of deviations (LnVar), lag-1 autocorrelation of deviations (r-auto), and skewness of deviations (skew).

Figure 3

Pearson correlation coefficients between RBVs of sires considering resilience indicators and RBVs considering milk production traits and the Czech Holstein Selection Index (SIH). Resilience indicators are log-transformed variance of daily milk yields (Var), log-transformed variance of deviations (LnVar), lag-1 autocorrelation of deviations (r-auto), and skewness of deviations (skew). Milk production traits include milk yield (kg), protein yield (kg), fat yield (kg), and protein and fat contents in milk (%). Higher RBVs are desirable; therefore, positive correlations are favourable.

Figure 4

Pearson correlation coefficients between RBVs of sires considering resilience indicators and RBVs considering selected exterior traits. Resilience indicators were log-transformed variance of daily milk yields (Var), log-transformed variance of deviations (LnVar), lag-1 autocorrelation of deviations (r-auto), and skewness of deviations (skew). Exterior traits include four feet- and leg-type traits and six udder-type traits.

Figure 5

Pearson correlation coefficients between RBVs of sires considering resilience indicators and RBVs considering fertility traits. Resilience indicators are log-transformed variance of daily milk yields (Var), log-transformed variance of deviations (LnVar), lag-1 autocorrelation of deviations (r-auto), and skewness of deviations (skew). Fertility traits include daughter fertility recorded in heifers, cows, and in both groups (daughters) together, direct calving ease, maternal calving ease, and direct 1st calving ease. Higher RBVs are desirable; therefore, positive correlations are favourable.

Figure 6

Pearson correlation coefficients between RBVs of sires considering resilience indicators, RBVs considering health traits, RBVs considering body condition score, and longevity and health indexes. Resilience indicators are log-transformed variance of daily milk yields (Var), log-transformed variance of deviations (LnVar), lag-1 autocorrelation of deviations (r-auto), and skewness of deviations (skew). Longevity index combines functional longevity with fertility of cows, body depth, udder depth, foot and leg score, and somatic cell count. Health index consists of resistance to clinical mastitis, infectious claw diseases, claw horn lesions, and overall claw disorders. Higher RBVs are desirable; therefore, positive correlations are favourable.