Beyond milk yield: could resilience be revealed as deviations through the first lactation curve in Holstein cows?

Selection for increased milk yield in cattle, often prioritized in dairy breeding programs, has negatively affected other economically important traits, such as health and fertility, thereby reducing animal resilience to environmental disturbances. This study aimed to estimate the heritability of resilience using the natural logarithm of variance (LnVar), autocorrelation (Auto), and skewness (Skew), as well as the genetic correlations between resilience and productive, health, and fertility traits in first-lactation cows. Data from 45,292 first-lactation Holstein cows from 237 herds in Paraná, Brazil, were used. Resilience indicators (LnVar, Auto, and Skew) were calculated from milk yield deviations relative to expected lactation curves. Genetic analyses were performed using the REML method to estimate heritability for resilience (through each indicator) and genetic correlations between resilience and productive traits (305-day milk yield - MY305, 305-day protein yield - PY305, and 305-day fat yield - FY305), reproductive traits (age at first calving - AFC and age at first service - AFS), and health traits (somatic cell count - SCC). Heritability estimates for resilience were low across all indicators, ranging from 0.01 to 0.19, with the highest values observed for the LnVar indicator (h² = 0.14 to 0.19). Furthermore, LnVar showed moderate to high genetic correlations with production traits, ranging from 0.42 to 0.59 for MY305, 0.30 to 0.47 for PY305, and 0.15 to 0.37 for FY305. In contrast, genetic correlations with health and fertility traits were low, ranging from 0.02 to 0.18. Therefore, among the indicators evaluated, LnVar was identified as the most promising resilience metric in this population because of its higher heritability and consistent genetic associations with production traits.