# Genetic Parameters for Rumination Time, Daily Average Milk Temperature, and Milking Traits Derived from Automatic Milking Systems in Holstein Cattle

## TL;DR

This study uses data from automatic milking systems to explore the genetic basis of traits like rumination time and milk temperature in Holstein cattle.

## Contribution

The study provides genetic parameter estimates for AMS-derived traits, showing their potential for use in breeding programs.

## Key findings

- Heritability estimates for traits like milking time and rumination time were moderate to high.
- Milk temperature showed seasonal variation and measurable genetic variation.
- Genetic correlations among traits were observed but require caution due to large standard errors.

## Abstract

Automatic milking systems (AMSs) continuously record large amounts of information on milk production, cow behavior, and physiological indicators. These data offer valuable opportunities to study the genetic background of traits related to milking efficiency, health, and welfare. In this study, data from a Holstein herd managed under AMS were used to evaluate the genetic basis of production, behavioral, and physiological traits, including rumination time and milk temperature. The analyses revealed that these AMS-derived traits are influenced by genetic factors, showing potential for inclusion in breeding programs. Using routinely collected AMS data can help identify cows that are not only productive but also well adapted to automated systems. This study shows that AMS information can support the development of breeding strategies aimed at improving both productivity and animal welfare in modern dairy herds. These results are based on routinely collected AMS data from a single commercial herd, and further multi-herd and large-scale studies are required to confirm and extend these findings.

Automatic Milking Systems (AMSs) enable the continuous recording of production, milkability, behavioral, and physiological traits, offering new opportunities for genetic evaluation in dairy cattle. This study aimed to estimate variance components and genetic parameters for milk yield-related traits, milking efficiency traits, rumination time (RT), and daily average milk temperature (MTEMP) using AMS-derived data from 1252 Holstein cows. 65,475 weekly records from a single commercial herd were analyzed using repeatability animal models fitted by restricted maximum likelihood. Heritability estimates were moderate to high for milking time (MT) (0.31), milking speed (MS) (0.38), RT (0.30), and MTEMP (0.28), whereas behavioral traits such as number of milking (NoM) (0.26) and number of refused (NoREF) (0.11) showed lower but meaningful heritabilities. Repeatability was highest for MT and MS (0.77 and 0.79), indicating consistent milking performance across repeated records. MTEMP demonstrated clear seasonal variation, increasing in warmer periods and decreasing during colder months, indicating sensitivity to environmental conditions. Genetic correlations among traits revealed both favorable and unfavorable associations; however, several estimates were associated with relatively large standard errors and should therefore be interpreted with caution. The inclusion of MTEMP as a proxy physiological trait derived from AMS data showed measurable genetic variation, although its biological interpretation requires careful consideration. Overall, the results suggest that AMS-derived phenotypes may contribute useful information for genetic studies of functional traits, but the single-herd structure, limited pedigree depth, and data aggregation procedures restrict the generalizability of the findings. Further multi-herd and genomics-based studies are required to validate these results and assess their applicability in breeding programs.