Dairy cattle

Herd-level data quality control improves genetic evaluations for gestation length in dairy cattle

---

Abstract

The objective of the present study was to explore herd-level quality control approaches for gestation length data and assess their effects on genetic parameter estimation in dairy cattle. A total of 4,861,141 gestation-length records from 1,827,654 dairy cows spanning the years 2005 to 2024 were available. Herd-year distributions for gestation length were classified using Hartigan's dip test and kernel density estimation as either unimodal, bimodal or trimodal. A total of 77.2% of the herd-years were unimodal, 22.0% were bimodal, and 0.8% were trimodal; data from trimodal herd-years were not considered further. The mean number of days within herd-year between the peaks of the two gestation modes was 20 days pointing to service dates not having been recorded. Variance components were estimated for 261 randomly selected unimodal herds (76,586 records from 43,424 cows) and 523 bimodal herds (73,849 records from 57,977 cows) spanning from 2018 to 2022 using linear mixed models. Six ranges in gestation length were considered: 270 - 290 days, 270 - 300 days, 260 - 300 days, 260 - 305 days, 260 - 310 days, and 250 - 310 days. For unimodal herd-years, direct additive genetic standard deviation increased as the range of phenotypic gestation lengths widened, from 3.01 days for the 270 - 290 days range to 5.39 days for the 250 - 310 days range; the phenotypic standard deviation followed a similar trend (4.28 to 7.10 days). The direct heritability ranged from 0.49 - 0.63 while the maternal heritability ranged from 0.01 to 0.03. For bimodal herd-years, the additive genetic and phenotypic standard deviations were 1.4% to 17.0% and 3.3% to 24.7% higher, respectively, than those of unimodal herd-years. The direct heritability estimates were also higher than the corresponding values for unimodal herd-years, peaking at 0.67 for the 260 - 310 days range. Including a fixed effect reflecting which peak in the bimodal herd-year the gestation length record belonged to in an interaction with contemporary group resulted in variance component estimates and heritability estimates resembling those estimated from just the unimodal herds. In conclusion, applying appropriate adjustments to bimodal data can recover 22% more records. Therefore, integrating herd-level distributional quality control into national genetic evaluations can improve both precision and robustness of genetic evaluation for gestation length in dairy cattle.

Keywords: 2026

How to Cite:

Mwangi, S., Buckley, F., Evans, R. & Berry, D., (2026) “Herd-level data quality control improves genetic evaluations for gestation length in dairy cattle”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2285418. doi: https://doi.org/10.31274/wcgalp.23694

Rights: 1