Skip to main content
ARS Home » Research » Publications at this Location » Publication #190383


item Norman, H

Submitted to: Journal of Dairy Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/9/2006
Publication Date: 1/1/2007
Citation: Dechow, C.D., Norman, H.D. 2007. Within-herd heritability estimated with daughter-parent regression for yield and somatic cell score. Journal of Dairy Science. 90(1):482-492.

Interpretive Summary: Heritability is often assumed to be constant across herds, but management differences may impact the value of a herd’s records when generating genetic evaluations. Within-herd heritabilities that were generated rapidly with regression varied significantly across herds and could be incorporated into routine genetic evaluations to improve accuracy.

Technical Abstract: The objectives of this study were to compare estimates of within-herd heritability (WHH) generated with daughter-dam regression, daughter-sire predicted transmitting ability (PTA) regression and REML and to evaluate the effect of adjusting records according to WHH. Milk, fat and protein yield and SCS was provided by the Animal Improvement Programs Laboratory at USDA for Holsteins representing the Northeast, Southeast, Midwest and Western regions of the US. Four datasets with 457 to 499 herds were randomly selected and the 15 largest herds from the Western region and ten largest herds from the other regions were selected across samples to create a large herd sample. Sample heritabilities for yield and SCS were estimated with a regression model that included fixed covariates for dam yield or SCS, sire PTA for yield or SCS, herd-year-season of calving and age within parity effects. Dam records and sire PTA were nested within herd and included as random covariates to generate WHH estimates that were regressed toward the sample average. Heritabilities were also estimated using REML for 45 large herds. Phenotypic variance for each herd was estimated from herd residual variance after adjusting for year-season and age within parity effects. Deviations from herd-year-season average were standardized to constant genetic variance across herds and records were weighted according to estimated error variance to accommodate WHH when estimating breeding values. Average WHH tended to be higher when estimated with daughter-dam regression (0.35 for milk yield) than when estimated with daughter-sire PTA regression (0.24 for milk yield). Heritability estimates varied widely across herds (0.04 to 0.67 for milk yield estimated with daughter-dam regression) and WHH deviated from sample averages more for large herds than small herds. The correlation between WHH estimated with daughter-dam regression and REML was 0.67 for milk yield, whereas the correlation between the REML WHH and daughter-sire PTA WHH was 0.46. Correlations between estimated breeding values and official genetic evaluations for sires with 99% reliability generally increased after adjusting for WHH, while correlations between yield deviations and parent averages generally declined. Within-herd heritability can be estimated rapidly using regression techniques with moderate accuracy, but adjusting to records for WHH with the methods used here resulted in minimal improvements to genetic evaluation accuracy.