Skip to main content
ARS Home » Research » Publications at this Location » Publication #67885

Title: A TEST-DAY MODEL FOR GENETIC EVALUATION OF YIELD TRAITS: POSSIBLE BENEFITSAND AN APPROACH FOR IMPLEMENTATION

Author
item Wiggans, George
item GODDARD, M - UNIV NEW ENGLAND,NSW,AS

Submitted to: Journal of Dairy Science
Publication Type: Abstract Only
Publication Acceptance Date: 3/25/1996
Publication Date: N/A
Citation: N/A

Interpretive Summary:

Technical Abstract: Benefits possible from use of a test-day model for genetic evaluation of yield traits include 1) more accurate estimation of environmental effects from including influence of particular recording days; 2) optimal use of all test-day information; 3) greater stability of bull evaluations as daughters progress through lactation from accounting for genetic differences in shape of lactation curve and rate of maturity; and 4) improved evaluation accuracy for component yields through contributions from milk yield information. A multitrait analysis with 60 traits [3 yield traits (milk, fat, protein), 2 lactation types (first, later) per trait, and 10 lactation stages per type] could provide these benefits. The following strategy is possible. 1) Adjust data for test-day effect within herd and then estimate breeding values for each stage of lactation across herd. Solutions across herds are used to reestimate test-day effects within herd to provide true best linear unbiased prediction solutions. 2) Decrease number of traits to iterate across herds by reducing rank of genetic (co)variance matrix such that information needed for selection criteria is retained. 3) Create uncorrelated traits with a canonical transformation. 4) Replace missing values by their expectations at each iteration to meet canonical transformation requirement for equal design matrices. 5) Apply a repeatability model to accommodate >1 lactation per cow. Each lactation conceptually provides observations for first and later lactation performance, but all values are missing for the other lactation type. 6) Include 305-day records for cows without test- day information by making all values missing for test days. The 305-day record contributes to canonical variables through the covariance matrix.