|Hutchison, Jana - Edwards|
Submitted to: Journal of Dairy Science
Publication Type: Abstract Only
Publication Acceptance Date: 2/25/2008
Publication Date: 7/11/2008
Citation: Kuhn, M.T., Hutchison, J.L., Norman, H.D. 2008. Modeling nuisance variables for phenotypic evaluation of bull fertility. Journal of Dairy Science. 91(E-Suppl. 1):7–8 (abstr. T20).
Technical Abstract: This research determined which (available) nuisance variables should be included in a model for phenotypic evaluation of US service sire conception rate (CR), based on DHI data. Models were compared by splitting data into records for estimation (n=3,613,907) and set-aside data (n=2,025,884), computing predictions using the estimation data, and then comparing predictions to bulls' average CR in set-aside data. There were 803 bulls used for comparison, after requiring a minimum of 50 records for estimation and 100 breedings in at least 30 herds in set-aside data. Correlations and mean differences were used to compare models. Nuisance variables considered were management groups based on herd-yr-season-parity-registry (HYSPR) classes, yr-state-mo, cow age, DIM, a short-cycle variable to account for lower CR for matings preceded by a breeding only 10 to 17 days prior, and various combinations of lactation, service number, and milk yield. Preliminary analyses led to selection of 305d-2x-ME milk yield as the production variable for consideration, and also showed that for each quantitative independent variable, categorical factors provided better bull fertility evaluations than modeling the effects as covariates. Two strategies for management groups were tested, one where HYSPR groups were required to have an absolute minimum number of records and a second where groups were combined across registry, season, and parity subclasses until a minimum group size was achieved. Combining groups to a target size of 20 but still including herd-years with at least 10 breedings maximized correlation with CR in set-aside data. Combining groups implies that some groups have multiple seasons and parities, hence consideration of yr-state-mo and lactation as additional factors. The final variables selected for inclusion were, in addition to HYSPR, yr-state-mo, lactation, service number, milk yield, cow age, short-cycle (yes/no), and the cow effect, partitioned as permanent environment and breeding value. This model maximized correlation with CR in set-aside (55.2%), minimized mean square error (3.25), and mean difference was 0.