View article

Health traits are some of the most important cost factors in dairy cattle production. Eight important health traits were chosen for data collection in Canada. They were mastitis, lameness, cystic ovarian disease, left displaced abomasum, ketosis, metritis, milk fever, and retained placenta. Data collected by producers on these 8 diseases were stored in a central database. These recordings were the basis to prepare genetic evaluations for health in Canada. Effect of the quality of the data was analyzed by using 2 different sampling frames for the inclusion of herds in the analysis: a stringent sampling frame requiring all herds to have collected at least one case of the disease analyzed and a second sampling frame requiring herds to have collected one case of any disease. Variance components were estimated with a linear model. Heritability estimates of all health traits were lower than 0.03. The second sampling frame gave lower estimates than the first one. Correlations between predicted transmitted abilities (PTA) calculated with both sampling frames were higher than 0.9. A second analysis compared the effects of using a threshold model instead of a linear model. Health traits were also grouped according to biological aspects. Heritability estimates calculated with the threshold model were higher than those of the linear model, but when they were transformed to the observable scale, results from both modelling approaches were similar. Use of indicator traits was investigated in analyzing body condition score (BCS) and health traits simultaneously. A longitudinal and a multiple-trait approach were used. BCS was positively correlated with resistance …