Genetic evaluation for calving ease trait using linear and threshold models in first parity dairy cows

Abstract

Calving ease (CE) is a categorical trait which can be analyzed by both linear and threshold models. Though it has been assumed that threshold models qualify to be the best procedure for analyzing categorical traits, its efficiency with field data has not been clearly confirmed. This study was aimed at comparing linear and threshold models for estimation of CE genetic parameters. Data consisted of 10,575 first parturition CE records collected from 1985 to 2006 in Isfahan, Iran. CE scores were classified into 5 categories depending upon parturition situation. A sire-maternal grand sire model, consisting of birth herd-year-season and sex-age as fixed effects and parturition herd-season, sire and maternal grand sire additive genetic values as random effects, was applied in MATVEC software. Mean, standard deviation, coefficient of variation, skewness, and kurtosis for first parity CE records were 0.233, 0.644, 276.640, 2.950, and 8.097, respectively. CE scores varied from 0 (86%) to 4 (0.1%). CE means for males (0.285) and female calves (0.190) were significantly different. Parturitions during winter and summer were associated with the highest and lowest calving difficulties, respectively. Estimations for direct and maternal heritability were 0.1082, 0.0486 using linear model and 0.0689, 0.0100 using threshold model, respectively. Even though departure from normal distribution was expected to favour the threshold model, implementation of this model in field data analysis did not eventuate to better genetic estimations. Key words: Calving ease, Holstein, Sire-MGS, threshold, genetic parameters.