ENGLISH VERSION

ZESZYT 357

Iwona Rozempolska-Rucińska
Threshold reproductive traits in improvement of laying hens

Application of appropriate mathematical methods in predicting the breeding value of reproductive traits enhances the reliability of estimates of (co)variance components, which, in turn, affects the quality of interpretation of the genetic mechanisms of determination of quantitative traits, the quality of selection response prediction, adequate selection of breeding programs and accuracy of breeding value estimation. 

OBJECTIVE. The aim of the study was to answer the question: how, i.e. with which biometric tools and traits included in the selection criterion, can efficient breeding programs be implemented in terms of hatchability of laying hens?
MATERIAL AND METHODS. The study was conducted on data from seven generations of rhode island white laying hens. Biometric modelling was applied to the number of chicks hatched from eggs during the setting period, both from eggs set to hatch, and eggs discarded due to structure defects or shell cracks. The performance traits were not the object of the analysis, but they were incorporated into mathematical models as assisting traits, thus raising the reliability of the estimates. While developing the models to assess the (co) variance, the significance of permanent environmental effects was taken into account, which had previously been verified by variance analysis performed with fixed effect models (the GLM procedure). Genetic parameters and prediction of the breeding value for the number of hatched chicks was assessed using linear and threshold-linear models with varied arrangement of traits and changing the analysis object and the assisting additive effect of the individual in the AM model. The genetic parameters were estimated and BLUP assessment of the number of chicks hatched was performed in accordance with the mathematical model. The reliability and consistency of ranking of individuals based on their breeding value obtained with various models were compared.

RESULTS AND DISCUSSION. The highest accuracy of genetic parameter estimates and predicted breeding value of hatchability was obtained using the threshold-linear model. The model is stable when there is a full set of information, as is the case while predicting the reference breeding value. In such a case, the change in animal ranking, depending on the accompanying trait used, was very insignificant, not exceeding 7% of the assessed individuals. The threshold-linear model proved to be more “sensitive” to lack of information than to the changed set of information. Which of the assisting linear traits would be incorporated into the model was less important than the lack of performance of the selected generation. Estimates obtained with the threshold-linear model were very close to those obtained with the linear model taking into account the random effect of the specific individual environment (LL_PERM). Although this model provides lower accuracy of BLUP values, the bird ranking in terms of the breeding value ​​was analogous with differences not exceeding 4%. It was shown that hatchability is a trait that can be defined in two categories, as a maternal trait and an embryo trait. In the first case, measurement of the trait is made by determining the hatchability level recorded for the hen; in the other case, the event of hatching or no-hatching of an individual from a particular egg is recorded. The quality of the eggs produced as hatching material determines the additive effect of an individual in approximately 18%; simultaneously, this is a high repeatability trait. As the embryo trait, hatchability is additively conditioned only to a slight degree. A high value of maternal heritability was found, which indicated that the probability of hatching is largely determined by the effect of the laying hen. High maternal repeatability confirms that it is the environment in which an embryo develops that has a huge impact on hatching success.