versión impresa ISSN 0365-2807
URIBE, Héctor A.. Test day model: A new statistical tool for genetic evaluation of dairy cattle. Agric. Téc. [online]. 2001, vol.61, n.4, pp. 500-511. ISSN 0365-2807. http://dx.doi.org/10.4067/S0365-28072001000400012.
Molecular genetics is making enormous improvement in identifying major effect genes or quantitative trait loci. However, quantitative genetics still plays an important role on modern animal breeding. Animal models using a single record per lactation assume that environmental effects do not vary through the entire lactation. A test day model permits environmental changes through the lactation. Test day models were developed in Canada and are being used for genetic evaluation in at least two countries. Chile, as a dairy frozen semen importer, uses genetic evaluations done abroad. Although domestic application of such evaluations is limited, it is necessary a minimal understanding of the methodology used in the estimation of breeding values. The objective of this work is to discuss basic aspects of the statistical theory used in test day models. A random regression test day model example is presented. A test day random regression model has been implemented in Canada for dairy cattle genetic evaluation, this allows better modeling of the lactation curve. It is concluded that breeding value estimation, using test day models, is better as compared to single record animal models, however, computer requirements are increased. Persistence, as a sub product of random regression test day models, may be an important production trait.
Palabras clave : breeding values; lineal models; random regression.