Abstract
Meat quality (MQ) and carcass composition (CC) are traits with high economic value, but their phenotyping costs makes direct selection for these traits economically unsustainable. Evaluation of traits with lower cost of collection in crossbred (CB) and purebred (PB) animals for use in selection for MQ and CC remains limited. The aim of this study was to measure the response to selection of economically important MQ and CC traits using different selection indices built on cost-effective correlated traits. Phenotypic and genomic information (using 60K SNP chips) was collected for three-way CB (n = 1227 to 5117 phenotyped, n = 1252 genotyped) and PB (n = 3,251 phenotyped, n = 1035 genotyped) Duroc animals belonging to 28 paternal half-sib families. The three breeding objectives (for which response was estimated) were intramuscular fat (IMF), slice shear force (SSF), and loin weight (LW). In total 7 different selection indices (A-G) were used to estimate the selection response for each breeding objective (Table 1). Heritability and genetic correlation parameters for all traits were estimated using GIBBS1F90. Selection response was obtained as expected response per generation weighed by the accuracy of breeding values estimated from a 4-fold cross validation, where animals were grouped based on sire relatedness. For all three breeding objectives, the index which provided best selection response was A (direct selection), with values of 0.35%, -0.72 kg, and 0.85 kg for IMF, SSF, and LW respectively. The index which provided the second-best selection response was B (color and pH traits) for all three breeding objectives, with values of 0.15%, -0.26kg, and 0.30kg for IMF, SSF, and LW respectively. Results for selection response using indices C-G can be found in Table 1. In the absence of direct selection due to high costs of phenotyping, respectable selection response can be achieved by indirect selection using traits with lower cost of collection.
Expectation and variance of the estimator of the maximized selection response of linear selection indices with normal distribution
