Principal component analysis of breeding values for growth and reproductive traits and genetic association with adult size in beef cattle1

Abstract

Principal component analysis was applied to evaluate the variability and relationships among univariate breeding values predicted for 9 weaning and yearling traits, as well as suggest functions of the traits that would promote a particular breeding objective. Phenotypic and pedigree information from 600,132 Nelore animals was used. Genetic parameters and breeding values were obtained from univariate analyses of birth to weaning weight gain; weaning to yearling weight gain; conformation, finishing precocity, and muscling scores at weaning and at yearling; and yearling scrotal circumference. The principal component mainly associated with maturity (precocious vs. late animals) was used as a pseudophenotype in bivariate analyses with either adult weight or adult height of cows. Direct heritability estimates ranging from 0.19 ± 0.01 to 0.41 ± 0.01 indicate that these 9 traits are all heritable to varying degrees. Correlations between the breeding values for the various traits ranged from 0.14 to 0.88. Principal component analysis was performed on the standardized breeding values. The first 3 principal components attained the Kaiser criterion, retaining 48.06%, 18.03%, and 12.97% of the total breeding value variance, respectively. The first component was characterized by positive coefficients for all traits. The second component contrasted weaning traits with yearling traits. The third component represented a contrast between late maturity animals (better for weight gain and conformation) and early maturity animals (better for finishing precocity, muscling, and scrotal circumference). Thus, the first 3 components represent 3 different potential selection criteria. Selecting for the first principal component would identify animals with positive breeding values for all studied traits. The second principal component may be used to identify animals with higher or lower maturation rates (precocity). Animals with negative values in the third principal component are regarded as early maturing. Genetic correlations between maturity (third principal component) with adult weight and adult height were 0.19 ± 0.02 and 0.32 ± 0.02, respectively. These results showed that the weaning and yearling bovine maturity is genetically associated with the adult size of the cows used in reproduction.

Colour variability of beef in young bulls from fifteen European breeds

“…This is in part due to differentially expressed genes associated either with muscle mass or fat deposition in the carcass (Bernard et al, 2009). Specialised beef breeds are generally late-maturing and fatten later than unimproved and dairy breeds (Albert ı et al, 2005;Boligon et al, 2016). However, the variation in meat colour observed the present study is not related to fatness as assessed by intramuscular fat deposition (Christensen et al, 2011) Red, Jersey and Holstein as unimproved and dairy breeds.…”

“…Specialised beef breeds are generally late‐maturing and fatten later than unimproved and dairy breeds (Albertí et al ., ; Boligon et al ., ). However, the variation in meat colour observed the present study is not related to fatness as assessed by intramuscular fat deposition (Christensen et al ., ) or the dissected rib fat content (Albertí et al ., ).…”

Principal components for morphometric traits in Campolina horses

“…(2013) and Boligon et al. (2016) suggested that principal components can be used as a selection index. The indices commonly used in horses for morphometric traits have empirical weight and do not consider the genetics correlations between variables (breeding values) (Cabral, 2004; Lucena et al., 2016).…”

“…According to Boligon et al. (2016), this approach addresses the correlations between variables and may be considered a new composite trait. According to Vargas et al.…”

“…According to Roso and Fries (1995), some traits might be redundant because of their high genetic correlation. The presence of several phenotypes requires extensive analytical work and often does not increase the accuracy of the estimates (Boligon et al., 2016).…”