A genetic evaluation of production traits in Nguni cattle

English: Genetic parameters may well aid the prediction of selection and correlated response when selection is based on production traits. The animal model, fitting an additional random factor such as the permanent maternal environmental effect is common practice to increase the accuracy of estimates. As selection in a beef breed such as the Nguni takes place in various environmental conditions, the inclusion of an additional random factor i.e. herd-year-season x sire interaction seems justified. Production traits analyzed included records of birth weight (BW), weaning weight (WW), yearling weight (YW) and eighteen-month weight (EW). Herd-year-season, sex, age at recording (except BW) and age of dam had a significant effect. Herd-year-season x sire interaction (HYSxS) was included as an additional random factor. Estimates obtained by univariate analyses were used as starting values in the multivariate analyses. In multivariate analyses another data set was used which included only herds linked with at least two other herds through sires used. This data set was also used in two bivariate analyses being the analyses between the maternally influenced traits (BWand WW) and WW and YW to gain knowledge as to a possible carry-over effect of the maternal genetic effect. Genetic trends were calculated using the regression of the average breeding value estimates . (EBV) on year of birth. Direct heritability estimates for BW, WW, YW and EW obtained by multivariate analyses were 0.31, 0.32, 0.23 and 0.27 respectively. Genetic correlations between direct and maternal effects for BW and WW were mostly low to moderate and negative (-0.35 to - 0.59 and -0.17 to -0.50 for BW and WW respectively), while direct genetic correlations among the four growth traits were moderate to high (0.51 to 0.95). The maternal genetic correlation between BW and WW was moderate (0.49). It was concluded that selection on WW alone would be undesirable as all other weights would increase. In review, breeding values for production traits should be seen in conjunction with one another so as to take into account undesirable effects on other traits. Genetic trends derived showed no major change. Nguni cattle, like most other beef producing breeds, are kept under a wide variety of extensive environments throughout South Africa. Based on this, a genotype x environment interaction (G x E) in this breed may be expected. Three most important aspects while considering the nature of interactions are the ranking order of the genotypes, the relative magnitudes of genotypic, environmental and interaction effects and the statistical significance of interaction effects. Because of limitations in the data set, only weaning weight (WW) records were used. Records from herds were allocated to one of three veld types that described the specific region in which each berd was situated, i.e. sweetveld (Region 1), sourveld (Region 2) and mixed veld (Region 3). Bivariate analyses of WW in the three different regions were carried out to determine the (co)variance components and weighted direct heritability estimates obtained ranged from 0.23 to 0.45. Direct genetic correlations of Regions 2 and 3 with Region I were significant and were 1.00 and 0.71 respect ivcly. This indicates that no Cl x E occurred between Regions I and 2, but indeed between Regions I and 3. The genetic correlation between Regions 2 and 3 (0.60) were non-sign ificant. Correlation coefficients between breeding values of sires, obtained by multivariate analyses and the different bivariate analyses, was also calculated (range 0.77 to 0.8]) indicating no major re-ranking of sires when selection is based on breeding values obtained by multivariate analyses which include a HYSxS interaction. Top performing sires showed more re-ranking.