Sire by finishing environment interactions for beef cattle carcass and meat quality traits

P.L. CHARTERIS, D.J. GARRICK & S.T. MORRIS
Institute of Veterinary, Animal and Biomedical Sciences,
Massey University, Palmerston North, New Zealand

ABSTRACT

Genotype by environment interactions can result in bulls having different genetic merits applicable to different environments. Beef cattle have traditionally been finished on pasture in New Zealand. The desire to meet perceived market requirements for meat quality has encouraged progeny testing and led to feedlotting. The objectives of this study were to evaluate sires on the basis of performance of their pasture-finished steer progeny, their feedlot-finished steer progeny and to compare the assessments of these sires from offspring in each environment.

Carcass and meat quality measures were obtained from 300 purebred steers representing the progeny of 23 Angus sires. Fourteen of these sires had progeny in both environments comprising 54 pasture-finished steers and 148 feedlot-finished steers. Estimated Breeding Values (EBVs) of each sire were obtained separately for eight traits that were measured for both pasture and feedlot progeny. Evaluations were obtained using single-trait Best Linear Unbiased Prediction procedures. Product-moment correlations between sire EBVs based on pasture-finished progeny and EBVs based on feedlot-finished progeny averaged 0.11 (range -0.16 meat colour to 0.50 subcutaneous fat depth) and rank correlations between sire EBVs averaged 0.11 (range -0.13 fat colour to 0.49 subcutaneous fat depth).

The expected distributions of these correlations between EBVs were obtained by simulation for a range of null hypotheses determined by the true genetic correlation between performance in the two environments using Monte Carlo simulation. The expected distributions of correlations between environments were determined for differing numbers of progeny per sire and for the effect of preselecting prior to entering the progeny test. The distributions of correlations were used to test whether the low observed correlations between EBVs estimated from each environment were likely to have arisen from chance sampling or were truly indicative of GxE interaction. The observed correlations were not sufficiently low to reject the null hypothesis at P=0.05 since critical values were -0.18 or -0.36 for true genetic correlations of 1.0 or 0.5. Increasing the number of progeny per sire would increase the chance of detecting an interaction should one truly exist. In that case separate progeny testing programmes for each feeding environment may be necessary.

Presented at the 57th Annual conference of the New Zealand Society of Animal Production, Lincoln University, Feb 11-14, 1997.