Genomic prediction of consumer satisfaction traits of Australian beef

Abstract

<p>Consumer satisfaction has become a key focus for beef producers as eating quality traits such as tenderness and flavour dictate purchasing choices and, ultimately, the price consumers are willing to pay. Due to the difficulty in measuring eating quality traits and the inability to predict those traits prior to slaughter, beef producers opt to select for correlated traits and indirectly select for eating quality. Genotyping of animals offers the opportunity for the selection of cattle with superior eating quality directly for both breeding and market allocation. The aim of this study was to determine the accuracy of genomic prediction along with heritabilities for eating quality traits" tenderness, juiciness, flavour and overall liking as well as the overarching consumer satisfaction trait known as MQ4 in a 10-fold cross validation. Phenotypes from 1,701 cattle recorded in eating quality trials held across Australia were collected for the 5 eating quality traits. Those same cattle were genotyped using varying Illumina SNP arrays between 50k and 100k density and then imputed up to high density 700k using a reference set of 4,506 cattle representing most breeds and crossbreds composites of the Australian beef herds. A linear mixed model was used with cohort, days aged, carcase weight, principal components 1-4 and heterozygosity fit in the model. Heritabilities ranged from 0.21 to 0.32 between juiciness and tenderness respectively, while tenderness and MQ4 had the highest accuracy of 0.27 from the cross validation and juiciness and flavour having the lowest accuracies of 0.23. While accuracies observed in this study were low, moderate heritabilities indicate selection for eating quality traits is feasible.</p>