approach (i.e., with in individual traits) was also tested and produced very similar results. Highest posterior densities
for variances and covariances of these results are shown in
Table5 as well. Results were successfully obtained by the
Gibbs sampling method for all tested single-trait and
multi-trait models. In the single-trait model, the most in fluential effect on
jumping performance of horses was the permanent environment, i.e., rearing,training,nutrition and treatment.The
effect of the rider had a lower in fluence, and the lowest
influence of all was the additive genetic value of the horse.
As shown in Table 4, the residual effect accounted for
73.67% of total phenotypic variance, heritability was 0.07
and repeatability was 0.18.
In the multi-trait model, influence of the rider and the
additive genetic effect increased with increasing difficulty
of the event, and the variance of the residual effect
decreased.The results for the first trait(i.e.,low difficulty;
fence height 90–110 cm) were similar to those of the
single-trait model, including estimates of heritability and
repeatability. For the second trait(II;height 120–135 cm),
heritability increased to 0.10,and repeatability increased
to 0.22.For the third trait(III;height 135–150 cm),the
most influential random effect was rider followed by the
additive genetic effect of the horse, and the least influential effect was the permanent environmental effect.The
heritability (0.16) and repeatability (0.25) were highest for
third trait.
Correlations of the additive genetic effects among traits
in the multi-trait model were high,ranging from 0.83 to
0.96 Correlations among the traits for the rider effect and