Random regression model estimation of genetic parameters for show‐jumping results of Hungarian Sporthorses

Abstract

The aim of this study was to estimate the genetic parameters for show-jumping competition performance of Hungarian Sporthorses using a random regression model. There were 21,210 records from 739 horses collected in Hungary between 1996 and 2004. Performance was expressed as shifted Blom normalized ranks and as the difference between fence height and fault points. The random regression model (RRM) included fixed effects for sex, year, location, and obstacle height and random effects for animal, rider and permanent environment. Regressions for the random effects in the RRM were modelled with Legendre polynomials from first to fifth order of fit. The model focused on performance of horses from 4 to 11 years of age, with heterogeneous residual variances considered. The heritabilities were low to moderate for both variables. Genetic and phenotypic correlations between different ages decreased with increasing distance between the ages.

Comparison of models to estimate genetic parameters for scores of competitive sport horse events in the Czech Republic

“…5 years of competition). Slightly higher heritabilities (0.03-0.30) were reported by Posta et al (2010), who used a similar RR model but with regression based on the age of the horse.…”

“…The increasing impact of the permanent environmental effect and the increasing level of difficulty of the competitions can be explained by the fact that all the horses competing at a high level of difficulty were well experienced and had extensive preparation and quality care, leading to reduced variation among the horses. Posta et al (2010) also reported that the performance of show jumping horses later in their career was less influenced by advantages of the MT model, it is worthwhile using it for genetic evaluation of sport horses in equestrian competitions. Table 4 shows the relative number of horses competing in the given trait according to the number of years of horse's experience.…”

“…This may be due to the small number of observations from 4100 days (11 years) of competitive experience. Posta et al (2010) used random regression on age of the horse and published curves of predicted breeding values at five advancing ages for the most popular stallions in Hungary at that time. They concluded that estimated breeding values for the stallions were relatively stable across time.…”

“…The random regression model has been used to evaluate performance of German Trotters (Bugislaus et al 2006), Thoroughbred horses in Brazil (Buxadera and Da Mota 2008), and Hungarian Sporthorses (Posta et al 2010). The multi-trait animal model was used by Janssens et al (1997) in Belgium, Alridge et al (2000) in Ireland, Viklund et al (2011) in Sweden, and Luehrs-Behnke et al (2002) in Germany.…”

“…Show-jumping competition results can be modelled as a repeated measurement of a trait for which the phenotype of an animal is represented as a continuous function of time (Posta et al 2010), which can be characterized by a trajectory with a theoretically infinite number of measurements. The random regression model has been used to evaluate performance of German Trotters (Bugislaus et al 2006), Thoroughbred horses in Brazil (Buxadera and Da Mota 2008), and Hungarian Sporthorses (Posta et al 2010).…”