Can early training of show jumpers bias outcome of selection events? Susana Santamarı ´a a,1 , Maarten F. Bobbert b , Wim Back a , Albert Barneveld a , P. Rene ´ van Weeren a , * a Department of Equine Sciences, Faculty of Veterinary Medicine, Utrecht University, Yalelaan 12, NL-3584 CM Utrecht, The Netherlands b Institute for Fundamental and Clinical Human Movement Sciences, Vrije Universiteit, van der Boechorstraat 9, NL-1081 BT Amsterdam, The Netherlands Received 8 April 2005; received in revised form 30 December 2005; accepted 11 January 2006 Abstract Evaluation of free jumping at sub-maximal heights is common practice within selection procedures for young breeding stallions. Early training might cause an unjustified bias. To investigate this, data from a 5-year longitudinal study on 30 horses were used. Half of these horses (experimental group) had received early training between 6 months and 4 years, the other half (control group) had not. Between 4 and 5 years, all horses had received standard training under saddle. At the age of 5 years, the horses were tested in a puissance competition and the 7 best and 6 worst jumpers were used for the present study. Kinematic variables that were different at ages 5years and 6months, and had been shown to be predictive for performance in earlier studies, were analysed at the age of 4years. It showed that early training had effaced the differences between potentially good and less good show jumpers in 3 of 3 predictive variables and had introduced a (false, because not related to performance) difference between the trained and untrained horses in one of them and a nearly significant trend in another. Early training may to a certain extent obscure differences in talent among individuals at the age at which selection events occur. Experienced judges may be able to account for this, but studbooks and judges should be aware of this possible pitfall. D 2006 Elsevier B.V. All rights reserved. Keywords: Horse; Selection for breeding; Early training; Kinematics; Show jumping 1. Introduction The goal of breeders of Warmblood horses is to produce excellent performance horses that can perform in top-level competitions. It is commonly accepted that performance traits are heritable (Barneveld, 1996; Barrey and Langlois, 2000; Koenen et al., 1995), so 1871-1413/$ - see front matter D 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.livsci.2006.01.003 * Corresponding author. Tel.: +31 30 2531062; fax: +31 30 2537970. E-mail address: R.vanWeeren@vet.uu.nl (P.R. van Weeren). 1 Present address: Department of Veterinary Clinical Studies, Faculty of Veterinary Medicine, University of Edinburgh, Veteri- nary Field Station, Roslin, EH25 9RG Edinburgh, UK. Livestock Science 102 (2006) 163 – 170 www.elsevier.com/locate/livsci