Horse phenotyping based on video image analysis of jumping performance for conservation breeding.

Background: Many horse breeds in the world are reserved as genetic resources; however, their characteristics seem to be insufficiently clarified, especially in terms of horse performance. Two jumping ability evaluation methods have been used to compare different types of performance breeds and on this basis their applicability for precision phenotyping has been determined.

Methods: Jumping data of 186 young Polish Warmblood stallions (27 with an endangered status) bred for sport and multipurpose use was collected during their performance tests organised under identical environmental conditions following the same guidelines. Jumping data consisted of objective measurements of free jumping parameters and the marks for jumping. Video recordings of 514 jumps (73 records for 27 stallions with an endangered status) were collected using a digital Panasonic AG-EZ 35 camera (25 fr/sec). Filming was recorded during a free jumping test in the line on a doublebarre obstacle (100-120 cm × 100 cm). Spatial and temporal variables of the jump were measured. The analysis of variance was performed (SAS, General Linear Model and Mixed procedures) using the statistical model, which included the random effect of the horse and fixed effects of the year of test, breeding status, height of jump and the successive number of the jump for objective kinematic data. The fixed effects of the year of test and breeding status were included in the model for subjective performance test data.

Results: Performance marks for free jumping were lower in the endangered group of stallions in the trainers' opinion ( p  ≤ 0.05), while no statistically significant differences were found in the judges' opinions. Statistically significant differences in jumping variables were measured for the bascule points-the elevations of the withers and croup were higher in the endangered group ( p  ≤ 0.001) and the take-off time was prolonged ( p  ≤ 0.05), which explained the subjective evaluation.

Discussion: The use of objective evaluation methods provides important information for practice, as phenotypic differences between horses may be unclear in the subjective evaluation. The objective evaluation should be used to characterise the performance potential of different breeds, because the information from the evaluators might not be consistent. Such characteristics should be recorded at least for every new population.