PERFORMANCE DEVELOPMENT IN ENGLISH FULL-B LOODED HORSES DURING THEIR RACING LIFE

As a part of more extensive investigations into genetic effects of par ents and grandparents on their offspring performance in the herd of En glish Full-blooded Horse, a partial problem of the development of hors e performance during their racing life was studied. The study included 207 studs and 206 mares taking part in faces held on German race cour ses. At least four-year racing life was a condition for inclusion of a horse into the group under study. In the present paper performance de velopment is analyzed in dependence on horse's age at the outset of ra cing life and on its duration. These relationships were evaluated sepa rately according to the sex. An evaluated rate of performance are the values of general handicap (Gh), that means maximum values Gh values e ach horse has achieved over the racing season. The results of the anal ysis show that in the horses with the outset of their racing life at t wo years of their age stud performance increases in the third and four th year of their age, and then it begins to decrease. The maximum perf ormance of mares was observed at the age of two years. At the age of f our years their performance decreases to a certain extent, in the foll owing years the decrease is steeper. The studs with the outset of thei r racing life at the age of three years have the highest performance a t the age of four and five years; the mares have their maximum perform ance as four years old. A decrease in performance between 4 and 11 yea rs of age makes 4.37 % in studs (i.e. 3.29 kg Gh) and 2.91 % in mares (i.e. 1.86 kg Gh). It has been demonstrated that the horses, and parti cularly studs, with their racing life outset at two years of age have much higher performance than the horses with the first starts at three years of age. This difference in performance corresponds to 7 kg Gh i n studs and 4 kg Gh in mares in the ,,Derby Year''. The horses' perfor mance depends on the duration of their racing life to a great extent. Figs. 1 an 2 illustrate this dependence: they show the averages of stu d and mare performance in relation to age and outset of racing life. I n the course of our study, Gh values of each horse were recorded durin g the whole racing life. An interesting question is what is the depend ence of the values of horse's performance in various racing seasons. T he correlation coefficients which described the dependence of horse pe rformance values in the different seasons were almost identical for ho rses with different duration of racing life. Hence the relations of pe rformance values in the different seasons are in fact independent of t he duration of racing life. But it has been demonstrated that these re lations depend on the sex and age of the horse at the outset of racing life to a great extent. In dependence on these factors, the correlati on coefficients of horses' performance are given for the different sea sons (Tabs. II and III). The correlation coefficients in these tables are mostly statistically significant. The strength of the relations is highest for performance in two following years; then its value is as much as r = 0.9. If the years (seasons) are more distant from each oth er, the values of correlation coefficients decrease. The correlations of horses' performance in the different seasons are so high that the k nowledge of the history of horse's Performance can successfully be use d to predict the performance in the season to come. Tab. III shows a c orrelation between the actual Gh value and the predicted value of hors e's performance: firstly, prediction was based on the values from all preceding years, secondly, prediction was based on horse's performance in the preceding year only. It is apparent that the correlation coeff icients for predictions and actual Gh values are almost identical for both methods of prediction. That is why the knowledge of performance i n all preceding seasons (the best prediction possible) provides only a slight improvement in comparison with the prediction based on perform ance in the immediately preceding season only. We also tried to answer a question whether it is possible to assess the duration of racing li fe from horse's performance at the young age. An analysis of the data obtained has given a negative answer.