Heritability and repeatability of speed for 2- and 3-year-old standardbred racehorses

Repeatabilities (t) and heritabilities (h2) of racing time were estimated from data on 7,206 2- and 3-yr-old Standardbred pacers and trotters competing in 1-mile (1.6 km) charted races at six tracks between 1975 and 1978. A total of 38,487 records representing 2,387 sire progeny groups were divided into subsets by gait, age and track. Initially, the designation "class of race" was recognized as a subjective categorization that reflected nonrandom assignments of horses to races. After extensive investigation, we concluded that racing times should be adjusted by linear regression for the time of the winning horse (defined at the "pace" of the race). Residual regression coefficients of time on pace ranged from .61 +/- .08 to .79 +/- .05 s/s and appeared to adequately account for the environmental effects of class. Repeatabilities of adjusted times ranged from .32 +/- .08 to .55 +/- .04 and heritabilities ranged from -.25 +/- .22 to .66 +/- .16. Overall estimates of t and h2 (pooled over gaits, ages and tracks) were .44 +/- .01 and .29 +/- .03, respectively. When data were not adjusted for pace, the sire, horse and error variance component estimates were inflated by environmental effects (pooled t = .62; pooled h2 = .36). Racing time adjusted for pace is recommended for use as a selection criterion in Standardbred horses.