Genetic Parameters for Racing Performance of Thoroughbred Horses Using Bayesian Linear and Thurstonian Models

The objective of this study was to estimate genetic parameters for rank in Thoroughbred horses using a Bayesian linear model (BLM) and a Thurstonian model (TM) to provide data that contribute to the selection and consequent genetic improvement of the breed in Brazil. Data were provided by the company Turf Total Ltda and consisted of 250,809 records for rank obtained from 40,300 horses and from 34,316 Thoroughbred races (distances of 1,000, 1,300, 1,600, and 2,000 m) that occurred between 1992 and 2011 on six tracks. The rank records at each distance were considered different traits and were submitted to single-trait analysis using BLM and TM. Fixed effects included sex, age, postposition, race, and level of difficulty. The heritability estimates for rank ranged from 0.228 to 0.032 when BLM was used and from 0.293 to 0.047 when TM was used. These estimates tended to decrease with increasing race distance in the two analyses. The TM estimated higher heritability for rank than BLM, indicating the possible use of this model in selection programs of Thoroughbred racehorses in Brazil.