RESUMEN

REASONS FOR PERFORMING STUDY: The pathogenesis of osteochondrosis (OC) and palmar/plantar first phalanx osteochondral fragments (POFs) is multifactorial, but specific knowledge of heritability is limited. OBJECTIVES: To improve the precision of heritability estimates and to estimate the genetic correlation between tarsocrural OC and POFs in Standardbred trotters. Further aims were to examine whether the prevalence of OC/POFs was different in the American and French lineages that have contributed to the Norwegian population, and if the prevalence was affected by heterozygosity. STUDY DESIGN: Retrospective cohort study. METHODS: Categorical data on tarsocrural OC and POFs from 2 radiographic studies performed in 1989 and 2007/2008 (n = 1217) were analysed with sire threshold models that included 230 sires. RESULTS: Heritability of OC at the distal intermediate ridge of the tibia and/or the lateral trochlear ridge of the talus was estimated at 0.29 ± 0.15. For OC at the distal intermediate ridge of the tibia only, the estimate was 0.40 ± 0.17. Heritability of POFs in all 4 limbs was estimated at 0.23 ± 0.13; for metatarsophalangeal POFs this was 0.26 ± 0.13 and for medial metatarsophalangeal POFs 0.32 ± 0.14. Estimates of genetic correlation between OC and POFs ranged from 0.68 ± 0.27 to 0.73 ± 0.28 but were not significantly different from a zero-genetic correlation. Effects of lineages or heterozygosity were not observed. CONCLUSIONS AND POTENTIAL RELEVANCE: This study confirmed a moderate to high heritability of tarsocrural OC and POF, providing further evidence of the heritable nature of these diseases. Examination of specific lesions yielded the highest heritability; therefore, breeding programmes and future genome-analysis studies should focus on predilection sites rather than the entire disease complex.

Asunto(s)

RESUMEN

The aim of this study was to examine how to apply optimal contribution selection (OCS) in the Norwegian and the North-Swedish cold-blooded trotter and give practical recommendations for the future. OCS was implemented using the software Gencont with overlapping generations and selected a few, but young sires, as these turn over the generations faster and thus is less related to the mare candidates. In addition, a number of Swedish sires were selected as they were less related to the selection candidates. We concluded that implementing OCS is feasible to select sires (there is no selection on mares), and we recommend the number of available sire candidates to be continuously updated because of amongst others deaths and geldings. In addition, only considering sire candidates with phenotype above average within a year class would allow selection candidates from many year classes to be included and circumvent current limitation on number of selection candidates in Gencont (approx. 3000). The results showed that mare candidates can well be those being mated the previous year. OCS will, dynamically, recruit young stallions and manage the culling or renewal of annual breeding permits for stallions that had been previously approved. For the annual mating proportion per sire, a constraint in accordance with the maximum that a sire can mate naturally is recommended.

Asunto(s)

RESUMEN

There have been several approaches to the estimation of breeding values of performance in trotters, and the objective of this study was to validate different alternatives for genetic evaluation of racing performance in the North Swedish and Norwegian cold-blooded trotters. The current bivariate approach with the traits racing status (RACE) and earnings (EARN) was compared with a threshold-linear animal model and the univariate alternative with the performance trait only. The models were compared based on cross-validation of standardized earnings, using mean-squared errors of prediction (MSEP) and the correlation between the phenotype (Y) and the estimated breeding value (EBV). Despite possible effects of selection, a rather high estimate of heritability of EARN was found in our univariate analysis. The genetic trend estimate for EARN was clearly higher in the bivariate specification than in the univariate model, as a consequence of the considerable size of estimated heritability of RACE and its high correlation with EARN (approximately 0.8). RACE is highly influenced by ancestry rather than the on-farm performance of the horse itself. Consequently, the use of RACE in the genetic analysis may inflate the genetic trend of EARN because of a double counting of pedigree information. Although, because of the higher predictive ability of the bivariate specification, the improved ranking of animals within a year-class and the inability to discriminate between models for genetic trend, we propose to base prediction of breeding values on the current bivariate model.

Asunto(s)

Asunto(s)