| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ANIMAL GENETICS |
Department of Animal Breeding and Genetics, University of Veterinary Medicine, Hannover (Foundation), Hannover, Germany
 |
Abstract |
|---|
 Top Abstract INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONIMPLICATIONSLITERATURE CITED |
|---|
Key Words: gait • genetic parameter • jumping • performance • radiography • Warmblood horse
|
INTRODUCTION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONIMPLICATIONSLITERATURE CITED |
|---|
). Accordingly, performance evaluation plays an important role in selection of horses for breeding. Performance tests may be held in the field or at station. The natural abilities of the horses are most objectively judged if the environmental influences on the performance of the individual horse are minimized. Opportunities for minimization of environmental influences on performance argue for station tests of several weeks in duration, but economic and practical reasons (e.g., limited test capacities) argue against station tests. Accordingly, 1-d field tests of mares are much more common than station tests.
Dressage and jumping performance are also important for appointment of horses for sale at riding horse auctions, with presentation and judgment procedures of horses being similar to mare performance tests in the field. Results of mare performance tests and inspections of auction candidates are therefore considered for routine genetic evaluation of Hanoverian Warmblood horses in Germany (Verband hannoverscher Warmblutzü chter e.V., 2005). However, the implied genetic identity of analogous traits has not been proven yet.
The presumed indicative value of radiographic findings in the equine limbs for future performance capacity is reflected by their effect on the sales value of a horse (Van Hoogmoed et al., 2003). Horses with abnormal radiographic findings in their limbs are considered to be at a greater risk to develop orthopedic problems than horses without such alterations. However, studies on the relationship between radiographic findings detected at a young age and present or later performance are missing for the riding horse.
Therefore, it was the aim of this study to determine the genetic correlations between prevalent radiographic findings in the limbs of young Warmblood riding horses and performance traits routinely evaluated in young horses of the same breed.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONIMPLICATIONSLITERATURE CITED |
|---|
Performance Data
Mare Performance Test.
Results of mare performance tests of the Association of Hanoverian Warm-blood breeders (Verband hannoverscher Warmblutzüchter e.V.; VhW) were available for the current study. Data included information on 12,661 performance-tested mares that were born in 1992 or later. Of these mares, 10,949 had completed their performance tests in the course of 1-d events in the field. Only 1,712 mares had participated in stationary performance tests, consisting of 18 d of standardized training and a final test on d 19. If a mare completed more than 1 performance test, only the last test results were considered in the data set provided.
Mare performance tests (MPT) were held in 1995 to 2004. The number of tested mares/yr ranged between 524 and 1,350 in mare performance tests in the field (MPT-F), and between 94 and 218 in mare performance tests at station (MPT-S). Information on mares of other breeds that were registered as broodmares by the VhW was included, but Hanoverians contributed more than 98% of the data.
The mares were judged for quality of gaits (walk, trot, canter) under rider, jumping talent (ability and style of free-jumping), rideability, and character using scores on a 0.5-point scale from 0 (not shown) to 10 (excellently shown). Ability and style of free-jumping were scored individually and averaged to a total score for free-jumping. Separate rideability scores were allocated by the judging commission and a test rider in MPT-F and MPT-S. Scores for interior (character) were allocated only in MPT-S.
Base information on the horses included place and date (d, mo, yr) of MPT and the age of the tested mare. There were 57 places of MPT-F, with on average 27.72 ± 14.51 (range 1 to 89) tested mares/d and place, but only 4 places of MPT-S, with on average 18.61 ± 5.90 (range 1 to 29) tested mares/d and place. In 3 of the 4 places of MPT-S, 3 or more evaluations/yr (mean 5.11 ± 3.38 evaluations/yr) were held, whereas in 95% of places of MPT-F only 1 or 2 evaluations/yr took place (mean 1.22 ± 0.59 evaluations/yr). Most of the mares completed their performance test before they were 4 yr of age (mean age 3.47 ± 0.85 yr for MPT-F and 3.35 ± 0.67 yr for MPT-F).
Auction Inspection.
Horses intended for sale at riding horse auction of the VhW must be presented to a judging commission, which decides on preliminary appointment of auction candidates. In a procedure similar to MPT-F, the horses receive scores for basic quality of gaits (walk, trot, canter) under rider, jumping talent (ability and style of free-jumping, total score for free-jumping), and rideability.
Information on place and date (d, mo, yr) of auction inspection (AU), sex, and the age of the inspected horse were available. Between 1999 and 2004, 4,773 Hanoverian Warmblood horses with a mean age of 3.87 ± 0.85 yr (range 2 to 12 yr) were judged on the occasion of AU, including 55 stallions, 1,308 geldings, and 3,410 mares. There were 108 places of AU, with on average 13.64 ± 16.04 (range 4 to 111) inspected horses/d and place. In 95% of places, AU took place only once or twice a year (mean 1.37 ± 1.08 evaluations/yr).
Radiographic Data
Horses preliminarily appointed for sale at riding horse auctions of the VhW routinely undergo a thorough, standardized, veterinary medical examination, which includes radiography of their limbs. In the course of selection for sale at riding horse auctions in 1997 to 2004, 5,102 Hanoverian Warmblood horses (137 stallions, 2,823 geldings and 2,142 mares) were examined at between 3 and 6 yr of age (mean age 3.14 ± 0.86 yr). More than 99% of these horses were born in 1992 to 2001. The number of radiographed horses per year, for which radiography results (RR) were available, ranged between 548 and 740.
Routine radiographic examination of the auction candidates included 10 standard radiographic projections of the limbs. Scrutiny of radiographs was independently carried out by 2 experienced radiologists. Greatest prevalence was determined for the following radiographic findings, which were considered for further analyses: osseous fragments in fetlock joints (OFF), osseous fragments in hock joints (OFH), deforming arthropathy in hock joints (DAH), and distinct radiographic findings in the navicular bones (DNB). Binary coding was used for OFF, OFH, DAH, and DNB, with 1 denoting the presence and 0 denoting the absence of the respective radiographic finding. Radiographic appearance of the navicular bones (RNB) was additionally considered as a quasi-linear trait with 8 categories (code 0 to VII), based on the location, number, and shape of radiologically visible canales sesamoidales (i.e., synovial invaginations along the distal border of the navicular bone, os sesamoideum distale) and the radiographic appearance of navicular bone structure and contour. Details on the radiographic examination procedure and the definition of radiographic findings can be found elsewhere (Stock and Distl, 2006a,b).
Combination of Performance and Radiographic Data
All radiographed horses and their contemporaries; i.e., horses with performance records (MPT, AU, or both) that were born between 1992 and 2001 (n = 18,609), were included in these studies. The number of performance records was 17,433, and the number of horses with performance records was 16,472. For 10,134 mares, only MPT-F records, for 1,566 mares only MPT-S records, and for 3,811 horses only AU records were available. The MPT-F and AU records were available for 815 mares, and MPT-S and AU records were available for 146 mares. The MPT-F, MPT-S, and AU records were used to investigate correlations between analogous performance traits.
Pooled performance records were used to investigate correlations between performance traits and radiographic findings. In the mares for which AU and MPT results were recorded, AU scores were used for the correlation analyses. There were 13,507 horses with performance records only, 2,137 horses (84 stallions, 1,526 geldings, and 527 mares) with RR records only, and 2,965 horses with both performance and RR records.
Pedigree Data
Pedigree information of all horses was provided by a unified animal ownership database (Vereinigte Informationssysteme Tierhaltung w.V.) in Verden on the Aller, Germany. Three ancestral generations of all horses with performance, radiography records, or both were considered for the genetic analyses. The relationship matrix comprised 50,725 horses, including 6,600 base animals.
The 18,609 horses included in this study descended from 932 sires and from 1,419 maternal grandsires. The distribution of all horses and of the horses with MPT-F, MPT-S, MPT, AU, and RR records among sires and maternal grandsires is given in Table 1. The percentage of stallions that occurred as both sires and maternal grandsires relative to the respective total number of sires ranged between 42.6 and 50.5%, and relative to the respective total number of maternal grandsires between 28.9 and 32.1%.
|
Distributions of MPT and AU scores and of residuals in a model with the fixed effect of age at performance evaluation were analyzed, including tests for normality using Kolmogorov-Smirnov statistics of the UNIVARI-ATE procedure (SAS Inst. Inc., Cary, NC). Information on 815 mares with AU and MPT-F records and 146 mares with AU and MPT-S records was used to compare AU and MPT evaluation results by means of Spearman rank correlation coefficients of the CORR procedure of SAS.
The influence of the presence or absence of OFF, OFH, DAH, and DNB, and of the value of RNB on the distributions of MPT and AU scores was investigated using data of radiographed horses with performance records and the GLM procedure of SAS:
 | [1] |
where yij = the MPT or AU score of the individual horse, µ= the model constant, RRi = the radiographic status with regard to OFF, OFH, DAH, or DNB (i = 0–1), or RNB (i = 0–7), and eij = the residual.
Simple and multiple ANOVA were performed. The performance traits and the quasi-linear RR trait RNB were analyzed in GLM and mixed linear models using the GLM and MIXED procedures, and the binary RR traits OFF, OFH, DAH, and DNB were analyzed in generalized linear models with binomial distribution and probit link function, using the GENMOD procedure of SAS.
The following effects were tested for their influence on the distribution of MPT and AU traits: Age group at MPT or AU evaluation (3-, 4-, or 
5-yr-old), evaluation year (individual years from 1995–2004), evaluation month (individual months), evaluation season (February through April, May through July, August through October, or November through January), evaluation place (108 places of AU, 56 places of MPT-F, and 3 places of MPT-S), combined year-place effect (238 levels for AU, 322 levels for MPT-F, and 17 levels for MPT-S), and combined date-place effect (350 levels for AU, 394 levels for MPT-F, and 91 levels for MPT-S). Sex effect (male, female) was tested with respect to the AU traits. Regression analyses were performed on age group and evaluation year.
For the simple and multiple ANOVA for the performance traits sex, age at MPT or AU evaluation, evaluation year, evaluation month, and evaluation season, the combined year-place was considered as a fixed effect, and the combined date-place was considered as a random effect.
Fixed effects tested for their influence on the prevalences of OFF, OFH, DAH, DNB, and RNB included examination age (3-, 4-, or 5-yr-old), sex (stallion, gelding, mare), interaction between sex and examination age, year of examination (individual years from 1997 to 2004), year of birth (1991–1992 and individual years from 1993 to 2001), season of birth (November to March, April, or May to October), interaction between sex and season of birth (18 combinations of sex and season of birth), individual date of presumable auction (49 riding horse auctions), and type of auction (elite auction, subsidiary auction, or Equitop-auction).
Estimation of Genetic Parameters
Models for the genetic analyses were developed on the basis of the results of significance tests and model fit test statistics. Genetic parameters were estimated multivariately in linear animal models with REML using VCE-5, version 5.1.2 (Variance Component Estimation; Kova
et al., 2003). The following models were used for the genetic analyses of the performance traits (Trot, Canter, Walk, RideC, RideR, FJA, FJS, FJT, Char) and the radiographic findings (OFF, OFH, DAH, DNB, RNB):
 | [2] |
 | [3] |
where y...s = the AU, MPT-F, or MPT-S score or radio-graphic finding of the individual horse, µ= the model constant, AGEAU/MPT I = the fixed effect of age group at performance evaluation (i = 1–3), SEXj = the fixed effect of sex (i = 1–2), dateAU/MPT x placeAU/MPT kl = the random effect of interaction between date of performance evaluation (kAU = 1–350, kMPT-F = 1–394, kMPT-S = 1–91, kAU/MPT = 1–535) and place of MPT evaluation (lAU = 1–108, lMPT-F = 1–56, lMPT-S = 1–4, lAU/MPT = 1–135), AUCTIONm = the fixed effect of presumable date of auction (m = 1–48), SEX x AGERR no = the fixed effect of interaction between sex (n = 1–3) and age at radiographic examination (o = 1–3), BIRTHn = the fixed effect of season of birth (P = 1–3), ar = the random additive genetic effect of the individual horse (r = 1–50,725), and e...s = the residual.
Fixed sex effect in model 2 was relevant only for AU, but not for MPT data. Fixed effects of the interaction between sex and age and season of birth were included for genetic analyses of OFH, DAH, DNB, and RNB but removed from model 3 for genetic analysis of OFF because they were not significant for this trait. Model 2 was used for the first set of genetic analyses including equivalent performance traits of MPT-F, MPT-S, and AU of horses with performance records and that were born after 1991. Analyses were performed trait-by-trait, so that genetic parameters were estimated in 7 trivariate analyses (Trot, Canter, Walk, FJA, FJS, FJT, RideC, evaluated on the occasion of MPT-F, MPT-S, and AU), 1 bivariate analysis (RideR, evaluated on the occasion of MPT-F and MPT-S), and 1 univariate analysis (Char, evaluated on the occasion of MPT-S).
Models 2 and 3 were used for the second set of genetic analyses, which included performance and RR traits. Bivariate analyses of all possible combinations of RR and performance traits were performed to obtain variance and covariance estimates. In each case, 1 RR trait and 1 performance trait were considered, so that each RR trait was included in 9 analyses and each performance trait was included in 5 analyses. Because differences between corresponding heritability estimates were negligibly small (mostly 
0.01; 0.05 for Char), only mean heritabilities (h2) and mean SE of heritabilities (SEh2) will be reported for each data set.
Underestimation of heritabilities and residual correlations in analyses of binary traits in linear models was compensated for by transformation of estimates and SE of heritabilities and residual correlations relating to the radiographic findings (OFF, OFH, DAH, RNB) to the underlying liability scale, according to Dempster and Lerner (1950) and Vinson et al. (1976). Applicability of transformation factors to the analyzed data has been proven in a previous study (Stock et al., 2005).
 | [4] |
 | [5] |
where h2liab (re liab) = the heritability of trait i (residual correlation between traits i and j) on the underlying continuous (liability) scale, h2obs (re obs) = the heritability of trait i (residual correlation between traits i and j) on the observed (binary) scale, pi (pj) = the frequency of outcome 1 for trait i (j), and zi (zj) = the ordinate of a standard normal distribution at the threshold point corresponding to a fraction pi (pj) of the population having the character.
|
RESULTS |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONIMPLICATIONSLITERATURE CITED |
|---|
Distribution of MPT-F, MPT-S, and AU scores is shown in Table 2. Ranges of scores were larger in MPT-F and AU than in MPT-S. Smallest differences between minimum and maximum score were observed for Char, Trot in MPT-S, and Canter in MPT-S. Maximum range of 1.0 to 10.0 was observed for FJA in AU and for RideR in MPT-F. Mean AU scores of all traits were considerably lower than both mean MPT-F and mean MPT-S scores. Mean scores for gaits (Walk, Trot, Canter) were lower in MPT-F than in MPT-S, mean scores for RideC were similar in MPT-F and MPT-S, and mean scores for RideR and jumping (FJA, FJS, FJT) larger in MPT-F than in MPT-S.
|
0.05) for scores and residuals for Trot, Canter, and Walk in MPT-S.
Results of comparison between AU and MPT scores in those mares of birth years 1992–2001, for which information on more than 1 performance test were available, are given in Table 3. Mean scores for gaits (Walk, Trot, Canter) and RideC differed by 0.37–0.78 between AU and MPT-F and by 0.62–1.07 between AU and MPT-S. Differences between mean scores for jumping (FJA, FJS, FJT) were considerably larger, namely 1.76–2.16 between AU and MPT-F and 1.65–2.10 between AU and MPT-S. However, correlations between analogous performance traits were all significantly positive (P < 0.001), ranging between R = 0.30 and R = 0.55 for AU and MPT-F and between R = 0.24 and R = 0.65 for AU and MPT-S.
|
Analyses of Variance and Model Development
Performance Data.
In the simple fixed effect AN-OVA all tested effects, except for sex in respect of the jumping traits (FJA, FJS, FJT) and season of AU in respect of RideC, had an influence on AU scores (P = 0.01 for season of AU in respect of Canter, P < 0.001 in all other cases). The MPT-F scores were influenced by age group, year, month and place of MPT-F (P < 0.001). Dependence of scores on season of MPT-F was determined for Trot, Canter, Walk, and FJA (P 0.02). Most MPT-S scores were dependent on year and place of MPT-S (P < 0.01). Age group at MPT-S had an influence on scores for Trot, Canter, and FJA (P 0.07).
Probability to receive high scores for Trot, Canter, Walk, and RideC in AU and MPT-F decreased, and probability to receive high scores for FJA, FJS, and FJT in AU and MPT-F increased significantly with age group. The MPT-S scores showed less age dependence than MPT-F and AU scores. Regression coefficients calculated for year of performance evaluation were mostly close to zero (b = – 0.01 to 0.06), but clearly negative for jumping scores FJT, FJA, and FJS in AU (b = – 0.16).
In the multiple ANOVA fixed effect of age group explained similar proportions of observed variance of AU, MPT-F, and MPT-S scores in combination with random combined year-place effect and in combination with random combined date-place effect (1.57 to 9.69%).
Radiographic Data.
The OFF prevalence in the 5,102 radiographed horses was influenced by date of presumable auction, year of examination, and year of birth (P < 0.01), OFH prevalence by date of auction, year of examination, year of birth, season of birth, sex, and sex x age interaction (P < 0.05), DAH prevalence by date of auction, year of examination, year of birth, season of birth, sex, and sex x age interaction (P < 0.01), and DNB prevalence and RNB score by date of presumable auction, year of examination, year of birth, season of birth, sex, examination age, and sex x age interaction (P < 0.001).
Joint Analysis of Performance and Radiographic Data.
Investigation of the influence of presence or absence of radiographic findings on the distributions of scores for performance traits revealed no significant differences between horses affected and horses not affected by OFF, OFH, and DAH in respect of scores for Trot, Canter, Walk, RideC, FJT, FJA, and FJS (P = 0.09 to 0.94). However, horses affected by DNB received on average lower scores than horses not affected by DNB for Trot (LSmean 6.77 in affected and 6.88 in unaffected horses; P = 0.004) and Walk (LSmean 6.85 in affected and 6.97 in unaffected horses; P = 0.005).
Genetic Analyses
Performance Data.
Results of multivariate genetic analyses of performance traits evaluated on the occasion of MPT-F, MPT-S, and AU are given in Tables 4 and 5. Additive genetic variances of MPT scores for Walk, Trot, Canter, RideR, RideC, and Char (
2a = 0.18 to 0.28 for MPT-F, 
= 0.07 to 0.28 for MPT-S) were smaller than of MPT scores for FJT, FJA, and FJS ( = 0.37 to 0.49 for MPT-F, = 0.52 to 0.55 for MPT-S) and of all AU scores ( = 0.08 to 1.33 for AU). Heritability estimates were h2 = 0.22 to 0.35 for Walk, h2 = 0.34 to 0.48 for Trot, h2 = 0.26 to 0.41 for Canter, h2 = 0.14 to 0.46 for RideC, h2 = 0.25 to 0.27 for RideR, h2 = 0.39 to 0.61 for FJT, h2 = 0.39 to 0.60 for FJA, h2 = 0.32 to 0.58 for FJS, and h2 = 0.30 for Char. Heritability of scores for Trot, Canter, Walk, and RideC were lowest in AU data, intermediate in MPT-F data, and greatest in MPT-S data, and heritability estimates for RideR were larger in MPT-S than in MPT-F data. Conversely, heritability estimates for FJT, FJA, and FJS scores were considerably lower in MPT-F than in AU and MPT-S data. The SE of heritabilities ranged between S
= 0.01 and S = 0.05.
|
|
0.75. Additive genetic correlations between MPT-F and MPT-S scores were closed to unity (rg = 0.87–1.00) but slightly lower between MPT-F and AU scores (rg = 0.81–0.90) and between MPT-S and AU scores (rg = 0.75–0.92).
Joint Analyses of Performance and Radiographic Data.
Results of multivariate genetic analyses of performance and RR traits are given in Tables 6 and 7. Heritabilities for scores for gaits, rideability, jumping, and character closely resembled those that were estimated for respective MPT-F scores. Range of heritabilities was 0.23 to 0.39 (S = 0.01 to 0.03). Heritability estimates for radiographic findings ranged between 0.05 and 0.11 (S = 0.01 to 0.02) before transformation and between 0.10 and 0.34 (S = 0.02 to 0.08) after transformation.
|
|
0.10). Significantly negative additive genetic correlations were determined between occurrence of OFH and scores for Char (rg = – 0.17, SErg = 0.11), and between occurrence of DAH and scores for Canter, RideR, FJT, FJA, FJS, and Char (rg = – 0.31 to – 0.18; SErg = 0.06 to 0.16). Ranges of residual correlations between performance traits except for Char and RR traits were re = – 0.03 to 0.07 (SEre = 0.01 to 0.03) before transformation and re = – 0.04 to 0.10 (SEre = 0.02 to 0.05) after transformation, whereas residual correlations between Char and RR traits ranged between re = 0.06 and re = 0.14 (SEre = 0.01 to 0.09) before transformations and between re = 0.08 and re = 0.27 (SEre = 0.01 to 0.15). |
DISCUSSION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONIMPLICATIONSLITERATURE CITED |
|---|
The breeding aim of the Hanoverian Warmblood horse is defined as follows: a horse that is on the basis of its temperament and character, rideability, conformation, movement, jumping abilities, and health suitable as a performance horse as well as a pleasure horse (VhW, 2005). Focus of breeders is on production of sport horses, particularly of horses suited for use in the disciplines dressage, show-jumping, and eventing. Different testing schemes have been developed for stallions and mares. Performance testing of stallions has a long tradition, is obligatory for sires of most breeds, and usually takes place at station during a predefined period of time. Performance testing of mares has been introduced in Germany in 1983 for the Holstein Warmblood (Nissen and Kalm, 1986) and in 1987 for the Hanoverian Warmblood (Christmann, 1996) on a voluntary basis and might be done at station or in the field. Moderate heritability estimates, mostly ranging between 0.15 and 0.60, have been reported for those traits that characterize the performance of a riding horse and are routinely evaluated in large numbers of young horses (e.g., Huizinga et al., 1990; Willms et al., 1999; Gerber Olsson et al., 2000). Results of this study agree with previous estimates found in the literature.
Reduced genetic variance (i.e., increased genetic homogeneity) in respect of the basic characteristics of a riding horse, namely gaits, rideability, and character, might indicate successful selection and breeding progress of the last decades regarding these traits (Wright 1935, Crow and Kimura, 1970). Average breeding values for dressage increased, average breeding values for jumping were almost constant in the last generations, and remarkable discrepancies between breeding values for dressage and show-jumping are frequently seen in the current Warmblood horse population (Stock and Distl, 2005b). Specialization, i.e., preference for breeding dressage or jumping horses rather than all-round horses and 1- rather than 2-trait-selection, may have increased the difference between best and worst jumpers and therefore be responsible for the large overall genetic variance in respect of the jumping traits. However, comparison with results of genetic analyses of MPF data from 1989–2004 (data not shown) revealed that decrease of genetic variances had little impact on the heritability estimates. Only contemporaries of radiographed horses, i.e., horses born after 1991 and accordingly performance-tested from 1995 to 2004, were considered further in order to ensure data homogeneity and consistency of genetic analyses. Results of regression analyses revealed absence of significant trends of scores over the years for most traits. Only average jumping scores showed a significant decrease, which may indicate dominance of selection for dressage performance over selection on jumping performance in this breed of Warmblood horses.
Given the relevant genetic determination of riding horse features in the Warmblood horse, much importance is attached to the performance of breeding animals and their offspring. Although specific testing procedures differ between countries and breeding organizations, there are basically 2 different types of performance tests: 1-d field-tests and stationary performance tests of at least 2-wk duration. In this study these test types were represented by auction inspections (AU), i.e., inspections of young horses intended for sale at riding horse auctions, and mare performance tests (MPT) in the field (MPT-F) on the one hand and mare performance tests at station (MPT-S) on the other hand. For horse breeders riding horse auctions provide an opportunity to present their young horses to a large number of potential buyers; for breeding organizations they provide an opportunity to watch large numbers of young horses, regardless of their intended use for breeding. Use of mares for breeding requires studbook registration, but participation in MPT is voluntary. The percentage of performance-tested Hanoverian mares increased from 26% in 1987 to 58% in 1995, so that performance of the tested mares should be rather representative for the whole population of Hanoverian Warmblood horses. Marked differences between AU and MPT-F or AU and MPT-S means and ranges of scores for performance traits may not necessarily be referred to some general preselection effect. Considerable differences of scores were also seen in those mares with AU and MPT records.
Own performance of sires is considered for routine genetic evaluation of German Warmblood horses. However, the number of tested mares per year is far beyond that of tested stallions. Furthermore, most mares absolve their MPT at 3 or 4 yr of age. The MPT results therefore provide an early and valuable source of information for breeding purposes and are considered for national genetic evaluation of German Warmblood horses since 1998 (Von Velsen-Zerweck and Bruns, 1998). Highly positive additive genetic correlations between MPT-F and MPT-S scores, which were previously determined (rg = 0.72 to 0.92; Christmann, 1996) and substantiated by the results of this study (rg = 0.86 to 1.00), justified their equivalent consideration. Deviation of distributions of scores from normality is generally ignored because influence on the results of genetic analyses is not considered to be of importance. Performance evaluation procedures are very similar in AU and MPT-F. Accordingly, AU scores were considered in addition to MPT-F and MPT-S scores in the internal genetic evaluation of Hanoverian Warmblood horses since 2002. Investigations on adequacy of equivalent consideration of AU, MPT-F, and MPT-S data had not been performed yet. According to the results of this study, genetic equivalence of performance traits evaluated on the occasion of AU, MPT-F, and MPT-S can be assumed, but systematic influences of type of performance test and age (MPT-F, MPT-S) or age and sex (AU) have to be allowed for.
Horses preliminarily appointed for sale at riding horse auctions of the VhW routinely undergo a standardized veterinary medical examination. Osseous fragments in fetlock and hock joints, deforming arthropathy in hock joints, and distinct radiographic findings in the navicular bones represent radiographic findings that occur with greatest prevalences in the limbs of these young Warmblood horses (Stock and Distl, 2005a). Relevant genetic determination of these radiographic findings has been identified previously (Stock et al., 2004a,b, 2005), and heritabilities estimated in this study are in good agreement with previous results.
Favorable dressage and jumping performance was found to have a positive effect on longevity of riding horses (Wallin et al., 2001), but investigations on relations between health and long-term performance of riding horses are rare and mostly confine to general health aspects. There is some controversy on the clinical relevance of deviations from normal radiographic appearance of parts of the equine skeleton. Conflicting results have been obtained in studies on the long-term effect of particular radiographic findings on the performance of racehorses (Grøndahl and Engeland, 1995; Storgaard Jørgensen et al., 1997; Roneus et al., 1998) and riding horses (Möller, 1993). Results of previous studies on the performance of young Warmblood riding horses in relation to their orthopedic status did not supply strong evidence for a negative effect of limb alterations on performance (rg = – 0.04 to 0.20; Holmström and Philipsson, 1993; Wallin et al., 2003). However, significantly negative additive genetic correlations that have been determined between prevalent radiographic findings in the equine limbs argue against consideration of overall orthopedic status without distinction between individual conditions in respect of type and location of possible alterations. In one study on Holstein Warmblood mares individual radiographic findings were considered, but low numbers of mares with information on radiographic findings (n = 472) and performance in MPT (n = 1,240) interfered with reliable estimation of genetic parameters (Willms et al., 1999). Absence of significant genetic correlations between occurrence of osseous fragments in hock joints (osteochondrosis dissecans tarsi) and MPT scores largely agreed with our results. Disagreement between study results refers to radiographic findings in the navicular bones (radiologically diagnosed podotrochlosis) and hock joint deformations (radiologically diagnosed spavin). Low, but significant, positive additive genetic correlations that were reported between navicular bone alterations and MPT scores for walk and free-jumping (Willms et al., 1999) were not seen in the current study. Reported absence of significant genetic correlations between spavin and MPT scores for walk, trot, canter, rideability, free-jumping, and character (Willms et al., 1999) were opposed to clearly negative additive genetic correlations between hock joint deformations and scores for trot, canter, rideability, free-jumping, and character determined in this study. However, character scores were available only for the station-tested mares. Genetic correlations between character and radiographic findings therefore require further investigation.
Standardized testing conditions are more likely to be encountered in MPT and AU than in regular tournament competitions. Use of MPT and AU data should therefore facilitate the distinction between genetic and environmental variation of performance traits and the identification of genetic correlations between performance traits and concurrently considered radiographic health traits. The results of this study support the results of previous investigations on relationships between radiographic findings observed in the limbs of young horses and performance in tournament sport (Stock and Distl, 2005c). Genetic correlations were mostly close to zero, but there were indications of some negative genetic correlations, amongst others between deforming arthropathy in hock joints and performance of young riding horses in respect of use and success in basic build-up competitions. Negative or absent genetic correlations between prevalent radiographic findings and performance of riding horses substantiate reported compatibility of selection for radiographic health and performance traits (Stock and Distl, 2005a,b).
|
IMPLICATIONS |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONIMPLICATIONSLITERATURE CITED |
|---|
1 Corresponding author: Kathrin-Friederike.Stock{at}tiho-hannover.de
Received for publication October 20, 2005. Accepted for publication August 29, 2006.
|
LITERATURE CITED |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONIMPLICATIONSLITERATURE CITED |
|---|
, M., E. Groeneveld, and A. Garcia-Cortéz. 2003. VCE-5 User’s Guide and Reference Manual Version 5.1.2. Institute for Animal Science and Animal Husbandry, Federal Agricultural Research Centre (Bundesforschungsanstalt fü r Landwirtschaft, FAL), Mariensee /Neustadt, Germany.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
