82 E
. Gerber Olsson et al. Livestock Production Science 65 2000 81 –89
for soundness, conformation, gaits under rider, free genetic trend in the population based on BLUP-
jumping and jumping under rider. The stallions were Animal model evaluations is estimated with infor-
also scored for temperament and general appearance mation from the RHQT and the trend has been
´ in the tests for gaits under rider, free jumping and
increasing rapidly during the last 10 years Arnason jumping under rider. The stallions usually had to be
et al., 1997. When considering horses born from approved at two out of three tests to be licensed for
1988 to 1992, the annual genetic progress for gaits breeding. Annually about 200 stallions were pre-
and jumping, respectively, have been about 0.05 and inspected for conformation and soundness. About
0.03 genetic standard deviations per year. 40–60 were selected to take part in the stallion
The purpose of the SPT is to select sires of the performance tests and on average 8–12 of these
coming generation of successful sport horses. As stallions were finally approved for breeding.
Ohlsson and Philipsson 1992 have concluded, In 1997 some changes were introduced in the
performance at RHQT is highly correlated to later stallion performance test. The test is now done as a
competition performance, it is of great interest to 9-day-long station test at 4 years of age at Flyinge.
find out to what extent the SPT and RHQT results The stallions are preselected in the autumn at 3.5
are genetically related. years of age, for conformation, soundness and free
The objectives of this study were to estimate jumping. The latter trait is therefore excluded from
heritabilities, genetic and phenotypic correlations and the regular stallion performance test. The stallions
repeatabilities for traits scored in the stallion per- are also judged by foreign test riders in the new test.
formance tests as well as correlations between The SPT is used for phenotypic selection among
stallion performance test results and 4-year-old field prospective breeding stallions, but until now it has
performance records of their offspring. These param- not been investigated if heritabilities and variances
eters are necessary for construction of integrated for the traits tested are high enough for an efficient
indexes to be used for performance evaluations and selection. Because the SPT was done as repeated
continued evaluations of stallions based on early tests
it was
important to
investigate the
re- progeny results. The parameters are also useful for
peatabilities of the results, and thereby determine the determining the opportunities of improving or sim-
effects of reducing the number of tests. The correla- plifying the stallion performance tests.
tions between traits in the Swedish stallion per- formance test have not been estimated before and it
would be interesting to know their interrelationships
2. Materials
in order to make optimal use of the information for selection purposes, which may indicate if the tests
2.1. Stallion performance test data can be further simplified and less costly.
Riding horse quality tests RHQT, a 1-day field A dataset comprising 378 stallions evaluated at the
test for 4-year-old horses, have taken place in SPT during 1979–1993 was analysed. The stallions
Sweden since 1973. Annually about one third, or took part in one to three performance tests. The
500–1000, of all the 4-year-old riding horses in dataset contained observations from a total of 683
Sweden participate at the tests which take place at tests and was used to estimate heritabilities, genetic
about 20 different places every autumn. The traits and phenotypic correlations and repeatabilities be-
scored are health status, conformation, gaits under tween tests. The number of observations in the
rider and jumping. A temperament and general dataset is shown in Table 1. All traits were given
appearance score is given at the latter two parts of scores between 1 and 10. The temperament and
the tests. All traits are given scores between 1 and general appearance scores for the two jumping traits
10. The tests are more precisely described by Gerber have only been recorded since autumn 1985 and the
et al. 1997. The RHQT system aims at genetic corresponding score for gaits under rider since
evaluations of stallions and mares and of providing autumn 1988. Also within a test some stallions were
information about the quality of the individual young not shown in jumping under rider if the free jumping
horses as a guideline for choice of sport horses. The result was poor. The scores for conformation and
E . Gerber Olsson et al. Livestock Production Science 65 2000 81 –89
83 Table 1
Number of observations per trait and combining of traits in the data from the stallion performance test Trait
1 2
3 4
5 6
1 Walk, trot, gallop 536
a
2 Temp. gaits 217
217 3 Free jumping
458 596
4 Temp. free jumping 189
478 478
5 Jumping under rider 315
278 235
315 6 Temp. jumping under rider
174 239
236 264
266
a
Temp., temperament and general appearance.
soundness were not included in the material as the tions from the SPT contained 683 records for gaits
stallions were pre-selected mainly on these criteria. and jumping. There were 234 stallions with offspring
Means and standard deviations for the different traits in the RHQT. Of these, 136 had a minimum of five
scored in SPT are given in Table 2. offspring in the test and were included in the
The evaluated stallions had 160 sires and 167 correlation analysis with RHQT results. There were
maternal grandsires, 60 of them appeared as both 2964 mares tested in the RHQT with offspring, that
sires and maternal grandsires. Among the sires and also had participated in the RHQT.
maternal grandsires 56 had themselves participated For estimation of the genetic correlations between
in the SPT. SPT and RHQT traits new combined traits were
created from the original ones. The trait ‘gaits’ was 2.2. Stallion performance test and riding horse
created as the average of the four scores for walk, quality test data
trot, gallop and temperament and general appearance, for the data from the RHQT and SPT. The score for
To estimate the genetic correlations between traits ‘jumping’ was in the RHQT defined as the average
tested at SPT and RHQT, respectively, the material of either two or four scores for free jumping and
from the SPT, described above, and results from the jumping under rider both were not always re-
RHQT during the period 1983–1993 were merged. corded, and temperament and general appearance
Results for mares with foal at foot when tested were for the two disciplines whenever recorded. Approxi-
excluded from the analysis. The final dataset for this mately 50 of the horses were tested for free
analysis comprised 6674 horses tested for gaits under jumping only, 40 for jumping under rider only, and
rider and for jumping in the RHQT. In the SPT, 327 the remaining had scores for both traits. In the SPT
and 359 stallions were tested for gaits under rider data the average of all four scores for the jumping
and jumping, respectively, and another 8559 horses traits were used. However, at the age of 4.5 years
occurred in the pedigrees. The dataset with observa- and above, the free jumping test was exchanged for
Table 2 Means, standard deviations S.D., minimum Min and maximum values Max and coefficients of variation CV of the different traits in
the data from the stallion performance test Trait
Means S.D.
Min Max
CV Walk
6.91 1.18
3 10
17.1 Trot
6.72 1.29
4 10
19.2 Gallop
7.09 1.09
5 10
15.4
a
Temp. gaits 6.52
1.21 1
10 18.6
Free jumping 6.49
1.72 1
10 26.5
Temp. free jumping 6.56
1.87 1
10 28.5
Jumping under rider 6.70
1.56 2
10 23.3
Temp. jumping under rider 6.61
1.68 1
10 25.4
a
Temp., temperament and general appearance.
84 E
. Gerber Olsson et al. Livestock Production Science 65 2000 81 –89 Table 3
Number of horses N , means, standard deviations S.D., minimum Min and maximum Max values and coefficients of variation CV for the average scores of different traits in the combined data set from the stallion performance test SPT and riding horse quality test
RHQT
Trait average score N
Mean S.D.
Min Max
CV Gaits RHQT
6674 6.52
0.89 3.00
9.33 13.7
Jumping RHQT 6674
6.79 1.50
0.50 10.00
22.1 Conformation RHQT
6674 7.47
0.49 5.00
9.60 6.6
Gaits SPT 136
6.83 0.92
4.33 9.67
13.5 Jumping SPT
136 6.24
1.63 1.00
10.00 26.2
another jumping test under rider. All tests at the the random effect of the kth tested stallion, mean50
2
different ages were included and means were calcu- and variance5 As ; pe
is the random effect of the
a kl
lated for each created trait. The trait ‘conformation’ permanent environment, mean50 and variance5
2
in the RHQT data was defined as the average of five s ; e
is the random residual effect, mean50 and
pe ijkl
2
scores comprising type, head-neck-body, legs, walk variance5s .
e
and trot at hand. Table 3 contains means and Repeatability for the traits scored in the SPT was
2 2
2 2
2
standard deviations for the traits scored. estimated
as s 1 s s
where s 5 s 1
a pe
P P
a 2
2 2
s 1
s , and s is the additive genetic variance.
pe e
a
Standard errors of the variance and covariance components were provided by the DMU programme
3. Methods