Weaning results of beef Hungarian Fleckvieh calves 2 . Genetic parameters , breeding values

Weaning weight, preweaning daily gain and 205-day weight of Hungarian Fleckvieh calves (n=8 929, bulls =4 539, heifers =4 390) born from 232 sires between 1980 and 2003 were examined. Variance, covariance components and heritability values and correlation coefficients were estimated. The effect of the maternal permanent environment on genetic parameters and breeding values were examined. Two animal models were used for breeding value estimation. The direct heritability (h d) of weaning weight, preweaning daily gain and 205-day weight was between 0.37 and 0.42. The maternal heritability (h 2 m) of these traits was 0.06 and 0.07. The direct-maternal correlations (rdm) were medium and negative −0.52 and −0.74. Contribution of the maternal heritability and maternal permanent environment to phenotype is smaller than that of direct heritabilities (h 2 m+c <h d). The ratio of the variance of maternal permanent environment in the phenotypic variance (c2) changed from 3 to 6 %. Estimated breeding values changed whether the permanent environmental effect of dam wasn’t taken into consideration but the rank of the animals was not modified. The genetic value for weaning results of Hungarian Fleckvieh population has increased since 1997.


Introduction
The efficiency of beef cattle, resp.slaughter cattle production can be considerably influenced by the weaning weight of calves in the Hungarian Fleckvieh farms, too.Therefore the evaluation of these traits and the estimation their genetic parameters are very important.From these parameters the value of heritability is important, which is the ratio of total phenotypic variance of a given trait, which can be ascribed to the genetic variance.
The heritability is typically of a population kept in a certain region, therefore if we want to make a breeding program for a breed we have to calculate it for the given population and environment.Besides, the environment the heritability of a given trait is influenced by the height of genetic variance.It considerably depends on the sire, the relatives and the grade of inbreeding.
The heritability value (h 2 ) depends on the method of calculation, too.The ways of evaluation take apart the diverse variance components with a different exactness whereby the residual variance can be lower or higher.
The Table 1 shows the heritability values of gain and weaning weight at Simmental breed.These results were found by some researchers.The table includes the name of author(s), breed, and country, method of estimating and direct and maternal heritability of traits.
At present the connection between maternal and genetic effects is of the mentioned in the foreign literature.The (co)variance, resp.correlation between the two effects has different degrees and senses.
Having examined the weaning weight of Angus population MEYER (1992) found the correlation between the direct and maternal genetic effect 0.22.In their examinations NÚNEZ-DOMINGUEZ et al. (1993) found in Angus and Hereford populations the value of correlation in case of weaning weight 0.25 and 0.63 respectively.VAN VLECK et al. (1996) examined the correlation between the direct and the maternal genetic effect in Charolais, Simmental, Red Poll and Braunvieh population.As a result they got a correlation 0.46, 0.16, 0.31 and 0.25 respectively.CUBAS et al. (1991) found a correlation −0.93 between the two effects in an Angus population.DODENHOFF et al. (1999) report on a correlation value −0.12, −0.37, −0.18, and −0.10 respectively in case of weaning weight in Charolais, Hereford, Limousin, and Simmental population.(2001( ), LEE et al. (1997b)), MEYER (1992), NÚNEZ-DOMINGUEZ et al. (1993), VAN VLECK et al. (1996)  The aim of the present investigation was to estimate the variance, (co)variance components and genetic parameters of weaning weight, preweaning daily gain and 205day weight by the means of two animal models in two Hungarian Fleckvieh populations.Furthermore the aim was to make a comparison between the two animal models, i.e. to estimate whether the maternal permanent environmental effect has an influence on the genetic parameters and the estimated breeding value of animals.Bedsides, the aim was to know what genetic trend is to be observed in the weaning performances of Hungarian Fleckvieh populations.

Material and methods
In the present examinations three traits were evaluated: the weaning weight, the preweaning daily gain and the 205-day weight.The evaluation includes the data of 8 929 calves (4 539 bulls and 4 390 heifers) of 232 Hungarian Fleckvieh sires born between 1980 and 2003.The data were given by the Association of Hungarian Fleckvieh Breeders.The composition of the examined population is showed in the Table 2.The following variance and (co)variance components and genetic parameters have been estimated: T ) -direct maternal genetic correlation (rdm) -ratio of permanent environmental variance in the phenotypic variance(c 2 ) -ratio of residual variance in the phenotypic variance (e 2 ).The parameters and breeding values were estimated with animal model.The applied model (model 1) was in case of all traits the following: where y is the vector of observation (trait), b is the vector of fixed effect(s), u is the vector of random effect (animal), m is the vector of maternal genetic effect, pe is the vector of maternal permanent environmental effect, e is the vector of random residual effect, X is the matrix of fixed effects, Z is the matrix of random effects, W is the matrix of maternal genetic effect, and S is the matrix of maternal permanent environmental effect.
In the course of evaluation by an animal model the fixed effects were herd, age of dam, year, season and sex.In case of weaning weight and preweaning daily gain the effect of weaning age as (co)variant was taken into consideration, too.
Having examined the maternal permanent environmental effect two models were applied.One of them was an animal model, identical with the above mentioned model, the other animal model (model 2) was different, and it didn't include the maternal permanent environmental effect.
With the two animal models the effect of maternal permanent environmental effect was analysed on estimated parameters, breeding value of animals and their rank if it is taken into account or if it is left out of consideration.The influence on the rank of animals was determined by calculation of rank correlation.
The genetic trend of weaning results of Hungarian Fleckvieh was stated according to the means of breeding values referring to the year of birth.The animal model estimates a breeding value for each animal (dam, sire, calf ).To determine the genetic trend the breeding value of the animal, born in the same year based on the direct genetic effect was averaged and the obtained points were represented in a system of co-ordinates.
The variance and (co)variance components, genetic parameters and breeding value were estimated with the DFREML (MEYER 1998) and MTDFREML (BOLDMAN et al. 1993) programs.For the rank-correlation calculation the SPSS 9.0 (1996) program was applied.

Results and discussion
The overall mean weaning results are shown by the Table 3.The mean weaning weight of the Hungarian Fleckvieh calves was 217 kg, the preweaning daily gain 1.009 kg/day, the 205-day weight 242 kg and the mean weaning age 181 days.

Variance and (co)variance components, genetic parameters
The Table 4 includes the variance and (co)variance components and the genetic parameters estimated with the two different animal models (model 1 and model 2).
The table shows that the (co)variance between the direct additive genetic effect and the maternal genetic effect was negative at all the three traits, so the sense of the correlation between the two effects is negative, too.The correlation coefficient, rdm changed between −0.52 and −0.72, i.e. there is a close negative connection between the two effects, which is similar to the results of VAN VLECK et al. (2002).
The direct heritability of weaning weight, preweaning daily weight and 205-day weight h 2 d =0.40, 0.42 and 0.37 respectively, these values are higher, than the results of LENGYEL et al. (2003) (h 2 d =0.10, 0.13 and 0.14 respectively).The maternal heritability of the examined traits h 2 m =between 0.06 and 0.07.The maternal permanent environmental effect in the phenotype (c 2 ) is between 3 and 6 %.The maternal genetic effect and the maternal permanent environmental effect are together (h 2 m +c 2 ) 0.09 -0.12, which is lower than the result of LENGYEL et al. (2003) in Limousine populations (0.15 -0.18).The total heritability of the examined traits h 2 T =0.24 -0.27.The residual variance in the phenotype (e2) changed between 0.59 and 0.62.

Comparison the models
The results prove that there is a difference between the breeding values estimated by two different models.Therefore it was examined by rank-correlation whether the said difference causes a change in the rank of the animals.
According to the obtained rank-correlation coefficients (Table 5) (r=0.95,0.94, 0.92; P<0.01) it seems that the rank of animals stated according to their breeding value wasn't influenced by the model in spite of the mentioned significant difference, because there is a close connection between the two ranks.

Breeding values
The Table 6 includes the estimated breeding value of the examined sires according to the additive direct and maternal genetic effect.According to the breeding value estimated on the direct genetic effect the best was the sire No. 12316, whose breeding value was by +24.0 kg, +0.15 kg/day and +22.2 kg higher than the mean value of the population.The worst was the breeder bull No. 12565, whose breeding value was −51.6 kg, −0.25 kg/day and −56.3 kg in the examined traits.The Table 7 includes the rank-line of sires according to the breeding value estimated on the direct maternal effect, taking the two different models into consideration.It can be laid down that the sires whose breeding values estimated on maternal effect (e.g.No. 6455 and No. 12565) is the best, are the last in the rank-line according to the breeding value estimated on the direct effect, and it is true in the opposite direction too (e.g.9683,12316).This can be explained with the mean, resp.close negative correlation (rdm= between −0.52 and −0.72).The sire No. 10811 can be counted as an exception because it is the 6th in the rank-line according to the breeding value estimated on the maternal effect, but he is in the middle of the rank-line (11th) according to the breeding value estimated on the direct effect.In the table it can be seen that there is not any considering deviation between the ranklines, established according to the breeding values based on the direct effect.In the rank line of breeding values estimated on the maternal effect can be found a higher deviation, e.g. the sire No. 12938, which is on the 4th place according to the 205-day weight in the model 1, but he is on the 8th place, when he was estimated by the model 2. It can be observed in case of the sires No. 5385 and 14180 too.In spite of the mentioned differences the rank-correlation in the rank-line of the breeding values estimated on the maternal effect didn't show significant deviation.
According to the obtained results it can be stated that the sire whose breeding values estimated on the maternal effect are good, has similar breeding values according to the preweaning daily gain and 205-day weight (e.g. the sire No. 10812 was 5th in the rankline according to the weaning weight, 7th according to the preweaning daily gain and 8th according to the 205-day weight).

Change of the genetic value of the population (genetic trend)
The Figure 1 shows the change of the breeding value of the examined population estimated according to the additive direct genetic effect in case of the 205-day weight by years.The decrease in the years 1993-1994 could be caused by the sire No. 12565 (above in the Tables 7 and 8), which had a spoiling influence.From the 316 calves, born in 1994 124 calves descended from it.Calves born in the same year from sire No. 10457, 12929 and 12930, which had a spoiling effect, too and which weren't included in the tables owing to the lower number of calves.Almost 60 % of the calves born in 1994 descended from a sire with lower breeding value.The remaining 40 % of the calves descended from sires with mean breeding value (e.g.No. 12916) and they had a breeding value near 0. The genetic trend was determined from the mean value of the breeding value of the animals born in the same year, therefore the program estimated lower breeding value for the calves born in 1994 owing to the predomination of sires with low breeding value (e.g. the calf 08099/0016/940 whose breeding value estimated on the 205-day weight was -48.7 kg, but their own 205-day weight was 140 kg) and so became the mean value of their breeding value very low.
Summing up the results it can be say that heritability estimated on the additive direct genetic effect (h 2 d =0.37 -0.42) is medium.The maternal value of heritability of the examined traits is low (h 2 m =0.08 -0.12).The correlation between the direct and the maternal genetic effects is negative (rdm=−0.52 -0.74), therefore it is practical to take in account both effects together.
The ratio of the maternal permanent environmental variance (maternal effect of environmental origin) in the phenotypic variance (c 2 ) changed between 0.03 and 0.06.According to the obtained values can it be stated that the maternal permanent environmental effect is at least as important, as the maternal genetic effect.Therefore in the course of examining the weaning performances it is important to take in account the maternal permanent environmental effect, too.
The estimated maternal genetic effect, so the value of the maternal heritability depends on the applied animal model.If the maternal permanent environmental effect isn't taken in account, this effect will appear in the maternal genetic effect and the

Table 2
Composition of the examined population Zusammensetzung der untersuchten Population

Table 3
Overall results of weaned calvesGesamtergebnisse der abgesetzten Kälber

Table 5
Comparison of the models with rank-correlation Vergleich der zwei Modelle aufgrund der Rangkorrelation

Table 6
Estimated breeding value of the tested sires by animal model Der Zuchtwert der Bullen geschätzt mit Einzelmodell