Population structure of Czech cold-blooded breeds of horses

In order to estimate effective population size, generation interval and the development of inbreeding coefficients (Fx) in three original breeds of cold-blooded horses kept in the Czech Republic: Silesian Noriker (SN), Noriker (N) and Czech-Moravian Belgian horse (CMB) all animals of the particular breeds born from 1990 to 2007 were analysed. The average values of generation interval between parents and their offspring were: 8.53 in SN, 8.88 in N and 8.56 in CMB. Average values of effective population size were estimated to be: 86.3 in SN, 162.3 in N and 104.4 in CMB. The average values of inbreeding coefficient were 3.13 % in SN stallions and 3.39 % in SN mares, in the N breed 1.76 % and 1.26 % and in the CMB breed 3.84 % and 3.26 % respectively. Overall averages of Fx were: 3.23 %, 1.51 % and 3.55 % for the breeds SN, N and CMB. The average value of inbreeding coefficient Fx increased by 1.22 % in SN, by 0.35 % in N and by 1.01 % in CMB, respectively. This may lead to a reduction in genetic variability. Reduction in genetic variability could be either controlled in cooperation with corresponding populations of cold-blooded breeds in other European countries or controlled by number of sires used in population.


Introduction
Small populations such as the three breeds of cold-blooded horses in the Czech Republic -Silesian Noriker (SN), Noriker (N) and Czech-Moravian Belgian Horse (CMB) -bear the risk of inbreeding and reduction in genetic diversity with the population.The mating of related animals (inbreeding) was previously aimed at strengthening the required characteristics and traits in the population and at the concentration of appropriate genes in the population.Another aim was to increase offspring uniformity.However, it was observed at the same time that an increase in the inbreeding coefficient caused an increase in mortality, fertility and adaptability of farm animals (Falconer & Mackay 1996).The objective of conservation breeding is to maintain genetic diversity in the population of farm animals.A reduction in genetic diversity is connected with the parameter of effective population size which is used to design the strategy for the conservation of endangered animal species (Bijma et al. 2001).The Food and Agriculture Organization (FAO) classifies breeds into different categories according to the type of endangerment, e.g.whether population size increases, decreases or remains stable.
The populations of cold-blooded horses in the Czech Republic are small sized, which has led to an increase in the population inbreeding coefficient of these breeds, especially in the 1990s when these populations were secluded from the import of genes from other populations.The increase in inbreeding coefficient may bring about undesirable inbreeding depression that is manifested in characteristics associated with fitness and reproduction and in other characteristics related.
The cold-blooded breeds Silesian Noriker, the Noriker and the Czech-Moravian Belgian Horse, along with Old Kladruber Horse and Hucul, belong to the group of endangered horse breeds in the Czech Republic.In the last approx.120 years CMB has been formed in the Bohemian territory on the basis of imports of mainly original Belgian stallions and Walloon stallions to a lesser extent.Breeding goal of CMB is a horse with a medium square frame and earlier maturity.Silesian Noriker has originated in the last approx.100 years from imports of original Noriker stallions and Bavarian cold-blooded stallions.Breeding goal of SN is a horse with a longer frame and later maturity.Noriker breed was formerly used for the formation of SN breed and has been kept in the Bohemian territory for approx.150 years.The two N breeds were further separated geographically to a greater extent.Breeding goal of N is a horse with a medium-sized frame and earlier maturity.
The objective of the present paper was the analysis of development of the inbreeding coefficient, of effective population size and of generation interval in the populations of coldblooded horses in the Czech Republic.

Material and methods
Inbreeding coefficients (F x ) were estimated for all horses regardless of their sex on the basis of complete pedigrees, i.e. pedigrees used comprised all generations reaching back to the founders of the particular breeds -more than 8 generations of ancestors in each horsealtogether 50 845 horses.Horses without known ancestors were included as unrelated animals with zero inbreeding coefficient (F x =0).Data sets were provided by the Horse Breeders Association of the Czech Republic (http://www.aschk.cz).The coefficient F x was estimated as a coancestry coefficient (Falconer & Mackay 1996) using the INBREED procedure of SAS package (SAS 2005): where ƒ is the coancestry coefficient between two horses, e.g.W and Z when A and B represent the parents of horse W while C and D are the parents of horse Z.
The estimated inbreeding coefficients were used to determine the expected change in inbreeding coefficient over a year (ΔF y ) and to estimate the effective population size (N e ) (Falconer & Mackay 1996).
The effective population size (N e ) was estimated from the quotient of a change in inbreeding coefficient per generation (ΔF), calculated as the product of a change in inbreeding coefficient per year (ΔF y ) and the generation interval (L) (Sørensen et al. 2005): Generation interval (L) comprised four real ways of gene transfer: from sire to son (L mm ), from sire to daughter (L mƒ ), from dam to son (L ƒm ) and from dam to daughter (L ƒƒ ) (Nomura et al. 2001): The analysis of the studied populations was done by a comparison of animals born in 1990 and 2007.

Breed development
Tables 1 to 3 show the structure and population size of the particular breeds.Population sizes have been gradually growing since 1993, mainly due to an increase in the number of stallions which has risen from the original 3 to 5 to 34 to 39 stallions.The increasing number of stallions was caused by including more stallions into the breeding scheme.On the other hand, while the numbers of mares remained relatively similar within the period of observation for the breeds SN and CMB, mare numbers have been reduced to a half in the N breed.

Inbreeding
The values of inbreeding coefficient were as follows: F x =0 for 23.68 % of horses, F x values 0<F x <0.0625 for 70.01 % of horses, F x values 0.0625<F x <0.125 for 5.79 % of horses and F x values >0.125 for 0.52 % of horses.The first two groups (F x <0.0625) comprised 93.69 % of all horses.Similar distributions of F x values were shown in other farm animals (Bezdíček et al. 2008, Maiwashe et al. 2008, Biedermann et al. 2009, Köck et al. 2009).
Figure 1 illustrates the development of the mean of F x values according to year of birth in horses of the particular breeds in the period 1990 to 2007.The lowest F x value was determined for all breeds of horses born in 1994 and 1999.A steep increase was observed between years 1994 and 1999.Since 1999, there has been an increase of different intensity in F x in the breeds studied.The general increase in F x values since 1994 was caused by the closing of populations of cold-blooded horses in the Czech Republic in that year against the immigration of genes of the other phylogenetically related horses populations.The peak of F x in the SN breed for horses born in 1995 compared to the other breeds resulted from the low number of horses of this breed which, in addition, also had a higher average F x value.Different values of average inbreeding coefficients among breeds determined in 2007 were caused by different breeding methods used in these breeds.Tables 1 to 3 also document an increase in inbreeding coefficient in the particular breeds during the period of observation (1990)(1991)(1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007).The largest change was found in the CMB breed and the smallest in the N breed.Stallions of the SN breeds had the lowest increase in the F x coefficient and stallions of CMB had the highest similar to the values for the whole population.In mares the lowest change was observed for the CMB breed while the highest change was recorded for the SN breed.As documented by the tables, the increase in the F x coefficient of SN horses born in 1990-2007 was caused by an increase in the average value of F x coefficients in mares whereas in stallions the average F x coefficient did not changed.In the CMB breed the average value of F x in CMB breed increased as a result of its increase in stallions while in mares the average coefficient F x showed a smaller change.In the N breed the increase in F x was similar in both sexes.Similar values of F x were estimated by Jakubec et al. (2009) in Old Kladrub horses and by Wolc & Balińska (2010) in the Polish Konik population.Sierszchulski et al. (2005) found average values of F x =0.88 % in the Polish Arab horse.The rate of F x is of a great importance for the future development of any population.It is considered the decisive measure for estimation of the degree of endangerment within a population.High rates of inbreeding are linked to a progressive loss of genetic variability.The trends of inbreeding found in the present study are highly dependent on the use of individuals of Czech cold-blooded breeds for reproduction.In the year 1990 only 3 to 5 stallions were used for reproduction in the dependence on the breed.As a consequence in the years 2007 already 34 to 55 stallions were used for reproduction in the dependence on the breed.This corresponds with the increase of F x in 1990 and F x stagnation after 2005.

Generation interval
Table 4 shows the values of generation interval (L), expressed as the average age of a horse at the time of birth of its offspring (Falconer & Mackay 1996), over the whole period of observation and over the period from 1990 to 2007, separately for both sexes.Average generation interval was 8.53 years in SN,8.88 in N and 8.56 years in CMB.Generation interval of horses born in 1990-2007 and their sires was 9.47 years in SN, 9.62 years in N and 9.17 years in CMB.The values of dams were 8.37 years in SN, 8.57 in N and 8.44 years in CMB.Tables 4 and 5 document that in horses born before 1990 the value of L was lower by approx.0.5 a year compared to animals born after 1990.CMB breed was the only exception in which a higher value of L was estimated in horses born before 1990.It was caused by mating in all breeds at a younger age by 1990.The birth of offspring was recorded already at 3 years of age in a higher number of horses born before 1990.Table 5 shows that in most cases the interval between the parent and the son is shorter than between the parent and the daughter.SN breed was the exception when the value of L was higher in the segment between sire and son compared to the segment between sire and daughter.Similar values were estimated by Druml et al. (2009) in the Austrian Noriker breed (L=7.9) and by Poncet et al. (2006) in the Franches-Montagnes breed (L=8.4).Slightly higher values were reported by Valera et al. (2005) in Carthusian horses.Compared to the other breeds of farm animals, horses reach a higher generation interval.The longer generation interval is typical for the horse breeding.These higher values of L are caused by older age at birth of the first offspring (4 years in males and females) and by older age at birth of the last offspring (from 19 to 29.6 years in males and from 19 to 23 years in females), depending on the breed.The values determined in this study correspond to the above mentioned data when the youngest age at birth of the offspring was 4 years.Within small populations that primarily have the aim of breed conservation, long generation intervals are advantageous in order to minimize the increase of inbreeding per unit of time.On the other hand, the time span is longer before an increase in inbreeding rate because the low numbers of stallions and breeding mares used can be observed (Hamann & Distl 2008).However, the prolonging of the generation intervals in Czech cold-blooded breeds is not sufficient to compensate for the small number of individuals forming the breed.Nevertheless, markedly higher values of L were determined in English Thoroughbred (L=11.2 for males and L=9.7 for females) (Thiruvenkadan et al. 2009).A shorter generation interval is typical of cold-blooded horses.The main reason is that they need not to undergo long training and athletic performance.The higher value of L in warm-blooded horses is also due to the fact that outstanding sires with a high number of offspring maintain their athletic ability for a longer time and are included in reproduction at older age.A comparison of the values of N e indicates that the analysed breeds of cold-blooded horses belong to the group of endangered populations.Meuwissen & Woolliams (1994) revealed a fundamental relationship between the effective population size and genetic variances of fitness traits.These authors concluded that the critical size for N e , i.e. the size below which the fitness of the population decreases, is between 50 and 100.
The results document an increase in the coefficient of inbreeding that has been observed in original cold-blooded horses in the Czech Republic in the last years.Inbreeding coefficient in the original cold-blooded horses in the Czech Republic can be considered as low.The increase in inbreeding coefficient has been caused by small populations of the particular breeds.Compared with other breeds like the Andalusian, the Lipizzan or the Old Kladruber horse, this level can be considered as low.On the contrary, inbreeding in the Austrian Noriker draught horse is at a same level (1.2 % - Druml et al. 2009).The objective of further breeding work should be a gradual increase in the population size of breeds using a planed mating combination of parents in order to avoid an increase in the relationship between horses and a decrease in genetic variability.Other possibilities could be measures such as common breeding schemes in cooperation with corresponding populations in other European countries or controlled by number of sires used in population.All these measures might also help in a reduction of costs for keeping the numerous small populations of similar genetic background because the taxpayers want to know good reasons when they have to give their money for these populations.

Figure 1
Figure 1 Development of the average value of inbreeding coefficient by the years of birth (▲ -Silesian Noriker, ■ -Noriker and • -Czech-Moravian Belgian horse)

Table 1
Average inbreeding coefficients ofSN within stallions and mares in 1990 and 2007 (Druml et al. 2009)1996)ive population size (N e ) derived from the increase in the average value of inbreeding coefficient per generation(Falconer & Mackay 1996).The lowest value of N e was estimated inSN (43.14)while the highest value of N e was found out in N breed (79.11).CMB breed showed the value N e =52.21.The numbers of founders in the total reference population were 98, 139, 168 (SN, N, CMB respectively), but the effective numbers of founders were only 43.14, 79.11 and 52.21 (SN, N, CMB respectively) animals.Hence, it becomes evident that some founders could be used more intensively than others, which is always critical in terms of loss of genetic diversity.Similar values of N e were determined in breeds of German cold-blooded horses N e =81.4-332.2(Aberle2003)and in Austrian Noriker N e =from 119 to 195(Druml et al. 2009).Jakubec et al. (2006)estimated higher values of N e (140.3) in the Old Kladruber horse.