Agricultural and Food Science, vol. 19 (2010): 284-293 A G R I C U L T U R A L A N D F O O D S C I E N C E Vol. 19(2010): 284–293. 284 A G R I C U L T U R A L A N D F O O D S C I E N C E Vol. 19(2010): 284–293. 285 © Agricultural and Food Science Manuscript received August 2009 Microsatellite, blood group and transferrin protein diversity of Estonian dairy cattle breeds Sirje Värv1*, Juha Kantanen2 and Haldja Viinalass1 1Institute of Veterinary Medicine and Animal Sciences, Estonian University of Life Sciences, Kreutzwaldi 1, EE-51014 Tartu, Estonia, *e-mail: sirje.varv@emu.ee 2Biotechnology and Food Research, MTT Agrifood Research Finland, FI-31600 Jokioinen, Finland This study investigates genetic diversity within and among three Estonian dairy cattle breeds (Estonian Native, Estonian Red and Estonian Holstein). A total of 36 markers (25 microsatellites, 10 blood group systems and transferrin protein) were investigated and the within-breed diversity was quantified by expected heterozygosity, number of private alleles and mean allelic richness. The population structure was studied by computing the inbreeding coefficients, breed differentiation and relationships were investigated with random drift-based measures and a factorial correspondence analysis. In addition, a neighbour-joining tree was drawn summarising allele sharing distances for 195 individuals of the Estonian breeds, Western Finncattle, and Danish Jersey. The Estonian breeds displayed generally similar levels of within-population diversity. Depending on the set of markers used 6.2 or 4.3% of the total genetic variation can be explained by differences among the breeds. Construction of the tree for individuals revealed a distinctive pattern of grouping for Estonian Holstein, Estonian Red and Danish Jersey, but Estonian Native and Western Finncattle appeared on the same branches. This indicates that the gene pool of Estonian Native largely overlaps with that of Western Finncattle. However, our genetic marker analysis shows that the three Estonian breeds are genetically differentiated, suggesting that the current gene pool of Estonian dairy cattle is diverse. Key-words: blood group, cattle, genetic diversity, microsatellite, transferrin Introduction Population genetic structures of domestic cattle breeds are greatly influenced by human activities. Different ancestral and demographic histories can generate dissimilar patterns of genetic variation within and among breeds, which can be effectively measured using genetic marker analysis (Ibeagha- Awemu et al. 2004; Li et al. 2007; Mao et al. 2007). Three dairy cattle breeds of different origins and A G R I C U L T U R A L A N D F O O D S C I E N C E Vol. 19 (2009): 284–293. 284 A G R I C U L T U R A L A N D F O O D S C I E N C E Vol. 19(2010): 284–293. 285 census sizes are currently raised in Estonia. Estonian Native Cattle are of old Estonian origin and were officially recognised in 1914 when the herd book for the breed was established. The present population (500 breeding females) has been influenced by the use of Western Finncattle A.I. bulls during the 20th Century. Since 1995 a conservation programme for Estonian Native Cattle has been implemented, including the collection of embryos for an ex situ gene bank for the breed. The Estonian Holstein Breed, which comprises 75% of the dairy cattle population (98 500 dairy cows in 2009) in Estonia, and the Estonian Red (21 000 breeding females) are the main dairy cattle breeds in Estonia. Their herd books were established in 1885. The Estonian Holstein originally descended from Dutch Friesian cattle, with a marked introduction of international Holstein semen since the late 1970s. The Estonian Red Cattle contain genes sourced from a broad Angeln and Danish Red base. Estonian breeders, being incorporated with the European Red Dairy Breed Association, are using genetic material focal to all European red cattle breeds. Tapio et al. (2006) analysed genetic diversity using microsatellite markers in 35 North European cattle breeds, including Estonian Native and Es- tonian Red. Their study showed that the Estonian Native Cattle share common ancestries with the Finnish and Scandinavian native breeds, while the Estonian Red belongs to the group of Baltic red breeds. In addition, Tapio et al. (2006) demonstrated that the Estonian Native Cattle are of a high value in the conservation of cattle genetic resources when the prioritisation of cattle breeds is based simultane- ously on within- and between-breed components of genetic diversity. However, that study did not include the Estonian Holstein Cattle. Here we study for the first time the molecular genetic diversity of this breed and increase the number of microsatellites used to analyse all Es- tonian dairy cattle breeds. In addition, we analyse erythrocyte antigen (EA) systems (or blood groups) and blood protein in the Estonian dairy breeds. Our attempt is to apply genotypes in the blood group systems to make more diverse use of them by typing also the parents and/or offspring of the studied in- dividuals. In studies where genetic data have come from typings of one generation, simplification and modification of the mode of inheritance are typical- ly needed (Blott et al. 1998; Kantanen et al. 1999). In the blood-group systems EAA, EAB, EAC and EAS, the antigenic factors form complexes that are inherited as haplotypic blocks. Also a recessive al- lele segregates in these complex blood groups and in bi-allelic systems EAJ, EAL, EAM, and EAZ. Previously, Arranz et al. (1996), Moazami- Goudarzi et al. (1997) and Kantanen et al. (2000) compared microsatellites and biochemical markers in cattle, and e.g. Barker et al. (1997), Luís et al. (2007) and Tapio et al. (2003) studied polymor- phisms in these two types of markers in water buf- faloes, horses and sheep, respectively. These two marker types typically give congruent results for population divergence (Arranz et al. 1996; Luís et al. 2007). The aim of the present study is to analyse genetic diversity and differentiation of the three Es- tonian dairy cattle breeds by comparing genotypic and allelic data of microsatellites and blood groups - protein markers. In addition, we investigate the di- vergence of Estonian Native Breed from the breeds (Western Finncattle, Jersey and the two other Esto- nian breeds) that have had genetic influence on the gene pool of the breed in order to identify animals with the most pure Estonian origins. Materials and Methods Sampling We collected blood samples from 40 Estonian Na- tive, 40 Estonian Red and 34 Estonian Holstein cattle. The sampled individuals originated from 14, 7 and 17 farms located in distinct regions of Estonia. Animals were pre-selected using pedigree data kept by the Estonian Animal Recording Center. For each animal three generations were considered to avoid sampling closely related animals. Sires of Estonian Native and Estonian Red animals were of Estonian origin and therefore our samples represent characteristic present-day types of these breeds. For Estonian Holstein, in turn, it was difficult A G R I C U L T U R A L A N D F O O D S C I E N C E Värv, S. et al. Diversity among the Estonian dairy breeds 286 A G R I C U L T U R A L A N D F O O D S C I E N C E Vol. 19(2010): 284–293. 287 to find animals sired by the old type of Estonian Black-and-White Cattle and the animals included in the present study were descended from modern international Holstein bulls. DNA extraction and marker analysis For this study, DNA from the blood samples of Esto- nian Holstein Cattle was extracted using a Genomic DNA Purification Kit (MBI Fermentas, Lithuania). For the Estonian Native and Estonian Red Cattle, DNA samples for the typing of microsatellites were available from a previous study (Tapio et al. 2006). The individuals were analysed for 25 microsat- ellites, 10 blood group loci, and the transferrin pro- tein locus. Data for the following 20 microsatellites were available for the Estonian Native and Estonian Red breeds (Tapio et al. 2006): BM1818, BM1824, BM2113, CSSM66, ETH003, ETH010, ETH152, ETH225, HEL005, HEL01, HEL09, HEL13, IL- STS005, ILSTS006, INRA005, INRA023, INRA032, INRA035, INRA037, and INRA063. In addition to these microsatellites, we typed a further five micros- atellites: TGLA053, TGLA122, TGLA126, TGLA227, and SPS115. Moreover, these additional markers are recommended by the International Society for Ani- mal Genetics (ISAG) and the Food and Agriculture Organization of the United Nations (FAO) for cat- tle genetic diversity studies. The Estonian Holstein Cattle, which were not included in the previous mi- crosatellite study by Tapio et al. (2006), were typed for all 25 microsatellites. We carried out the PCR reactions to amplify microsatellite loci in a 25 μl reaction mixture including 7.5 – 20 pmol of each primer, 200 μM of each dNTP, DynaZyme™- buffer (Finnzymes, Finland), 50 ng of DNA template, and 1 U of DynaZyme™II DNA polymerase (Finnzymes). Annealing temperatures in the PCR for different mi- crosatellites ranged from 55 to 58 ºC and the ampli- fied products were separated on a MegaBACE™ 500 DNA Sequencer (Amersham Biosciences, UK). The consistency in size of microsatellite alleles was as- sured by comparison with control samples available from the study of Tapio et al. (2006). A total of 60 erythrocyte antigenic (EA) factors for the 10 systems were typed. These were: A1, and A2 in the blood group system A (EAA); B1, B2, G1, G2, G3, I1, I2, K, O1, O2, P1, P2, Q, T1, T2, Y2, A’, B’, D’, E’2, F’1, F’2, G’, I’1, J’2, K’, O’2, P’, Q’, Y’, B’’, G’’ (EAB); C1, C2, E, R1, R2, W, X1, X2, C’, L’ (EAC); F and V (EAF); J (EAJ); L (EAL); M (EAM); S, U1, U2, H’, U’, H’’, U’’ (EAS); Z (EAZ); S’ and R’ (EAR’). In the antigenic factor detection, the internationally accepted haemolysis test using monospecific reagents was used, the suitability of which was examined by biannual comparison tests organised by the ISAG in 1993 – 2004. For the blood group systems EAF and EAR’, codo- minance operates, while for other EA systems a recessive allele segregates. In addition, genes con- trolling the determination of erythrocyte antigens for multifactor systems are closely linked and are inherited as haplotypic complexes that determine the phenotypic appearance of several antigenic factors or a single antigen. Determination of the antigen complexes (considered here as alleles) in the genotypes of EAA, EAB, EAC, and EAS was carried out using family analysis. Thus, parents of all the 114 individuals were typed in our study to determine their genotypes. Horizontal polyacrylamide gel electrophoresis was used to separate transferrin (TF) alleles (A, D1, D2, E), as described by Juneja and Gahne (1987). Estonian Native and Estonian Red were typed also for amylase 1 (AMY1), amylase 2 (AMY2) and ceruloplasmin (CP) proteins (unpublished data) using starch gel electrophoresis (Smithies 1955). These proteins were not typed for Estonian Hol- stein. We used these additional data only in the calculation of the within-population inbreeding coefficient (see results). Statistical analysis In the statistical analysis, the microsatellite and biochemical marker (the 10 blood group systems and transferrin, denoted here and henceforth as EA systems/TF) data sets were examined separately. Locus-wise deviations from Hardy-Weinberg equi- A G R I C U L T U R A L A N D F O O D S C I E N C E Värv, S. et al. Diversity among the Estonian dairy breeds 286 A G R I C U L T U R A L A N D F O O D S C I E N C E Vol. 19(2010): 284–293. 287 librium (HWE) and pair-wise linkage disequilibrium (LD) between loci within each breed were computed using GENEPOP v.3.4 (Raymond and Rousset 1995a) with the following parameters of the Markov Chain Method: dememorization = 10 000, batches = 1 000 and iterations = 10 000. In the breed-wise LD tests, the frequency of significant results (p < 0.05) and the significance of pooled p-values of the exact tests using Fisher’s method for combining probabilities (Raymond and Rousset 1995b) was reported for mi- crosatellites and EA systems/TF separately. Basic di- versity indices, i.e. the unbiased estimates of expected heterozygosity, the number of private alleles and the allelic richness, were calculated and the calculation of the allelic richness was based on 22 (microsatel- lite data) and 15 individuals (EA systems/TF). The within-population diversity estimates were derived using FSTAT v.2.9.3 (Goudet 2001). This program was also used to compute within-population inbreeding coefficients (f) (Weir and Cockerham 1984). Genetic differentiation was computed using the variance based method (θ) of Weir and Cock- erham (1984) in FSTAT v. 2.9.3.2 (Goudet 2001). The significance of θ-estimates was determined with 5 000 permutations. Moreover, the pattern of population differentiation was described by a facto- rial correspondence analysis of the individual mul- tilocus scores using GENETIX4.05 (www.genetix. univ-montp2.fr/genetix/genetix.htm). The popula- tion clusters derived from the factorial correspond- ence analysis are identified graphically (Lebart et al. 1984). The first two major components were plotted on a scatter diagram for the three cattle breeds. In ad- dition, Chord genetic distances (Cavalli-Sforza and Edwards 1967) between the breeds were computed using GENETIX4.05. We conducted an additional genetic differentia- tion analysis by calculating the allele sharing distanc- es (Bowcock et al. 1994) between 195 individuals of the three Estonian breeds and Danish Jersey and Western Finncattle using the data for 19 microsatel- lites (INRA035 was excluded, see results). The data for Western Finncattle and Danish Jersey, the breeds which the Estonian Native Cattle Breed Society has used for upgrading of the Estonian Native, were ob- tained from the study of Tapio et al. (2006). Based on the allele sharing distance matrix, a neighbour-join- ing tree was constructed using SplitsTree4 V4.11.3 software (Huson and Bryant 2006). Results Markers All markers were polymorphic across the breeds (Table 1). A total of 209 microsatellite alleles and 122 blood group and transferrin alleles were detected. The number of microsatellite alleles per single locus ranged from 2 (ILSTS005) to 15 (TGLA053 and TGLA122), and that of EA systems/TF alleles from 2 (EAJ, EAL, EAM, EAF, EAR’, and EAZ) to 59 (EAB). The average expected heterozygos- ity for the microsatellite loci was 0.70, and for the biochemical markers 0.41. EAB and EAC displayed higher levels of expected heterozygosity than any microsatellite marker (Table 1). In the blood group systems, the genotyping was not totally successful due to discrimination difficul- ties between probable homozygotes for a dominant allele and heterozygous genotypes for blood groups where a recessive allele was segregating. The EAA and EAS were the most difficult markers to deter- mine an individual’s genotype from the antigenic phenotypes, with an overall genotyping success of 76 and 53%, respectively. In Estonian Native, the genotyping success at dominant marker loci ranged from 55 (EAS) to 90% (EAL), in Estonian Red from 58 (EAC) to 93% (EAJ) and in Estonian Holstein from 45 (EAS) to 92% (EAB). The most complex locus, EAB, was genotyped for 80 and 85% of Estonian Red and Estonian Native Cattle individuals, respectively. Nine of a total of 75 (12%) independent tests for Hardy-Weinberg equilibrium (HWE) at the mi- crosatellite loci were rejected at p < 0.05. When results of the microsatellite loci were pooled across the breeds, INRA035 showed significant (p < 0.05, adjusted with a Bonferroni correction) deviation from HWE. This marker showed also a high posi- tive f value (Weir and Cockerham 1984; Table 1). The Mendelian inheritance of microsatellite alleles A G R I C U L T U R A L A N D F O O D S C I E N C E Värv, S. et al. Diversity among the Estonian dairy breeds 288 A G R I C U L T U R A L A N D F O O D S C I E N C E Vol. 19(2010): 284–293. 289 was not investigated in this study, but we assume that the deficiency of heterozygotes at INRA035 was due to the presence of non-amplifying null alleles and therefore we excluded INRA035 from further analysis. In the data set of EA systems/TF, one deviation (transferrin in the Estonian Native) from HWE (3% of the independent tests) was recorded (p < 0.05). When the results were pooled and a Bonferroni correction applied to adjust the significance levels, none of the biochemical markers showed deviation from HWE. Genetic diversity and population structure of the breeds The within-population genetic diversity and popula- tion structure estimates are given in Table 2. The breeds showed similar levels of within-population diversity in terms of expected heterozygosity and allelic richness on the basis of the microsatellite data. For the EA systems/TF data, however, Estonian Holstein displayed lower within-population diversity than the two other breeds. Our data sets indicated that the number of private alleles was highest in Estonian Red (20 microsatellite alleles over 24 loci and 28 of EAB, EAC and EAS alleles totally). The within-breed population structure was in- vestigated by computing linkage disequilibrium es- timates and inbreeding coefficients for the breeds. For the Estonian Red and Estonian Holstein, the frequency for linkage disequilibrium was less than 5%, while in Estonian Native this frequency was slightly more than expected by chance (microsatel- lite data). No linkage disequilibrium was detected in the EA system/TF data and pooled p -values from locus-by-locus pair-wise comparison did not indicate any significant deviations from linkage equilibrium proportions. Within-population in- breeding estimates (f) based on microsatellite data did not deviate significantly from zero. However, the negative f-estimate obtained from the EA sys- tems/TF analysis suggested the influence of out- breeding in Estonian Native Cattle (95% CI for f [-0.163, -0.086]). Table 1. Microsatellites and Erythrocyte antigen sys- tems/Transferrin protein analysed in the present study, number of alleles (Na) detected, Nei’s gene diversity (H) and f estimates calculated according to Weir and Cockerham (1984). Marker Na H f BM1818 8 0.651 0.033 BM1824 5 0.751 0.036 BM2113 8 0.814 0.053 CSSM66 9 0.816 -0.022 ETH003 7 0.791 0.016 ETH010 10 0.810 -0.026 ETH152 8 0.743 0.055 ETH225 9 0.867 0.052 HEL001 9 0.720 0.068 HEL005 9 0.740 0.085 HEL009 12 0.731 0.004 HEL013 7 0.669 -0.152 ILSTS005 2 0.574 0.066 ILSTS006 9 0.797 0.030 INRA005 4 0.575 -0.255 INRA023 9 0.803 -0.043 INRA032 5 0.659 0.045 INRA035 6 0.579 0.255 INRA037 11 0.719 -0.009 INRA063 6 0.646 -0.198 SPS115 6 0.682 0.022 TGLA053 15 0.860 0.006 TGLA122 15 0.803 0.005 TGLA126 8 0.730 0.025 TGLA227 12 0.870 -0.023 EAA 3 0.328 -0.178 EAB 59 0.967 -0.020 EAC 39 0.941 -0.060 EAF 2 0.322 -0.162 EAJ 2 0.213 -0.124 EAL 2 0.092 -0.068 EAM 2 0.036 -0.021 EAR' 2 0.280 -0.103 EAS 5 0.441 0.016 EAZ 2 0.194 -0.107 TF 4 0.680 0.027 A G R I C U L T U R A L A N D F O O D S C I E N C E Värv, S. et al. Diversity among the Estonian dairy breeds 288 A G R I C U L T U R A L A N D F O O D S C I E N C E Vol. 19(2010): 284–293. 289 We calculated within-population inbreeding co- efficients for Estonian Native and Estonian Red, including three additional codominantly inherited blood protein loci (AMY1, AMY2 and CP) in the data set. Data on these proteins for the Estonian Holstein are not available. We obtained -0.050 and 0.004 for f-estimates, with 95% confidence inter- vals [-0.134, 0.062] and [-0.054, 0.054], respec- tively, suggesting that our estimate for the Estonian Native presented in Table 2 was not robust. Genetic differentiation The overall θ estimate (Weir and Cockerham 1984) for the microsatellite data was 0.062 (95% CI [0.045, 0.080]) and for the EA systems/TF data, 0.043 (95% CI [0.021, 0.067]). All pair-wise θ comparisons were statistically significantly different from zero (p <0.05) when computed from the microsatellite data (Table 3). For the EA systems/TF data, the respective pair-wise θ estimates were: 0.051 (p < 0.05), 0.040 (p = 0.06) and 0.043 (p < 0.05). In the factorial correspondence analysis of the microsatellite data, the first two principal compo- nents (PCs) explained 61.7 and 38.3% of the total variation, and in the analysis of the EA system/TF data, 53.7 and 46.3%, respectively (Fig. 1a and 1b). The two-dimensional plot constructed from the mi- crosatellite data indicated discrete grouping of the three Estonian cattle breeds with only two Estonian Native and one Estonian Red animals not being as- signed to their source population. Based on the EA systems/TF data the demarcation within Estonian Red, Estonian Native and Estonian Holstein clusters was lower than on the microsatellite-based plot, but still discriminated the breeds. However, individuals with intermediate component scores indicated the probable outbred origin of these animals. Table 2. Within-population diversity values and population structure derived from the microsatellite loci and EA systems/ TF data. Mean expected unbiased heterozygosity (Hexp), allelic richness (R), number of private alleles (A), the frequency of significant (P < 0.05) pair-wise linkage disequilibrium test (LD%), the pooled exact P-values in the LD-tests (χ2) and within-population inbreeding coefficient (f) with 95% confidence intervals (95% CI) are shown. Breed Microsatellite data EA systems/TF data Hexp R A LD% χ 2 f (95% CI) Hexp R A LD% χ 2 f (95% CI) Est. Native Est. Red Est. Holstein 0.715 0.699 0.694 6.01 5.97 5.87 16 20 15 5.4 4.7 3.3 573.1 NS 511.6 NS 481.2 NS -0.017 [-0.065, 0.003] 0.026 [-0.022, 0.048] -0.016 [-0.076, 0.009] 0.404 0.405 0.361 4.94 5.14 4.53 26 28 16 0 0 0 63.9 NS 37.9 NS 49.8 NS -0.107 [-0.163, -0.086] -0.010 [-0.083, 0.031] -0.034 [-0.119, 0.009] The significance of pooled P-values of the exact tests in LD analysis using Fisher’s method: NS Not Significant Table 3. Pair-wise θ and chord distances between the breeds based on microsatellite (given above the diagonal) and EA systems/TF (below the diagonal) Breed Chord distance θ Estonian Native Estonian Red Estonian Holstein 1 2 3 1 0.061 0.046 2 0.058 0.046 3 0.095 0.075 1 0.051 0.040 2 0.052 0.043 3 0.078 0.059 A G R I C U L T U R A L A N D F O O D S C I E N C E Värv, S. et al. Diversity among the Estonian dairy breeds 290 A G R I C U L T U R A L A N D F O O D S C I E N C E Vol. 19(2010): 284–293. 291 A) B) Figure1. Plotted representation of three breed clusters as defined by Analysis of Factorial Correspondence: analysis based on A) microsatellites; B) EA systems/TF data. Estonian Native Estonian Red Estonian Holstein Estonian Red Estonian Holstein Estonian Native A) B) Figure1. Plotted representation of three breed clusters as defined by Analysis of Factorial Correspondence: analysis based on A) microsatellites; B) EA systems/TF data. Estonian Native Estonian Red Estonian Holstein Estonian Red Estonian Holstein Estonian Native Fig. 1. Plotted representation of three breed clusters as defined by Analysis of Factorial Correspondence: analysis based on a) microsatellites; b) EA systems/TF data. The analysis of the genetic distances based on the proportion of shared alleles between indi- viduals (Fig. 2) confirmed the close relationship of Estonian Native with Western Finncattle. The grouping of 195 cattle revealed a large mixed group of individuals from the Estonian Native and West- ern Finncattle and distinct branches of Estonian Red, Estonian Holstein and Danish Jersey cattle confirming the discrete grouping of the Estonian breeds found in the factorial correspondence analy- sis. Discussion As suggested by our genetic marker analyses, the Estonian Native, Estonian Red and Estonian Holstein breeds are genetically divergent populations among which the gene flow appears to be restricted. The divergence indicates that the present-day gene pool of Estonian dairy cattle is diverse. The θ estimates (Weir and Cockerham 1984) showed that the Es- tonian cattle breeds are significantly differentiated and the factorial correspondence analysis (Fig. 1) and allele sharing distances (Fig. 2) confirm the grouping of individuals graphically according to their breed origin. However, as seen in Figure 2 the present Estonian Native cattle population forms an overlapped gene pool with Western Finncattle and it was not possible to determine a special Estonian Native group among the analysed individuals. This finding diminishes the conservation value of the Estonian Native cattle among the North Euro- pean cattle breeds (Tapio et al. 2006) in terms of genetic uniqueness, but despite this the breed can be considered as an important gene reservoir for agro-biodiversity in the Estonian context. As shown by the microsatellite data, 6.2% of the total genetic variation of the Estonian dairy cattle can be explained by differences among the breeds. The level of genetic differentiation among European cattle breeds has been slightly higher, around 10% (MacHugh et al. 1998; Kantanen et al. 2000), than the present estimate, but estimates have typically been based on a larger set of breeds, from a wider geographic region, than covered in the present study. The current subdivision of the Estonian dairy breeds at the microsatellite loci is comparable to the extent of genetic differentiation among 18 French, Spanish, and Portuguese cattle breeds (a proportion of 7% among the breeds ac- cording to Cañón et al. 2001). Our conclusion re- a) b) A G R I C U L T U R A L A N D F O O D S C I E N C E Värv, S. et al. Diversity among the Estonian dairy breeds 290 A G R I C U L T U R A L A N D F O O D S C I E N C E Vol. 19(2010): 284–293. 291 garding their genetic divergence, is in agreement with the previous study by Tapio et al. (2006), which showed that the North European native, Red and Holstein-Friesian breeds form discrete breed groups. The EA/TF data indicated a lower level of genetic differentiation among the breeds (4.3% of the total genetic variation) than the microsatel- lites. This could have been due to the lower num- ber of alleles found at the biochemical markers, which may increase the probability that alleles are identical by state but not identical by descent, but partly also due to the lower number of EA sys- tem/TF markers analysed in the present study. The standardised genetic differentiation measure G’ST presented by Hedrick (2005) allows a more ap- propriate comparison between loci with different mutation rates. We obtained overall G’ST values of 0.20 for microsatellites (20% of the maximum possible) and 0.06 for EA system/transferrin (6% of the maximum possible). Our data indicate that microsatellites are more valuable markers for di- agnostics in breed differentiation and individual assignment analysis of dairy cattle. In addition to microsatellites, the two highly polymorphic blood group systems, EAB and EAC, were found to dis- criminate cattle breeds more efficiently than other blood groups and even microsatellite loci (G’ST 0.80 at EAB and 0.54 at EAC). These loci add valu- able information for breed differentiation studies including for the private alleles found (e.g. alleles of the EAB Y2D’G’, B1G2KA’, B1G2KE’F’2 and I2G’Q’ in the Estonian Native breed and BP’ and O2QJ’K’O’ in the Estonian Red breed). Although recent demographic histories of the Estonian dairy cattle breeds differ considerably, the breeds in general show a similar degree of intra- breed genetic variation (Table 2). The estimates for the Estonian breeds at the microsatellite and at the EA systems/TF markers are comparable with those presented for other European cattle breeds (Kan- tanen et al. 2000; Li et al. 2007; Tapio et al. 2006). The molecular diversity of the Estonian Holstein breed was measured for the first time in the present study and we found that this effectively selected breed shows similar levels of within-population variation as e.g. Finnish, Russian and French Hol- stein or Black-and-White cattle populations (Li et Fig. 2. Neighbour-joining tree showing relationships between cattle of five breeds (Estonian N a t i v e , We s t e r n F i n n c a t t l e , Estonian Red, Estonian Holstein, and Danish Jersey) constructed by SplitsTree4 using the allele sharing distances (Bowcock et al. 1994). The individuals of Estonian Native and Western Finncattle located at the same branches on the figure are marked with large circles and small squares, respectively. A G R I C U L T U R A L A N D F O O D S C I E N C E Värv, S. et al. Diversity among the Estonian dairy breeds 292 A G R I C U L T U R A L A N D F O O D S C I E N C E Vol. 19(2010): 284–293. 293 al. 2007; Maudet et al. 2002; Tapio et al. 2006; 19 or 20 common microsatellites with our study). The present f-estimates calculated from the mic- rosatellite data (Table 2) do not suggest effects of inbreeding or outbreeding in the Estonian breeds. However, the biochemical marker data point to- wards outbreeding in Estonian Native Cattle. When additional protein loci were included in the analysis, the outbreeding was less apparent. We speculate that the test statistics did not reveal an outbreeding effect in the Estonian breeds (although they have been influenced by other breeds) because genetically closely related breeds have been used for crossing. The present microsatellite and biochemical marker data gave inconsistent results on breed re- lationships (Table 3; Fig. 1). The Estonian breeds did not show a fragmented population structure (Table 2), which could have been one source of discrepancy, as reported by Tapio et al. (2003). The microsatellite data may provide more reliable results on between-breed diversity compared with documented breed histories (Kantanen et al. 2000; Rendo et al. 2004; Wiener et al. 2004; Tapio et al. 2006). As pointed out by Bowcock et al. (1994), markers with a large number of alleles typically show less biased estimates than those based on low-polymorphic markers. On the other hand, we typed more microsatellite markers than biochemi- cal markers, which increases the reliability of analysis (Takezaki and Nei 1996). Our genetic marker analysis indicated that the Estonian dairy cattle gene pool is variable. The breeds have diverged as reflected also by the rela- tively high number of private alleles detected in each breed. However, the future trends may threat- en this diversity in the Estonian dairy cattle gene pool. For example, the old type of Estonian Red studied here is rapidly disappearing due to cross- breeding with Red Holstein and with other Euro- pean red cattle. We conclude that the DNA samples and genotyping data collected in the present study will be of value in future studies examining, for example, temporal changes in the genetic diversity of Estonian dairy cattle breeds. Acknowledgements. This work was supported by the Nordic Council of Ministers and the Nordic Genetic Resource Center (NordGen), targeted financing of research project SF1080022s07 and applied research of the Ministry of Agriculture of Estonia. References Arranz, J. J., Bayón, Y. and San Primitivo, F. 1996. Compar- ison of protein markers and microsatellites in differentia- tion of cattle populations. Animal Genetics 27: 415–419. Barker, J.S.F., Moore, S.S., Hetzel, D.J. S., Evans, D., Tan, S.G. and Byrne, K. 1997. Genetic diversity of Asian wa- ter buffalo (Bubalus bubalis): microsatellite variation and a comparison with protein-coding loci. Animal Ge- netics 28: 103–115. Blott, S.C., Williams J.L. and Haley, C.S. 1998. Genetic re- lationship among European cattle breeds. Animal Ge- netics 29: 273–282. Bowcock, A., M., Ruiz-Linares, A., Tomfohrde, J., Minch, E., Kidd, J.R. and Cavalli-Sforza, L.L. 1994. High reso- lution of human evolutionary trees with polymorphic mi- crosatellites. Nature 368: 455–457. Cañón, J., Alexandrino, P., Bessa, I., Carleos, C., Carrete- ro, J., Dunner, S., Ferran, N., Garcia, D., Jordana, J., Laloė, D., Pereira, A., Sanchez, A. and Moazami-Gou- darzi, K. 2001. Genetic diversity measures of local Eu- ropean beef cattle breeds for conservation purposes. Genetics Selection Evolution 33: 311–332. Cavalli-Sforza, L.L. and Edwards, A.W.F. 1967. Phyloge- netic analysis: models and estimation procedures. Amer- ican Journal of Human Genetics 19: 233–257. Goudet, J. 2001. FSTAT, a program to estimate and test gene diversities and fixation indices (version 2.9.3). Available from http://www.unil.ch/izea/softwares/fstat. html. Hedrick, P.W. 2005. A standardized genetic differentiation measure. Evolution 59: 1633-1638. Huson, D.H. and Bryant D. 2006. Application of Phyloge- netic Networks in Evolutionary Studies. Molecular Biol- ogy and Evolution 23: 254–267. Software available from www.splitstree.org. Ibeagha-Awemu, E. M., Jäger, S. and Erchardt, G. 2004. Polymorphisms in blood proteins of Bos indicus and Bos taurus cattle breeds of Cameroon and Nigeria, and de- scription of new albumin variants. Biochemical Genet- ics 42: 181–197. Juneja, R.K. and Gahne, B. 1987. Simultaneous pheno- typing of pig plasma α-proteaseinhibitors (PI1, PO1A, PO1B, PI2) and four other proteins (PO2, TF, CP, HPX) by a simple method of 2D horizontal electrophoresis. Animal Genetics 18: 197–211. Kantanen, J., Olsaker, I., Adalsteinsson, S., Sandberg, K., Eythorsdottir, E., Pirhonen, K. and Holm, L.-E. 1999. Temporal changes in genetic variation of Northern Eu- ropean cattle breeds. Animal Genetics 30: 16–27. A G R I C U L T U R A L A N D F O O D S C I E N C E Värv, S. et al. Diversity among the Estonian dairy breeds 292 A G R I C U L T U R A L A N D F O O D S C I E N C E Vol. 19(2010): 284–293. 293 Kantanen, J., Olsaker, I., Holm, L.-E., Lien, S., Vilkki, J., Brusgaard, K., Eythorsdottir, E., Danell, B. and Adalsteis- son, S. 2000. Genetic diversity and population structure of 20 North European cattle breeds. Journal of Hered- ity 91: 446–457. Lebart, L., Morineau, A. and Warwick, K.M. 1984. Multi- variate Descriptive Statistical analysis: Correspondence Analysis and Related Techniques for Large Matrices. John Wiley ansd Sons, New York. Li, M.H., Tapio, I., Vilkki, J., Ivanova, Z., Kiselyova, T., Marzanov, N., Ćinkulov, M., Stonajović, S., Ammosov, I., Popov, R. and Kantanen, J. 2007. The genetic struc- ture of cattle population (Bos taurus) in northern Eura- sia and the neighbouring Near Eastern regions: impli- cations for breeding strategies and conservation. Mo- lecular Ecology 16: 3839–3853. Luís, C., Juras, R., Oom, M.M. and Cothran, E.G. 2007. Genetic diversity and relationship of Portugese and oth- er horse breeds based on protein and microsatellite loci variation. Animal Genetics 38: 20–27. MacHugh, D.E., Loftus, R.T., Cunningham, P. and Bradley, D.G. 1998. Genetic structure of seven European cattle breeds assessed using 20 microsatellite markers. Ani- mal Genetics 29: 333–340. Mao, Y., Chang, H., Yang, Z., Zhang, L., Xu, M., Sun, W., Chang, G and Song, G. 2007. Genetic structure and dif- ferentiation of three Chinese indigenous cattle popula- tions. Biochemical Genetics 45: 195–209. Maudet, C., Luikart, G., and Taberlet, P. 2002. Genetic di- versity and assignment tests among seven French cat- tle breeds based on microsatellite DNA analysis. Jour- nal of Animal Science. 80: 942–950. Moazami-Goudarzi, K., Laloë, D., Furet, J.P. and Gro- sclaude, F. 1997. Analysis of genetic relationships be- tween 10 cattle breeds with 17 microsatellites. Animal Genetics 28: 338–345. Raymond, M. and Rousset, F. 1995a. Genepop (Version 3.1d, updated version of Genepop v. 1.2): population ge- netics software for exact tests and ecumenicism. Jour- nal of Heredity 86: 248–249. Raymond, M. and Rousset, F. 1995b. An exact test for pop- ulation differentiation. Evoution 49: 1280–1283. Rendo, F., Iriondo, M., Jugo, B.M., Aguirre, A., Mazón, L.I., Vicario, A., Gómez, M. and Estonba, A. 2004. Analysis of the genetic structure of endangered bovine breeds from the western Pyrenees using DNA microsatellite markers. Biochemical Genetics 42: 99–108. Smithies, O. 1955. Zone electrophoresis in starch gels: Group variation in the serum proteins of normal human adults. Biochemical Journal 61: 629. Takezaki, N. and Nei, M. 1996. Genetic distances and re- construction of phylogenetic trees from microsatellite DNA. Genetics 144: 389–399. Tapio, I., Värv, S., Bennewitz, J., Maleviciute, J., Fimland, E., Grislis, Z., Meuwissen, T.H.E., Miceikiene, I., Olsa- ker, I., Viinalass, H., Vilkki, J. and Kantanen, J. 2006. Prioritization for conservation of Northern European cat- tle breeds based on analysis of microsatellite data. Con- servation Biology 20: 1768–1779. Tapio, M., Miceikiene, I., Vilkki, J. and Kantanen, J. 2003. Comparison of microsatellite and blood protein diversi- ty in sheep: inconsistencies in fragmented breeds. Mo- lecular Ecology 12: 2045–2056. Weir, B.C. and Cockerham, C.C. 1984. Estimating F-sta- tistics for the analysis of population structure. Evolution 38: 1358–1370. Wiener, P., Burton, D. and Williams, J.L. 2004. Breed rela- tionships and definition in British cattle: a genetic analy- sis. Heredity 93: 597–602. Microsatellite, blood group and transferrin protein diversity of Estonian dairy cattle breeds Introduction Materials and Methods Sampling DNA extraction and marker analysis Statistical analysis Results Markers Genetic diversity and population structure of the breeds Genetic differentiation Discussion References