Livestock Research for Rural Development 21 (11) 2009 Guide for preparation of papers LRRD News

Citation of this paper

Population structure in Albanian sheep breeds analyzed by microsatellite markers

A Hoda*, P Dobi*, G Hyka* and Econogene consortium**

* Department of Animal Production, Agricultural University of Tirana, (AUT), Albania;


Classifications of breeds in Albania are based mainly on phenotypic traits. It is important the determination of population structure for breed conservation. Thirty-one microsatellite markers are used for population structure analysis and for the assignment of the individuals to their reference population. It was carried out a direct assignment of individuals and an exclusion analysis based on 10 000 simulated individuals.


The number of inferred populations was K = 2 for microsatellite markers. Population structure analysis revealed a high level of admixture between populations. Several reasons for the breed admixture are presented. A breeding strategy and policy has to be started in order to conserve valuable sheep breeds.

Key words: breed admixture, breed assignment, local breeds


Local sheep breeds are one of most important genetic resources of livestock species in Albania. There are some local breeds that traditionally are defined mainly by morphological characteristics and production performance. Sheep breed are well adapted to extensive management conditions, graze in the pasture, are resistant to various diseases and parasites. There is a high risk for the extinction of native sheep breeds, because there is a lack of herd book, breeding programs are absent and native breeds are displacing with other breeds according to the preferences of the farmers.  Therefore determination of population structure is very important for breed conservation.


Molecular markers like microsatellites are very useful in assessment of genetic diversity within and between breeds. There are many studies of genetic diversity among sheep breeds, based on microsatellite markers (Arranz et al 1998, Arranz et al 2001, Farid et al 2000, Pariset et al 2003, Peter et al 2007). 


The aim of this study is to assess the genetic structure of Albanian sheep breeds, using 31 microsatellite markers. The laboratory work is carried out in frame of ECONOGENE project.


Materials and methods 

A total of 93 randomly sampled animals representing 3 different Albanian sheep breeds, Bardhoka (Bar), Ruda (Rud) and Shkodrane (Sko) were analyzed. For each breed there were sampled maximum three unrelated individuals (two females and one male) per flock. The individuals were unrelated, older than 2 years and used several times for breeding. Sampling was carried out from an average of 11 flocks per breed. There were no relationships between different herds. The breeds are marginally farmed and autochthonous.


There were used 31 microsatellite markers. The molecular data are provided in the frame of ECONOGENE project


Assignment of individual to their reference population was evaluated using GeneClass (Piry et al 2004).   For all breeds, it was carried out a direct assignment of individuals and an exclusion analysis based on 10 000 simulated individuals. The likelihood method based on allele frequencies (Paetkau et al 1995), as well as Bayesian approach (Rannala and Mountain 1997) were used. By the likelihood method, an individual can be assigned to the population, where the likelihood of its genotype has the highest value. The calculations are carried out, using the “leave one out” procedure. It was considered as correctly assigned, only the individuals with a probability of 99% of assignment to their reference population.


The genetic structure of four sheep breeds is analyzed using the program STRUCTURE (Pritchard et al 2000). The program STRUCTURE uses the Markov Chain Monte Carlo method. The program was run 3 times independently, for K ranging from 2 to 4. All runs were carried out under “admixture model”, with a burn-in period of 300,000 iterations and a period of data collection of 300,000 iterations.


The hierarchical analysis was carried out using analysis of molecular variance (AMOVA) in the arlequin 2000 package (Schneider et al 2000). AMOVA yields estimations of population structure at different levels of the specified hierarchy.


The estimates of distances measured using Proportion of shared alleles, (Dps), are calculated using Population program (Langela 2002). A dendrogram based on UPGMA clustering method was generated from genetic distance matrix using the Phylip package (Felsenstein 1993). The dendrogram was visualized by TreeView program (Page 1996).



There were used two individual assignment/exclusion tests: assignment using a Bayesian-based Approach (Rannala and Mountain 1997), assignment based on allele frequencies (Paetkau et al 1995). The results of assignment of individuals to their reference population, by both methods, under simulation and direct approach, are presented in Table 1.

Table 1.  Correct assignment of individuals to their reference population by microsatellites markers


Number of
























































Both assignment tests gave similar results. Ruda is the breed with the higher rate of correctly assigned animals, using Bayesian method of assignment. Ruda is also the breed with the higher rate of excluded animals (29.03% and 32.26% under frequency based method and Bayes theorem respectively).With the likelihood based method, based on allelic frequencies only 68.82% of the individuals were assigned to their reference population, but based on Bayes theorem, were assigned correctly 65.59% of the individuals. Exclusion tests based on both methods give the same results.


The program STRUCTURE (Pritchard et al 2000) infers the number of populations into which the analyzed genotypes can be divided. The program estimates the natural logarithm of the probability Ln P(D) of given genotype, being part of a given population K. By the Bayes Model-based Clustering Analysis, the highest likelihood value (Ln P(D)), correspond to the most probable number of subpopulations. In order to find the best value of K, the program was run 3 times, fitting K from 2 to 4, (Table 2), under “admixture model”, for both types of markers.

Table 2.   Likelihood values Ln P(D) and its variance Var[LnP(D)]  for K from 2 to 4  



Ln P(D)











The likelihood values and variance are presented in Table 2 and Figure 1 and Figure 2. The best value of  Ln P(D) (-10067.7) was obtained for K = 2, for microsatellite markers. 

Figure 1.  Likelihood values Ln P(D) and its variance Var[LnP(D)]  for K from 2 to 4, for microsatellite markers

Figure 2.   Clustering assignment of the three sheep breeds provided by STRUCTURE analyses

Therefore we infer K= 2 as possible populations (Table 3). It is clear that the inferred populations, for both markers are formed by individuals of all three breed.  

Table 3.  Proportion of membership of Albanian sheep breeds when inferred clusters are 2 microsatellites markers.


Number of individuals

Inferred clusters - Microsatellites















 Each of the 93 animals is represented by a thin vertical line that is divided into two colored segments (K=2), which represent the membership of each individual to the two clusters. Breeds are separated by thin black lines. With K = 2 inferred clusters, it can be seen breed admixture patterns.


The dendrogram based on inter-individual genetic distances is presented in Figure 3.

Figure 3.  Dendrogram of genetic distances between each individual of three sheep breeds

It is clear that the individuals are dispersed among the nodes of different breeds.


From both the tests using Structure and Geneclass, it was also clear that Albanian sheep breeds have a high level of admixture. 

Table 4.  AMOVA analysis of Albanian sheep breeds based on 31 microsatellite markers

Source of variation

Sum of squares

Variance components

Percentage  variation

Among populations




Within populations








AMOVA analysis (Table 4) was carried out to analyze the variation within and between breeds. The AMOVA revealed that percentage of variation among populations was 1.12% and within populations was 98.83%.  



There are several sheep breeds in Albania that for a long time are characterized based on phenotypic traits, production performance, or on blood (Bleta et al 1985) or milk (Zoraqi 1991) protein polymorphism. Recently the genetic variability within and between breeds has been explored using molecular markers. Hoda et al (2009) have characterized the genetic variability of four local sheep breeds, using only 6 microsatellite markers. The breeds presented here have been part of genetic diversity study of 57 European and Middle Eastern sheep breeds (Peter et al 2007). The recent studies have revealed a small genetic distance between breeds and a high level of gene flow.


Therefore it is important to define the population structure, in order to design and establish a proper conservation program for local sheep breeds. The microsatellites markers proved very useful for assessing admixture in Albanian sheep populations. The results revealed using program STRUCTURE and Geneclass suggested a high level of breed admixture.


Assignment methods have multiple applications, like the identification of the source population of a given genotype, evaluation of population differentiation (Waser and Strobeck 1998) or in agriculture for the traceability of animals for breed confirmation (Maudet et al 2002. The analysis of inter-individual distances based on the proportion of shared alleles, made possible the construction of dendrogram that show the genetic relationship among individuals. The dendrogram based on proportion of shared alleles distance revealed frequent cases of animals dispersed among other breed nodes.


Two of these breeds, Bardhoka and Ruda are traditionally transhumant breeds. Their transhumance is oriented on the same region, which are western lowlands. This may bring to intercrosses that occur occasionally or voluntarily by farmers. Albanian farmers have never been organized in breeding associations that may apply breeding standards, identification and keep herd book. Nevertheless that several breeding stations were acting during the last 50 years, sheep herd book was not established and this has facilitated the breed admixture.


In Albania are applied crossing between Bardhoka and Shkodrane, known as Baca, in order to improve the milk production and body weight. All the sheep farms are private and the animal breeding is realized by the farmers himself, who buy the reproducing males in the farm animal market, having as a consequence lack of parentage control The results provided here, may be used to start a breeding strategy and a policy in order to conserve the important sheep breeds.



This work has been supported by the ECONOGENE project, funded by the European Union (project QLK5- CT2001-02461). The content of the publication does not represent necessarily the views of the Commission or its services.



Arranz J J, Bayon Y and San Primitivo F 1998 Genetic relationships among Spanish sheep using microsatellites. Animal Genetics 29: 435-440.


Arranz J J, Bayon Y, and San Primitiovo F 2001 Differentiation among Spanish sheep breeds using microsatellites. Genetics Selection Evolution 33: 529-542


Bleta V, Tartari T and Shteto Th 1985 Disa tė dhėna pėr shpėrndarjen e tipit tė transferinave (Tf) nė delet e racės “Shkodrane”. Buletini i shkencave zooteknike e veterinare 3 (11-13)


Econogene  Sustainable conservation of animal genetic resourses in marginal rural areas: integrating molecular genetics, socio – economics and geostatistical approaches”.


Farid A, O'reilly E, Dollard C and Kelsey C R 2000 Genetic analysis of ten sheep breeds using microsatellite markers. Canadian Journal of Animal Science 80: 9-17


Felsenstein J 1993 PHYLIP: Phylogeny Inference Package. Version 3.5c edition. Seattle, WA, University of Washington .


Hoda A, Dobi P and Hyka G 2009 Genetic diversity and distances of Albanian local sheep breeds using microsatellite markers. Livestock Research for Rural Development. Volume 21, Article #93. Retrieved August 28, 2009, from


Langella O 2002 Populations.


Maudet C, Luikart G and Taberlet P 2002 Genetic diversity and assignment tests among seven French cattle  breeds based on microsatellite DNA analysis. Journal of Animal Science 80: 942–950


Paetkau D, Calvert W, Stirling I and Strobeck C 1995 Microsatellite analysis of population structure in Canadian polar bears. Molecular Ecology 4: 347-354


Page R D M 1996 TREEVIEW: An application to display phylogenetic trees on personal computers ( Computer Applications In The Biosciences 12: 357-358


Pariset L, Savarese M C, Cappucio I and Valentini A 2003 Use of microsatellites for genetic variation and inbreeding analysis in Sarda sheep flocks of central Italy. Journal of Animal Breeding and Genetics 120: 425-432


Peter C, Bruford M, Perez T, Dalamitra S, Hewitt G, Erhardt G and The Econogene Consortium 2007 Genetic diversity and subdivision of 57 European and Middle Eastern sheep breeds. Animal Genetics 38: 37 – 44


Piry S, Alapetite A, Cornuet J M, Paetkau D, Baudouin L and Estoup A 2004 GeneClass2: a software for genetic assignment and first generation migrants detection. Journal of Heredity 95: 536- 539


Prichard J K, Stephens M and Donnelly P 2000 Inference of population structure using multilocus genotype data. Genetics 155: 945–959


Rannala B and Mountain J L 1997 Detecting immigration by using multilocus genotypes. Proceedings of the National Academy of Sciences. USA 98: 9197-9201


Schneider S, Roessli D and Excoffier L 2000 Arlequin ver. 2000: a software for population genetics data analysis. Genetics and Biometry Laboratory, University of Geneva, Switzerland.


Waser P and Strobeck C 1998 Genetic signatures of interpopulation dispersal, Trends in Ecology and Evoution 13: 43-44


Zoraqi G 1991 Studimi i strukturės gjenetike tė racave tė deleve autoktone “Bardhoka” e “Shkodrane” me anė tė markerėve gjenetikė polimorfė dhe lidhjet gjenetike midis racave me anė tė markerėve gjenetikė polimorfė dhe lidhjet gjenetike midis racave e me tiparet prodhuese. Dizertacion pėr marrjen e gradės shkencore “Kandidat i Shkencave

Received 6 September 2009; Accepted 16 September 2009; Published 1 November 2009

Go to top