Alu insertion polymorphisms in the Balkans and the origins of the Aromuns
Wednesday, 7. October 2009, 06:35:15
We have analysed 11 human-specific Alu insertion polymorphisms in the Balkans to elucidate the origins of the Aromuns, a linguistic isolate inhabiting scattered areas in the Balkan Peninsula.
Four Aromun samples (two from the Republic of Macedonia, one from Albania, and one from Romania) and five neighbouring populations (Macedonians, Albanians, Romanians, Greeks, and Turks) were analysed by means of genetic distances, principal components and analyses of the molecular variance (AMOVA).
Three hypotheses were tested: Aromuns are Romanophonic Greeks; the result of a Romanian southward migration; or local descendants of the Thracians.
The analyses show that the Aromuns do not constitute a homogeneous group separated from the rest of the Balkan populations.
Grouping by language or geography does not explain the genetic differences observed in the region, suggesting a lack of genetic structure in the area. Aromuns do not seem to be particularly related to Greeks, Romanians, or to other Romance speakers.
The Aromuns might have their origin to the south of the Danube river, with extensive gene flow with the neighbouring populations.
The present results suggest a common ancestry of all Balkan populations, including Aromuns, with a lack of correlation between genetic differentiation and language or ethnicity, stressing that no major migration barriers have existed in the making of the complex Balkan human puzzle.
.......
Scenario 1: Aromuns are Latinised Greeks
The origin of the Aromuns could be traced to a group of Greeks who changed their language and adopted Latin after the arrival of Romans into the Balkans. This assumption would imply that Aromuns should be genetically more similar to Greeks than to the rest of neighbouring populations. In the present analysis, Northeastern Greeks have been shown to be geneti- cally differentiated from the rest of the Balkan popula- tions (see also Huckenbeck et al. 2001). This agrees with the results shown by classical genetic markers in other Greek samples (Cavalli-Sforza et al. 1994; Schmidt & Scheil, 2001). The Aromun samples differ from Greeks in the PC analysis and also presented considerable ge- netic distances from them. Moreover, the genetic dis- tances found between Greeks and each Aromun sample are not significantly shorter than the distances found, for instance, between Aromuns and Romanians.
Within this scenario, a possible explanation for this difference found between Greeks and Aromuns would be that the origins of Aromuns could be traced back to Greeks who extensively interbred with Romans. This hypothesis would imply that the Aromuns present intermediate genetic features between Greeks and Southern and Western Romance populations. Since Alu allele frequencies are not available for Italians, Western Romance populations (represented in the present analysis by French, Catalans and Andalusians) can be used as proxies.
The genetic position of Aromuns was not intermediate between Greeks and the Western Romance speakers, which, with the caveat that the latter are not necessarily descendants of Romans, would be incompatible with Aromuns being the result of an admixture between Romans and Greeks.
Thus, the present data seems to refute this first hypothesis of the Aromuns being Latinised Greeks.
Scenario 2: Aromuns and Romanians are descendants of Dacians
This second scenario postulates that the Aromuns and Romanians have the same origin. Both may be descendants of ancient inhabitants living north of the Danube and mixed with Romans. This scenario would imply a close genetic relationship between Aromuns and Romanians. The present analysis shows that there is no close relation between the samples north of the Danube river (Romanians and Hungarians) and the Aromun samples, which agrees again with the results shown by classical genetic markers (Schmidt et al. 1998).
Even Romanian Aromuns are very differentiated from Romanians, as shown by genetic distances and PC analysis.
This could be explained by the fact that after 1920 many Aromuns coming from the south of the Balkans emigrated to Romania, where they settled in the Dobruja region and near Bucharest. This second hypothesis for the same origin of the Aromuns and Romanians, as descendants of the inhabitants north of the Danube and mixed with Romans, can also be rejected.
Scenario 3: Aromuns are Descendant of Thracians
The last hypothesis tested is the origin of the Aromuns as descendants of Thracians, who inhabited the region south of the Danube, and afterwards mixed with Romans. This scenario would be difficult to test since no extant population can be identified with the ancient Thracians. Nevertheless, this hypothesis would imply
a close genetic relationship between Aromuns and the rest of the populations living in the Balkans south of the Danube. The present analysis has shown a close relationship between Balkan populations and Aromuns,
although Aromuns of any given geographical region or political country do not present closer genetic distance to their neighbours, as seen in the AMOVA analysis.
The genetic similarity of Aromuns to other Balkan populations could be compatible with the present scenario, but
it is worth noting that Balkan populations do not differ distinctly from the other European populations. This is in agreement with the high homogeneity of European populations as shown by classical markers (Cavalli-Sforza et al. 1994) and mitochodrial DNA lineages (Simoni et al. 2000).
Nonetheless, a single origin for the Aromuns cannot be confirmed by the present data, given that the genetic heterogeneity found among Aromun samples is similar to that found among a set of linguistically unrelated Balkan populations. This pattern could be explained if Aromun populations had, in fact, different origins scattered in the Balkans, and had converged linguistically by adopting and preserving a Romance language.
Or, alternatively, genetic drift could have erased the traces of a putative common origin. Alu insertion polymorphisms represent multiple independent autosomal loci, and the general trends observed in them escape the random specificities of each single locus, a feature that makes them useful in population genetics. However, they lack the well-defined phylogeographic structure of variation at the mtDNA or the Y chromosome, which may provide more precise details to the intriguing question of the origin of the Aromuns.
Received: 18 July 2003
Accepted: 1 October 2003
(C) University College London 2004
Annals of Human Genetics (2004) 68,120–127
Alu insertion polymorphisms in the Balkans and the origins of the Aromuns
Four Aromun samples (two from the Republic of Macedonia, one from Albania, and one from Romania) and five neighbouring populations (Macedonians, Albanians, Romanians, Greeks, and Turks) were analysed by means of genetic distances, principal components and analyses of the molecular variance (AMOVA).
Three hypotheses were tested: Aromuns are Romanophonic Greeks; the result of a Romanian southward migration; or local descendants of the Thracians.
The analyses show that the Aromuns do not constitute a homogeneous group separated from the rest of the Balkan populations.
Grouping by language or geography does not explain the genetic differences observed in the region, suggesting a lack of genetic structure in the area. Aromuns do not seem to be particularly related to Greeks, Romanians, or to other Romance speakers.
The Aromuns might have their origin to the south of the Danube river, with extensive gene flow with the neighbouring populations.
The present results suggest a common ancestry of all Balkan populations, including Aromuns, with a lack of correlation between genetic differentiation and language or ethnicity, stressing that no major migration barriers have existed in the making of the complex Balkan human puzzle.
.......
Scenario 1: Aromuns are Latinised Greeks
The origin of the Aromuns could be traced to a group of Greeks who changed their language and adopted Latin after the arrival of Romans into the Balkans. This assumption would imply that Aromuns should be genetically more similar to Greeks than to the rest of neighbouring populations. In the present analysis, Northeastern Greeks have been shown to be geneti- cally differentiated from the rest of the Balkan popula- tions (see also Huckenbeck et al. 2001). This agrees with the results shown by classical genetic markers in other Greek samples (Cavalli-Sforza et al. 1994; Schmidt & Scheil, 2001). The Aromun samples differ from Greeks in the PC analysis and also presented considerable ge- netic distances from them. Moreover, the genetic dis- tances found between Greeks and each Aromun sample are not significantly shorter than the distances found, for instance, between Aromuns and Romanians.
Within this scenario, a possible explanation for this difference found between Greeks and Aromuns would be that the origins of Aromuns could be traced back to Greeks who extensively interbred with Romans. This hypothesis would imply that the Aromuns present intermediate genetic features between Greeks and Southern and Western Romance populations. Since Alu allele frequencies are not available for Italians, Western Romance populations (represented in the present analysis by French, Catalans and Andalusians) can be used as proxies.
The genetic position of Aromuns was not intermediate between Greeks and the Western Romance speakers, which, with the caveat that the latter are not necessarily descendants of Romans, would be incompatible with Aromuns being the result of an admixture between Romans and Greeks.
Thus, the present data seems to refute this first hypothesis of the Aromuns being Latinised Greeks.
Scenario 2: Aromuns and Romanians are descendants of Dacians
This second scenario postulates that the Aromuns and Romanians have the same origin. Both may be descendants of ancient inhabitants living north of the Danube and mixed with Romans. This scenario would imply a close genetic relationship between Aromuns and Romanians. The present analysis shows that there is no close relation between the samples north of the Danube river (Romanians and Hungarians) and the Aromun samples, which agrees again with the results shown by classical genetic markers (Schmidt et al. 1998).
Even Romanian Aromuns are very differentiated from Romanians, as shown by genetic distances and PC analysis.
This could be explained by the fact that after 1920 many Aromuns coming from the south of the Balkans emigrated to Romania, where they settled in the Dobruja region and near Bucharest. This second hypothesis for the same origin of the Aromuns and Romanians, as descendants of the inhabitants north of the Danube and mixed with Romans, can also be rejected.
Scenario 3: Aromuns are Descendant of Thracians
The last hypothesis tested is the origin of the Aromuns as descendants of Thracians, who inhabited the region south of the Danube, and afterwards mixed with Romans. This scenario would be difficult to test since no extant population can be identified with the ancient Thracians. Nevertheless, this hypothesis would imply
a close genetic relationship between Aromuns and the rest of the populations living in the Balkans south of the Danube. The present analysis has shown a close relationship between Balkan populations and Aromuns,
although Aromuns of any given geographical region or political country do not present closer genetic distance to their neighbours, as seen in the AMOVA analysis.
The genetic similarity of Aromuns to other Balkan populations could be compatible with the present scenario, but
it is worth noting that Balkan populations do not differ distinctly from the other European populations. This is in agreement with the high homogeneity of European populations as shown by classical markers (Cavalli-Sforza et al. 1994) and mitochodrial DNA lineages (Simoni et al. 2000).
Nonetheless, a single origin for the Aromuns cannot be confirmed by the present data, given that the genetic heterogeneity found among Aromun samples is similar to that found among a set of linguistically unrelated Balkan populations. This pattern could be explained if Aromun populations had, in fact, different origins scattered in the Balkans, and had converged linguistically by adopting and preserving a Romance language.
Or, alternatively, genetic drift could have erased the traces of a putative common origin. Alu insertion polymorphisms represent multiple independent autosomal loci, and the general trends observed in them escape the random specificities of each single locus, a feature that makes them useful in population genetics. However, they lack the well-defined phylogeographic structure of variation at the mtDNA or the Y chromosome, which may provide more precise details to the intriguing question of the origin of the Aromuns.
Received: 18 July 2003
Accepted: 1 October 2003
(C) University College London 2004
Annals of Human Genetics (2004) 68,120–127
Alu insertion polymorphisms in the Balkans and the origins of the Aromuns
WIDGETIZE