Mitochondrial DNA variation and language replacements in the Caucasus

Ivane Nasidze, Mark Stoneking

Abstract

Sequences of the first hypervariable segment of the mitochondrial DNA (mtDNA) control region were obtained from 353 individuals representing nine groups and four major linguistic families (Indo–European, Altaic and North and South Caucasian) of the Caucasus region. The diversity within and between Caucasus populations exceeded the diversity within Europe, but was less than that in the Near East. Caucasus populations occupy an intermediate position between European and Near Eastern populations in tree and principal coordinate analyses, suggesting that they are either ancestral to European populations or derived via admixture from European and Near Eastern populations. The genetic relationships among Caucasus populations reflect geographical rather than linguistic relationships. In particular, the Indo–European–speaking Armenians and Altaic–speaking Azerbaijanians are most closely related to their nearest geographical neighbours in the Caucasus, not their linguistic neighbours (i.e. other Indo–European or Altaic populations). The mtDNA evidence thus suggests that the Armenian and Azerbaijanian languages represent instances of language replacement that had little impact on the mtDNA gene pool.