ИСТИНА

Significance.The question of congruence between genetic and craniometric biodistances in humans is essential for reconstructing relationships between ancient populations or individuals using morphometric data. First studies of this issue revealed moderate to high and significant correlations between the two types of distances at the global level. Geographical distances are often used as a proxy for genetic distances and show moderate or high correlations with craniometric distances, again at the global level. But later research has shown that those correlations can vary substantially depending on a number of factors: cranial samples and polymorphisms used, part of the skull studied and, importantly, on the scale of comparison (i.e. global, continental, regional or in- tracemetery).The role of climate as an alternative factor shaping the human face has been assessed in numerous studies also showing the fluctuating nature of the association between cranial morphology and climate.Material and methods.We employ 35 male cra- nial samples from North Eurasia (mid-facial measurements), corresponding mtDNA and Y-chromosome samples (from the same populations, not the same individuals), matrices of geographical and climatic distances between the populations in order to sys- tematically assess the influence of the scale of comparison on the association between different interpopulation distance matrices: continental vs. sub-continental vs. regional. Results. At the continental level, when the populations frombothWest and East Eura- sia were analyzed together,mid-facial craniometric distances were equally strongly correlated withmtDNA (0.67), Y-chromosome (0.60) and geographical (0.65) distances while amuch weaker association with climatic distances (0.41) was observed.These values are very similar to those published previously for global cranial samples. When sub-continental datasets, i.e. West (20 samples) and East (15 samples) Eurasians, were analyzed separately, correlations with genetic and geographical distances dropped dramat- ically: mtDNA – -0.01 and 0.22; Y-chromosome – 0.42 and 0.45; geographical – 0.32 and 0.20 (Europe and Asia, respectively). Simultaneously, climatic distances became more strongly associated with morphological ones: 0.46 in Europe and 0.53 in Asia. This latter finding supportsmany previous studies reporting higher association between facialmorphology and climate at the con- tinental scale compared to the global scale. Finally, we considered three regional subsets of samples: North Asia (10 populations, all were demonstrated to be cold-adapted), Northeast Europe (10 populations, also all cold-adapted) andWest and South Europe (10 populations, adapted to temperate or Mediterranean climates). At this scale, all correlations between craniometric distances and either genetic or geographical distances remain low or become even lower than at the sub-continental scale (range from -0.17 to 0.38). One exception is mtDNA distances in North Asia (0.54). But the association with climate is low as well, ranging from 0.05 in Northeast Europe to 0.30 in North Asia. Thus at the regional scale, when populations from the same climatic zone are studied, mid-facial craniometric biodistances between the populations remain largely unexplained by either their neutral genetic relationships or climate. Discussion. Certainly, the main explanation of the results of this study seems to be the sampling strategy, wheremorphological and genetic data are obtained fromdifferent individuals.The smaller are the differences between populations (both morphological and genetic), the more important is the effect of sampling. But in our opinion, other important factors are the genetic basis of craniofacial variation and epigenetic effects, internal rather than external, affecting the path from genotype to craniofacial phenotype