Janus Pannonius Múzeum Évkönyve 37 (1992) (Pécs, 1993)

Régészet - Finnegan, Michael–Szalay, Ferenc: Population Distance Between Late Roman Period to 11th Century Arpadiam Age Populations as Determined by Non-metric Trait Analysis

101 matrixes for congruence. Elements of the first matrix are plotted against corresponding elements of the second matrix in order to test for distortion in the plot. This has been displayed as a bivariate scatter diagram (see Figure 2). The correlation between these two matrixes was also computed and found to be 0.791 (SOKAL and DERISH 1988). DERISH and SOKAL (1988) suggest that in order to be significant cophenetic correlations should be in the neighborhood of 0.85. While we have not attained that correlation with these matrixes, we still feel the cophenetic correlations are probably in the neighborhood and are therefore quite good. Discussion The distance phenogram (Fig. 1) displays one of a number of possibilities of various hierarchical grouping techniques. This phenogram is directly based on the distance matrix produced in Table 7, and on a pair-wise basis the data presented in this matrix is the analysis of choice. However, to see the interrelationship among these samples, the phenogram was constructed. If we arbitrarily draw a vertical line through the phenogram at 0.0400, we would divide the phenogram into two groups an two independent populations. In the first, more closely associated group, we find the samples of Eilend, an 8-9th century Avar sample, coalescing with Majs, a lOth-llth century Arpadian age population. The Avar population from Kékesd joins very shortly thereafter at a level of 0.0325. A second group, composed of the sample from Székesfehérvár Street and the sample from Nagypall, becomes a group at the level of 0.036. This second group then joins the first group at a level of 0.0428. These two groups, then represent 3-4th century late Roman populations through the Avar period into the Hungarian conquest period ending in about the 11th century. Somewhat extraneous to the above two groups are two populations. First, Zengővárkony, a Neolithic population which joins with the above two groups at the level of 0.0622. Following this, and the last population to join the phenogram, is the late Roman population called István Square, which comes into the phenogram at a level of 0.0807. It is of some interest that the samples from Pécs, István Square and Székesfehérvár Street samples, both aged to the late Roman period, do not join together to make a group. Even using our arbitrary group designation of 0.0400, these two populations would still not join as the pair wise MMD (Table 7) is 0.0420. It might well be that these two populations, which are here seen as significantly different, could have existed at slightly different times, or have been two distinct genetic groups who were alive at the same time an in roughly the same place, with little or no gene flow between them. The difference seen between the sample from Zengővárkony and the remaining samples while great, is expected, in that the Zengővárkony sample is Neolithic whereas the other samples are significantly later in time. The population samples from Kékesd, Majs and Eilend may show some ancestral relationship to each other in a genetic sense and therefore display the lowest MMD value. Still, it must be remembered that these samples are significantly different, based on pair-wise measures of biological divergence. In the above phenogram we have shown the biological relationships among a number of samples of earlier human populations, ranging in age from the late Roman age through the Hungarian conquest and ending in 11th century samples. At the present time the biological distances do not in themselves represent migration patterns which show the affinity among the populations used in this study. However, these results can be compared with similar material from FINNEGAN and MARCSIK (1979, 1989a). We are particularly interested, at a future date, of comparing samples which are of all the same relative age and simply vary by geographic location. Such a study may allow further insight into the interaction of human populations which are separated geographically but not temporally. By doing this, we may better able to identify migrations of earlier human populations or their genes in this Transdanubian region of Hungary. Acknowledgements The authors are indebted to the personnel of the Archeological Museum of Pécs, for their aid, assistance and many kindness, and allowing us to work at odd times, including weekends, during the data collecting phase of this study.