Abstract:

Cranial morphological variability and potential European ancestry factors in human skeletal remains from the Northern Wei Dynasty cemetery at Hengyuan Road were examined in this study. A total of 62 craniums were analyzed, focusing on 18 male and 22 female adult individuals. The research aimed to address methodological limitations in traditional cranial analysis by integrating multivariate statistics and machine learning algorithms, thereby reducing subjectivity in ancestry estimation. The study began with an assessment of within-group variability using Principal Component Analysis (PCA) and biological distance analysis. Specifically, the mean pairwise Euclidean distance (d_m) was calculated to quantify intra- and intergroup variation, revealing that male craniums exhibited significantly lower variability compared to females. This disparity suggests that females may have originated from more diverse genetic or geographic backgrounds, potentially linked to intermarriage or migration practices during the Northern Wei period.

To identify non-Mongoloid individuals, discriminant analysis was conducted using W. W. Howells’ global craniometric database. The results suggested the presence of Europoid traits in some individuals, particularly those resembling the Mediterranean East Type (e.g., M35) and the Central Asian-Two Rivers Type (e.g., M13-2), which are two branches of Europoid. Machine learning algorithms, including random forest (via Ancestrees) and linear discriminant analysis (via Fordisc 3.1), were employed to validate these findings. For instance, M13-2 exhibited a 77.73% probability of Northern / Central European ancestry using Ancestrees’ tournamentForest algorithm, while Fordisc classified M35 as closest to ancient Egyptian and Zalavar groups (probabilities: 34.9% and 31.2%, respectively). These results highlight the advantage of algorithmic approaches in mitigating observer bias inherent in traditional visual assessments.

Biological distance analysis using unrooted neighbor-joining trees further contextualized these individuals within broader Eurasian populations. M35 clustered closely with Mediterranean East Type groups (e.g., Alagou I and Shampula), historically associated with ancient Silk Road populations like the Cheshi and Yutian. Conversely, M13-2 aligned with Central Asian-Two Rivers Type groups (e.g., Jilintai and Zhaosu), suggesting ties to Wusun or Xiongnu-affiliated populations. Notably, the mixed ancestry observed in females like M8-1 and M38 may reflect the multicultural dynamics of Pingcheng (modern Datong), which served as a hub for trade, warfare, and religious exchange during the Northern Wei Dynasty.

The socio-political context of Pingcheng as a cosmopolitan center underpins these findings. Between the 4th and 6th centuries CE, the Northern Wei court relocated millions of people from conquered territories, including Central Asian, Xianbei, and Xiongnu groups. Additionally, the presence of Buddhist artifacts in the Hengyuan Road cemetery aligns with historical records of Sogdian and Persian monks contributing to religious dissemination, possibly introducing Europoid genetic lineages. This multicultural exchange is reflected in the coexistence of European and East Asian cranial traits within the burial population.

Finally, the study underscores the transformative potential of advanced statistical methods in bioarchaeology. By combining PCA, machine learning, and phylogenetic networks, this research provides a replicable framework for analyzing cranial variability and ancestry in fragmented or incomplete datasets. However, limitations such as missing data emphasize the need for cautious interpretation. This work not only enriches our understanding of population diversity in medieval China but also challenges historical narratives by demonstrating that Europoid individuals in the Yellow River Basin were likely isolated cases rather than large-scale migrants, consistent with Han Kangxin’s earlier conclusions. Ultimately, these findings illustrate how cultural assimilation and the Silk Road’s connectivity shaped the biological landscape of ancient Pingcheng, reaffirming the integrative power of Chinese civilization.

Key words: The Northern Wei Dynasty, cranial morphology, Europoid race, biological distance analysis