Construction of worldwide database for craniofacial and dental morphology

基于来自世界各地共约18000个个体的牙齿和颅骨形态学数据，分析了16个地理类群的区域间和区域内变异模式。Fst估算的区域间差异在总方差的12.7-18.9%范围内（最小Fst=10.4%）。这一变异范围与遗传数据得到的结果是一致的。撒哈拉以南非洲人在16个主要地理群体中表现出最大的区域内多样性。我测试了撒哈拉以南非洲人的多样性是基于时间的函数，还是基于人口参数，如与其他地理群体相比，使用r矩阵方法时更大的人口规模。否定了不同地理区域长期有效人口规模相同的零假设。撒哈拉以南非洲序列的相对有效人口规模为8.3或更高，则残差（观测方差与预期方差之间的差异）变得微不足道。在排除人口规模的潜在影响后，撒哈拉以南非洲人与其他群体的距离越来越远。如果我们接受分支模型，那么从区域间和区域内变异分析中获得的结果与最近关于解剖学上现代人类起源的非洲模型是一致的。然而，采用区域间长期迁移模型，目前的结果并不一定与多区域模型不相容。强调指出，不同类型的数据应反映现代人口区域间和区域内变化的相似模式。

The patterns of inter- and intra-regional variation among 16 geographical groups were analyzed based on dental and cranial morphological data (totally ca.18,000 individuals from all around the world). The inter-regional variation estimated by Fst falls within the range of 12.7-18.9% (minimum Fst=10.4%) of the total variance. This range of variation is compatible with those obtained by genetic data. Subsaharan Africans show the largest intra-regional diversity among 16 major geographic groups. I test whether the diversity of subsaharan Africans is based on a function of time or on demographic parameters such as larger population size when compared with other geographic groups by applying R-matrix method. A null hypothesis that the long-term effective population sizes of different geographic regions were the same was rejected. Weighting a relative effective population size of 8.3 or more for subsaharan African series, the residual variance (the difference between observed and expected variance) becomes insignificant. After removing the potential effect of the population size, subsaharan Africans become more removed from other groups. If we accept the branching model, the results obtained from the analyses of both inter- and intra-regional variation are in agreement with the recent African model for the origin of anatomically modern humans. Taking a model for long-term migration among regions, however, the present results are not necessarily incompatible with multiregional model. It is stressed that different types of data should reflect similar patterns of inter- and intra-regional variations of modern human populations.