Franceschi M G, Paoli G
Dipartimento di Scienze del Comportamento Animale e dell'Uomo, Università di Pisa, Italy.
Hum Biol. 1994 Oct;66(5):905-16.
We investigate the influence of within-parish similarity level on between-parish similarity pattern in surname analysis through a study of the relative roles of various isolation factors accounting for the population structure within a municipality (21 parishes) in the hilly part of northern Tuscany. One surname per family was collected (N = 3052), and the pattern of kinship observed by isonymy was compared with the expected patterns based on demogeographic variables. The fit of the isolation by distance model is substantially improved with the inclusion of values at zero geographic distance (i = j). Differentiation in migration rate was clearly shown by the local kinship values, which are highly correlated with the demogeographic features of the parishes under study. Furthermore, the strong heterogeneity of the local kinship values seems to be the major factor responsible for the poor relation of the off-diagonal values of the kinship matrix to the expected isolation pattern. In fact, the kinship matrix weighted on the basis of the local kinship values points to a pattern of similarity among parishes congruent with the demogeographic structure of the population. On the whole, the results indicate that at the low hierarchical level of parishes the extreme localization of surnames is a disturbing factor in population structure analysis. A simple method to overcome this problem has been proposed.
我们通过研究托斯卡纳北部山区一个市镇（21个教区）内各种隔离因素对人口结构的相对作用，来调查教区内部相似性水平对姓氏分析中教区之间相似性模式的影响。每个家庭收集一个姓氏（N = 3052），并将通过同姓法观察到的亲属关系模式与基于人口地理变量的预期模式进行比较。通过纳入零地理距离（i = j）处的值，距离隔离模型的拟合得到了显著改善。当地亲属关系值清楚地显示了迁移率的差异，这些值与所研究教区的人口地理特征高度相关。此外，当地亲属关系值的强烈异质性似乎是亲属关系矩阵非对角线值与预期隔离模式关系不佳的主要因素。事实上，根据当地亲属关系值加权的亲属关系矩阵指向了一种教区之间的相似性模式，与人口的人口地理结构一致。总体而言，结果表明，在教区的低层次水平上，姓氏的极端本地化是人口结构分析中的一个干扰因素。本文提出了一种克服这个问题的简单方法。
