Référence bibliographique 
Gagnon, Alain et Heyer, Evelyne. 2001. «Fragmentation of the Quebec Population Genetic Pool (Canada): Evidence from the Genetic Contribution of Founders per Region in the 17th and 18th Centuries ». American Journal of Physical Anthropology, vol. 114, no 1, p. 30-41.
« More precisely, our study focuses on the genetic contribution of founders, per region, to a cohort made of spouses married between 1780-1799. » (p. 31)
« The Population Register of Early Québec contains almost every birth, marriage and death certificate (>700,000) recorded in parish registers between 1608-1800. » (p. 31)
Type de traitement des données :
« The 6 million French-Canadians of Quebec derive from a relatively small number of founders. Consequently, some hereditary diseases, which may or may not present a worldwide distribution, have been detected in high frequency in this population. Several studies, however, indicate a nonuniform distribution of these diseases through the population, suggesting that the French-Canadian founder effect has been geographically stratified. Here we explore this stratification by using a demographic database, the Population Register of Early Quebec, that contains almost all birth, marriage, and death certificates (>712,000) recorded in parish registers between 1608-1800. In this database, every genealogical link has been traced back to the founders of the population, so that we can compute the genetic contribution of founder per region, and then account for the early events that have shaped the distribution of diseases. Ten regions, comprising varying numbers of parishes, have been selected. We first describe each region in terms of homogeneity and concentration of its gene pool. For this purpose, a new concept is introduced, the founders’ uniform contribution number (FUN), i.e., the number of founders a population would have if all its founders had an equal contribution. Second, we estimate genetic similarity between regions on the basis of differential genetic contribution. To classify the regions, we use principal component and cluster analysis. Our results show a tripartite clustering of the population, and invite us to reconsider the results obtained from biomolecular and clinical studies, which show a bipartite clustering. » (p. 30)