Temporal variation in the mating structure of Sanday, Orkney Islands.

Pedigree and vital statistics data from the population of Sanday, Orkney Islands, Scotland, were used to assess temporal changes in population structure. Secular trends in patterns of mate choice were analysed for three separate birth cohorts of spouses: 1855-1884, 1885-1924 and 1925-1964. The degree to which mating was random or assortative with respect to both genealogical and geographic distance was determined by comparing average characteristics of all potential mates of married males with those of actual wives. We integrated this procedure, originally developed by Dyke (1971), into a three-fold investigation of population structure: (1) comparison of random and non-random components of relatedness as measured from pedigree data; (2) an analysis of marital distance distributions for actual and potential mates of married males; and (3) the relationship between genealogical relatedness and geographic distance. As population size decreased from 1881 to the present, total kinship and spatial distances between spouses increased. Whereas the random component of relatedness increased over time, consanguinity avoidance was sufficient to decrease the total coefficient of kinship over time. Part of the increase in consanguinity avoidance was associated with isolate breakdown, as distances between island-born spouses, as well as the total amount of off-island migration, increased from the mid-nineteenth century to the present. Mate choice was influenced by geographic distance for all time periods, although this effect diminished over time. Since decreases in population size, concomitant with increases in consanguinity avoidance and community exogamy, have probably occurred quite frequently in small human populations, as well as in rural Western communities in the past century, observed secular trends illustrate the potential for change in population structure characteristic of isolate breakdown.