Phenotypic evolution in prehistoric Ohio Amerindians: natural selection versus random genetic drift in tooth size reduction.

Many anthropologic investigations involve measurement and analysis of polygenic skeletal and dental traits in prehistoric populations from which genetic details cannot be inferred. However, population genetics concepts can be applied productively to analyses of phenotypic variation in prehistoric human populations. One potentially useful approach, derived from basic quantitative genetics (Lande 1976, p. 314), models the effects of natural selection and random genetic drift on the evolution of the average phenotype in a population. We apply this model to the problem of dental size reduction in three prehistoric Amerindian populations from Ohio. Conversion of mean log-transformed buccolingual diameters for six permanent teeth (maxillary and mandibular I1, M1, and M2) to phenotypic standard deviation units reveals significant size reduction in the maxillary teeth only. By assuming 40 generations (t) between the 2 populations and a narrow heritability (h2) range of 0.30-0.70, the estimated minimum selective mortality required to produce the reductions is 1.8 deaths per 100 persons per generation. Given the same t and h2 values, the effective population size (Ne) needed to reject the neutral hypothesis (i.e., random genetic drift) with 95% confidence is approximately 150. Because paleodemographic and ethnographic studies suggest minimum effective sizes of this magnitude for these populations, we tentatively reject random genetic drift and conclude that selective mortality is most probably responsible for the maxillary tooth size reduction observed.