| Abstract: | This analysis examines the association between genetic heterozygosity and individual morphologic variation in a captive population of Papio hamadryas hamadryas consisting of 403 juveniles and adults. The population structure of the colony was artificially generated and maintained and is thus rigorously defined. Subpopulations delimited by age, sex, and degree of inbreeding are also explored. Heterozygosity, as enumerated from six simple Mendelian biochemical loci, is compared with the residual morphologic variation of each individual for each of 20 quantitative traits. Use of a sequential Bonferroni technique nullifies all significant correlations. Principal-components analysis reduces the morphometrics to a single or few significant axes in each population. The first axis of the total population contains 86.07% of the variation in the sample and the absolute values of the factor scores exhibit a significant positive correlation with heterozygosity at P < 0.05. Correcting for age- and sex-related variation in the total population with a linear model subsequently demonstrates that no significant correlation between heterozygosity and morphologic variation exists. No significant relationship is found in the inbred animals or subpopulations when age and sex are controlled. Previous studies have indicated that individuals proximal to the population mean for a specific polygenic trait exhibit a higher biochemical heterozygosity than individuals distant from the mean. The results presented here, which are based on more loci than many studies and a well-defined population, do not support this relationship. Substructuring of a population by age and sex can lead to spurious correlations with univariate or multivariate techniques. Comprehensive indices of genetic variation and rigorous statistical techniques should be used in future analyses. Studies that fail to recognize these design elements should be interpreted with caution. |
