1Anthropology, University of Cincinnati, 2Anthropology, University of Michigan, 3Anthropology, University of Toronto, Mississauga
March 28, 2019 , CC Room 26 C
In order to better characterize pigmentation variation and associated genetic loci in admixed populations, we report on quantitative pigmentation phenotype data from three population samples (African Americans, n=259, US Hispanics, n=90, and Mexicans from the city of Palenque, n=101). All three samples are significantly different (F=215.5, df=2, p<2x10-16) from each other in skin Melanin Index (M). A Tukey post-hoc test indicates that this is not explained by the higher M of the African American sample alone, as the US Hispanic and Palenque samples are also significantly different (p<1x10-10). Hair M also differs significantly among the three populations (F=41.96, df=2, p <2x10-16), driven by the lower values of the US Hispanic sample, which is significantly lighter than the Palenque and African American samples (p<0.00001). We use high-resolution iris images to quantitatively assess iris pigmentation in a subset of the African American (n=147) and US Hispanic (n=48) samples. MANOVA shows that the two groups are significantly different (F=20.882, p<2.2x10-16) from each other in CIELab color space. While CIELab values are generally lower in the African American sample (indicating darker irises), they are more variable in the US Hispanic sample, consistent with previous reports. These results suggest that considerable variation in skin, hair, and iris pigmentation phenotype exists between populations, and that using social categories (e.g. “Hispanic” or “African American”) as a stand-in for phenotypic variation may mask important within-population variation. Finally, we use dense genotype data and admixture mapping methods to test for genotype-phenotype associations, conditioning on variation in individual ancestry.
Data collection was supported by a grant to HLN and EJP by the US Department of Justice (2013-DN-BX-K011).