In admixed populations, genetic contributions from males and females of specific parental populations can be of different proportions due to past directional mating during the process of genetic admixture. In this research paper, we provide evidence of such male- and female-specific differential admixture components of African, European, and American Indian origin in an admixed population from the city of Melo, in the northeastern region of Uruguay. From data on 11 autosomal markers from a sample of 41 individuals of mixed African descent, we estimated 47% African, 38% European, and 15% Amerindian contributions. In contrast, 6 mtDNA site-specific polymorphic markers showed that the mtDNA genome of these individuals was 52% African, 19% European, and 29% Amerindian, while from 3 Y-specific polymorphic sites, we estimated 30% African, 64% European, and 6% Amerindian contributions. We argue that this heterogeneity of admixture estimates results from disproportionate unions of European males with African and American Indian females from which this mixed African population was formed. Also, we argue that the asymmetry of the admixture estimates from the three sets of markers (autosomal, mtDNA, and Y-linked) is a result of the changes in the direction of mating during the history of the population. Implications of such evidence of directional mating are discussed, indicating the need of further demographic data for a quantitative assessment of the impact of directional mating on genetic structure of admixed populations.

Recent studies of the Uruguayan population revealed different amounts of Amerindian and African genetic contributions. Our previous analysis of Afro-Uruguayans from the capital city of the Department of Cerro Largo showed a high proportion of African genes, and the effects of directional mating involving Amerindian women. In this paper, we extended the analysis to a sample of more than 100 individuals representing a random sample of the population of the whole Department. Based on 18 autosomal markers and one X-linked marker, we estimated 82% European, 8% Amerindian, and 10% African contributions to their ancestry, while from seven mitochondrial DNA site-specific polymorphic markers and sequences of hypervariable segment I, we determined 49% European, 30% Amerindian, and 21% African maternal contributions. Directional matings between Amerindian women and European men were detected, but differences involving Africans were not significant. Data about the specific origins of maternal lineages were also provided, and placed in a historical context.

Results concerning HLA types and 22 other blood genetic systems are reported for the Parakanã Indians of northern Brazil, a tribe that is notable for the light color and pilosity of some of its members. No clear evidence of Caucasoid admixture was found, but the Parakanã show unusual frequencies in the EsD1, PGM1(1) Gc2, CpB, Fya, Dia, and LM genetic markers. In addition, the very rare Rh allele ry is present, as well as what seems to be a new PGM2 variant. There is very limited heterogeneity in the HLA system. All these distinctive features may have arisen through a combination of founder effects and genetic drift. However, low FIS values, as well as higher mean ages in heterozygous as compared to homozygous persons, suggest that an heterotic effect is counteracting these dispersive forces.

A total of 495 individuals from five different Argentinian tribes was examined for variation in 23 blood group and protein genetic systems, and the results were integrated with previous data on some of these systems. These tribes generally present RH * R1, PGM1 * 1, and ACP * A frequencies lower and RH * R2, ESD * 1, and GLO * 1 prevalences higher than those observed in other South American Indian groups. Earlier studies with mitochondrial DNA showed that haplogroup A was present in low frequencies in these tribes, but haplogroup B showed a high prevalence among the Mataco. Average heterozygosities are very similar in the five tribes, while estimates of non-Indian ancestry are generally low. Both the blood group and protein, as well as the mtDNA data sets, divide the five tribes into two groups, and the relationships obtained with the blood group and protein systems are exactly those expected on the basis of geography and language. However, the topology obtained with the mtDNA results was different, possibly due to sampling effects or diverse patterns of exchange between the groups related to sex.