Plos Genetics issue of March'08 is out and with it an already advanced interesting research: Geographic patterns of genome admixture in Latin American Mestizos, by Sijia Wang et al.
The article has been recieved with sort of a "nothing new" comment all around the genetic blogosphere but I think a couple of issues are of interest.
The "nothing new" non-finding is that Mestizos have overall more European than Native male ancestry, specially when compared with the female ancestry. In this study they researched autosomal SNP markers and X chromosome ones. The first is statistically inherited equally from male and female ancestors, while the second is more female than male (roughly 67% and 33%). In most Mestizo samples, the European ancestry happens to be significatively larger in the SNP bar than in the autosomal one (see fig. 2), meaning more male than female European ancestry. the inverse is true for all samples regarding Black African ancestry (i.e. their Black ancestors were systematically mostly females).
But there are at least two exceptions to the European male dominance "rule": in the samples of Oriente (Guatemala) and Salta (Argentine, near Bolivia) the European ancestry bars are about the same, what actually translates as Native American male ancestry larger than the female one, and European female ancestry unusually large. The same is true for the samples of Pasto and Peque (both in rural Colombia), though maybe not as exageratedly. So there are 4/14 where the standard machista Eurocentric mestizaje was modified up to inversion. Why? Don't know yet but a clue may be in the rural nature of these mestizo samples, as well as in the fact that they are among the 5 samples that have highest Native American blood overall. Probably European immigration was low in those areas and therefore Mestizo women (rather than purely European, who were always few) may have no choice but marrying natives (I mean it was, and surely still is, socially less favorable - I naturally do not oppose that personally, not at all: society does).
The other issue is the makeup of Native ancestry in these Mestizo groups. And, against what the authors argue, the data (fig. 5 and 6) shows that some populations actually have an extremely variegated non-local native ancestry. The most striking examples are also probably the most urban ones: Mexico City and Medellín.
Mexico City Mestizos have among their native ancestry, some 60% that does not belong to North America (North- and Meso-American linguistic groups) but to Isthmic and South America. The Andean and Equatorial-Tucanoan groups, that are definitively not Central American nor Caribbean, make up a 40% of their native ancestry. A very similar case is that of Medellín Mestizos, where the North American blood makes up almost 40% of their native ancestry. Much of the same can be said of other samples: Rio Grande do Sul (Brazil), Tucuman and Cajamarca (Argentine), Cundinamarca (Colombia, the department that includes Bogotá) and Oriente (Guatemala).
I wonder why and the easiest answer is that probably Latin American Mestizos travelled quite a lot in the Spanish Empire - after all they constituted the backbone of the Spanish administration and military, even if the ruling elite was of nearly pure European ancestry and even often new arrivals from the metropolis.