I've posted a fair amount about the new field of historical population genetics. Some of the most popular mass-market books in genetics deal with this field, for example Spencer Wells' Journey of Man. On the other hand, there's a lot of sloppy overreach on the part of some practitioners, especially due to the excessive reliance on uniparental lineages; the unbroken female and male lineages (mtDNA and NRY). Nevertheless, in specific narrow cases where hypotheses are being tested they can be very illuminating. For example, here is a question: do the mixed-race populations of the Caribbean exhibit any evidence of descent from the indigenous pre-Columbian populations? This is an open question because it was in the Caribbean that the first and most extreme die-offs of native populations occurred when exposed to Eurasian pathogens. The short answer seems to be yes, some indigenous ancestry does persist into the present. This was first confirmed in Puerto Rico, Reconstructing the population history of Puerto Rico by means of mtDNA phylogeographic analysis:
The haplogroup identities of 800 mtDNAs randomly and systematically selected to be representative of the population of Puerto Rico were determined by restriction fragment length polymorphism (RFLP), revealing maternal ancestries in this highly mixed population of 61.3% Amerindian, 27.2% sub-Saharan African, and 11.5% West Eurasian. West Eurasian frequencies were low in all 28 municipalities sampled, and displayed no geographic patterns. Thus, a statistically significant negative correlation was observed between the Amerindian and African frequencies of the municipalities. In addition, a statistically highly significant geographic pattern was observed for Amerindian and African mtDNAs. In a scenario in which Amerindian mtDNAs prevailed on either side of longitude 66°16 West, Amerindian mtDNAs were more frequent west of longitude 66°16 West than east of it, and the opposite was true for African mtDNAs. Haplogroup A had the highest frequency among Amerindian samples (52.4%), suggesting its predominance among the native Taínos....
Now Dienekes reports on Cuba, Genetic origin, admixture, and asymmetry in maternal and paternal human lineages in Cuba:
...In order to investigate the demographic processes that gave rise to the current Cuban population, we analyzed the hypervariable region I (HVS-I) and five single nucleotide polymorphisms (SNPs) in the mitochondrial DNA (mtDNA) coding region in 245 individuals, and 40 Y-chromosome SNPs in 132 male individuals...The Native American contribution to present-day Cubans accounted for 33% of the maternal lineages, whereas Africa and Eurasia contributed 45% and 22% of the lineages, respectively. This Native American substrate in Cuba cannot be traced back to a single origin within the American continent, as previously suggested by ancient DNA analyses. Strikingly, no Native American lineages were found for the Y-chromosome, for which the Eurasian and African contributions were around 80% and 20%, respectively.
The traditional model for the Caribbean is pretty straightforward, the natives die, and African slaves and European settlers produce a mixed population to various degrees. There is of course a noticeable sex-bias here due to the surplus of European males. Because of the incredibly high mortality rates any sex imbalance within the slave population was less relevant than skew in those who might have a chance to reproduce, in particular women who entered into relationships with free white males and so might be able to secure more resources for their offspring (slave mortality rates in the Caribbean were far higher than in the American South). These new data add some nuance into the picture: the persistence of female Amerindian lineages tells us that admixture occurred very early on. Additionally, I believe that the likely explanation here is that mixed-race children gained an enormous marginal return from their European ancestry when it came to resistance to Eurasian pathogens. Because of the strong sex-bias in male migration one could posit a circumstance where immigrants from Spain had a tendency to marry native-born women. Over a few generations these women could have been predominantly non-Amerindian in ancestry because of the constant migration of males from Africa and Europe, but traced back along the matrilineage the Amerindian heritage would persist. As I note above, one should be very cautious of extrapolating from the mtDNA and NRY toward total genome content. For example, an individual could have an Amerindian mtDNA lineage, and a West African NRY lineage, and yet still be overwhelmingly European in total genome content (the autosomal genome having most weight here) and manifest a European appearance. I suspect that the first approximation model of African-European admixture to varying degrees will be validated by autosomal surveys, but the mtDNA work gives us a very concrete window into the mating dynamics of the early generations in the Caribbean.