We analyzed two admixed populations that have experienced different demographic histories, namely, Caracas (n = 131) and Pueblo Llano (n = 219). The native American component of admixed Venezuelans accounted for 80% (46% haplogroup [hg] A2, 7% hg B2, 21% hg C1, and 6% hg D1) of all mtDNAs; while the sub-Saharan and European contributions made up ∼10% each, indicating that Trans-Atlantic immigrants have only partially erased the native American nature of Venezuelans. A Bayesian-based model allowed the different contributions of European countries to admixed Venezuelans to be disentangled (Spain: ∼38.4%, Portugal: ∼35.5%, Italy: ∼27.0%), in good agreement with the documented history.
From here (hat tip to Dienekes).
Thus, approxmiately 80% of the matrilineal ancestry of urban Venezuelans was Native American, 10% was African (presumably descendants of women brought to the New World in the slave trade), and 3.8% is Spanish, 3.6% is Portugese, and 2.7% is Italian.
One would expect that the patrilineal ancestry of Venezuelans (and also autosomal ancestry which is derived from all ancestors, rather than merely the patriline and matriline ancestors) would have a much larger Southern European component. Also, by implication, the percentage of urban Venezuelans who have exclusively European ancestry is less than (and probably significantly less than) 10%.
The Native American component, consistent with other studies of Native American mtDNA in North American and in Latin America, shows less genetic diversity than many North American Native American populations. It remains unclear how much of that difference is due to serial founder effects and how much is due to the impact of possible subsequent waves of migration to the Americas that did not penetrate to Central and South America in significant numbers.
It is also reasonable to expect, given historical patterns of colonization in Venezuela, that the Native American component of rural populations in Venezuela is probably larger than the population in two of its largest cities.
This data point also supports the frequent reality that a paternal superstrate population can give rise to language shift even in the absence of a large maternal genetic contribution.