A new paper in MBE that uses publicly available SNP and other sequence data to examine which of the candidate skin pigmentation loci are under positive selection. I guess that the main finding here is that both dark and light skin have adaptive value (not just relaxation of selection pressures). Among the African populations, since they find two loci that show positive selection, the authors suggest that dark skin has an adaptive value, possibly as protection against sun-induced damage. In addition:
"The mechanisms by which light skin confers a selective advantage cannot be inferred with these data alone. However, the relevance of mechanisms like vitamin D synthesis, which have been claimed to have a role in the evolution of light pigmentation (Robins 1991 and references therein), could be minimized as neither of the representative loci of this pathway (Omdahl et al. 2002) that have been analyzed, VDR (Vitamin D receptor) and VDRIP (Vitamin D3 receptor-interacting protein), CYP27B1 (P450C1), CYP24A1 (P450C24), and RXRA (Retinoid X receptor), showed clear evidence of positive selection."
A Scan for Signatures of Positive Selection in Candidate Loci for Skin Pigmentation in Humans
Neskuts Izagirre, Iker Garcia, Corina Junquera, Concepcion de la Rua, Santos Alonso
Molecular Biology and Evolution (2006); 23(9) 1697-1706
Abstract: Although the combination of pale skin and intense sun exposure results in an important health risk for the individual, it is less clear if at the population level this risk has possessed an evolutionary meaning. In this sense, a number of adaptive hypotheses have been put forward to explain the evolution of human skin pigmentation, such as photoprotection against sun-induced cancer, sexual selection, vitamin D synthesis or photoprotection of photolabile compounds, among others. It is expected that if skin pigmentation is adaptive, we might be able to see the signature of positive selection on some of the genes involved. In order to detect this signature, we analyze a battery of 81 candidate loci by means of phylogenetic and population genetic tests. Our results indicate that both light and dark skin may possess adaptive value. Of the main loci presenting this signature, TP53BP1 shows clear evidence of adaptive selection in Africans, whereas TYRP1 and SLC24A5 show evidence of adaptive selection in Caucasians. Although we cannot offer a mechanism that based on these genes explains the advantage of light skin, if TP53BP1, and perhaps RAD50, have truly conferred an adaptive value to the African population analyzed, photoprotection against sun-induced skin damage/cancer might be proposed as a mechanism that has driven the evolution of human skin pigmentation.