DARC (Duffy antigen receptor for chemokines) is a receptor expressed on red cells, key in the infection of those cells by vivax malaria. A version of the gene that eliminates expression of this receptor on red cells (and prevents vivax infection) has reached very high frequency (up to 99%) in sub-Saharan Africa. A new article in PLOS Genetics concludes that this was driven by some kind of strong selection, but a long time ago ( ~40,000 years). They estimate that the Duffy-negative mutation conferred a selective advantage of about 4.3%, leading to effective fixation in about 8,000 years. Not everywhere in Africa – didn’t happen in the Bushmen, Naturally they suspect that the selection was driven by vivax malaria , but they’re a bit mystified because today vivax is usually mild ( at last when compared with falciparum malaria). Vivax is often found in places where it can’t be transmitted year-round (because of winter) : it waits in the liver [hypnozooites] and reappears later, sometimes much later. I’ve heard of cases where it reemerged as much as 40 years later. Which means that it can’t afford to be too severe – it has to let the host survive.
However, in the tropics, it could get away with being more virulent, and perhaps it did, in the past.
An advantage of 4.3% is a lot, as selective advantages go, easily enough to drive a gene frequency from a tenth of a percent to 99% in a few thousand years. In much the same way, a behavioral tendency (heritable of course) that led men to have sexual intercourse with women other than their primary mate, when practical, could become very common even if it only resulted in a few percent extra surviving children per generation, say 2.2 instead of 2.1 . Compound interest. A behavior wouldn’t have to yield 100 extra children in order to become common (although producing an extra 100 would certainly have that effect!). In the same way, a heritable behavioral tendency that reduced the average number of offspring by a few percent would also become rare over a few thousand years.
What fraction of the practitioners in various human sciences understand this wee bit of population/quantitative genetics? Very few, I think.