Sorry, but those of you arguing blood groups are neutral or evolutionary remnants are just plain wrong. There are literally dozens and dozens of different types of blood types, each determined by specific, gene-level differences - they're all
polymorphic. Most of them are adaptations to different types of diseases.
Aside from the well-known ABO & Rh(esus) groups, there are the Duffy, MN, Diego, Colton, Kell, Lutheran, Lewis, Xg, Bombay,
S (sickle cell)... The list goes on and on. If memory serves all of them have at least some hypothetical reason for variation. Others, we know for sure exactly why they've evolved some variability. I'll list & describe some of those here:
ABO: This blood group has three alleles, A, B, & O. A & B are codominant, O is recessive; there are four phenotypes: A, B, AB, & O. B & O phenotypes are more resistant to such pathogens as venereal syphilis, B is more resistant to plague, A is more susceptible to smallpox, etc. Perhaps more importantly, the different phenotypes experience what's called prezygotic selection - sperm of certain blood groups are less likely to fertilize ova of another; for example, non-O sperm (A, B, or AB) are less likely to fertilize an O ovum than are O sperm. Haldane recognized in the 40s (I think) that this type of prezygotic selection can actually drive speciation - so this isn't just microevolution we're talking about here.
S: Sickling is a common adaptive response to malaria; there are multiple types of sickle cell (different alleles), but the most common is the
S allele. Here, there are three genotypes: AA (normal hemoglobin, normal red blood cells), AS (sickle cell trait, red blood cells sickle under certain stressful conditions), and SS (sickle cell anemia, red blood cells sickle most of the time). AA has no resistance to malaria caused by
Falciparum organisms, AS has a very high resistance to
Falciparum malaria (relative to AA), and SS is nearly 100% lethal before the carrier reaches adulthood. Thus, the S allele is most frequent in areas of the world where malaria is endemic (sub-Saharan Africa, the area around the Mediterranean, Southeast Asia) - there is no advantage to having an S allele where there is no malaria, because there's no malaria danger to outweigh the dangers of having either sickle cell trait or sickle cell anemia. The
S allele is a fantastic example of what's called a
balanced polymorphism. You can read more
here, from the man who recognized the adaptive nature of sickle cell anemia over 50 years ago.
Duffy: Duffy has three alleles: FyA, FyB, & Fy (the recessive). Individuals who are FyFy are nearly 100% resistant to malaria caused by
Vivax - further, there are no known ill effects of being FyFy (unlike AS or SS, where AS can have problematic sickling and SS is lethal). Frequency of FyFy is at or very near to 100% in certain populations, like the pygmies of Western-central Africa - vivax malaria, unsurprisingly, simply doesn't affect those populations, because they're completely resistant. FyFy individuals are practically non-existent in people from Northern Europe & Northern Asia - the Fy allele probably hasn't had time to make its way into those populations, and while there's no disadvantage to being FyFy, there's no advantage to it where there's no vivax, so this is not a balanced polymorphism.
If you want to read more about blood groups & their evolution, check out Molnar's
Human Variation: Races, Types, and Ethnic Groups or Relethford's
The Human Species: An Introduction to Biological Anthropology. Non-current editions can probably be had on Amazon for pennies.