The Eating of the Blondes - genes, that is!

No, this isn't a post about cannibalism!
Genophagy - literally the "eating of genes", is a discredited notion that was popular at the dawn of genetic science in the early part of the 20th century. Back then, it was believed that the dominant alleles for any given trait would eventually overwhelm the recessive ones - the dominant genes would seem to "eat" the recessive ones right out of the population. What is interesting to me is that this idea never really seems to be put to rest, at least in the popular perception of how genetics works, which helps to explain why so many news outlets (and very credible ones at that!) picked up the hoax story a few years ago about blondes going extinct in the near future.

Why aren't blondes going extinct? Aren't dominant genes "stronger" and more widespread?
Not really. Dominant just means that the allele gets expressed in heterozygotes' phenotypes. What's usually going on at the cellular and molecular levels is that both alleles are getting translated, but the dominant allele's functional product is obvious in the phenotype, whereas the other gets "hidden". A good example of this is the ABO blood system. Recall that both A and B alleles are codominant, but both are dominant to the O allele. This is because the A and B alleles code for a fully formed, functional antigen, whereas the O allele codes for a shortened molecule that does not function as an antigen, basically because it's missing a piece. Thus, in heterozygotes with one O allele, the other allele is making antigens, while the O is making non-antigen molecules. Phenotypically, only the antigens count.

The genes behind hair color are somewhat similar. Like the eyes and skin, hair color is all about melanin. More melanin = darker hair. So blonde people, like blue-eyed people, have few melanocytes. So dark hair is dominant just because dark hair genes actually code for melanin to be produced in the hair.

Dominant genes do tend to be more widespread, but not always. Take the case of "milk allergies", which are not usually allergies at all (a true allergy is when the body's immune defenses activate in response to common, non-disease causing things like pollen or peanuts). Most people who are "allergic" to milk are actually intolerant of the milk sugar lactose because they have lost the ability to produce the enzyme lactase, which breaks it down (digests it). Almost everyone in the world can digest lactose from birth to about 2 years old, but most people lose the ability to product the enzyme sometime between ages 2-6. The gene that causes this (it's a regulatory gene that "switches off" the gene that makes lactase) is recessive. In some populations, many people have inherited a mutated version of this gene, which keeps lactase production going into adulthood. This gene is dominant to the "switch off" gene. So, if most people in the world are lactose intolerant as adults, most people in the world are homozygous recessive.


Hardy-Weinberg to the rescue!
Yes, these two mathematicians can help us understand why Blondes aren't going away anytime soon. Or, probably, ever. These two gents actually proposed their equilibrium model as a way to disprove the genophagy theory. In most populations most of the time, most allele frequencies are fairly stable over time. There are only a few factors that can shift them around, those 5 forces of evolution we've been talking about (natural selection, genetic drift, gene flow, mutation, and assortative mating).
What about our world population of people and the gene locus for hair color - are we evolving?
If so, which force or forces are at work?
For blonde people to be selected out of the population, they'd have to have fewer babies that nonblondes (hey, that sounds like the name of a band!) This could happen if blonde people were somehow less able to have kids (blonde genes somehow related to problems with reproductive organs, perhaps?), if blonde people were less able to avoid predation (blondes more easily detected by bears?), if blondes tended to die before reaching child-bearing age (blonde hair = leukemia?) and so forth. Of course, none of my proposals above is really happening.
What about assortative mating - do gentlemen (or gentlewomen!) prefer blondes? If this were true, then we can imaging that blondes might have even greater reproductive success than other hair colors, thus increasing the blonde genes in the population!

At any rate, I don't think blondes are going anywhere anytime soon.