The birth of new species always involves a barrier to cross-breeding between two different groups of the same species. This barrier may start out as a geographical barrier (two raccoon populations on different sides of a mountain never encounter each other and thus fail to interbreed), but however it starts, reporductive barriers always turn into a genetic barrier. To form new species, two populations of organisms have to drift apart genetically.
The genetic split can happen in a variety of ways, as scientists are discovering in the their quest to find 'speciation genes.' It can happen because a selfish gene fails to be shut down in the offspring of cross-breeding flies, and it can happen because one mouse gene doesn't work right when it encounters genetic variants from another subspecies.
A report in Science describes one more speciation gene, this time in two sub-species of thale cress plants. In this case, the barrier to reproduction is the result of faulty gene copying.
At some point in thale cress history, a gene called HPA coding for a metabolic enzyme was accidentally copied. For a time, the two copies of this gene, each one at a different place in the genome, produced a functional enzyme. However, since two copies of the HPA gene are not needed, mutations eventually began to erode one of the copies. Since the mutations were not harmful to the plant, natural selection did not weed them out, and one copy of this gene was rendered completely non-functional.
The catch is that, in different populations of thale cress (all of the same species), a different copy of the HPA gene was destroyed. In one population, copy 1 (on chromosome 1) was knocked out, while in the other, copy 2 was destroyed (on chromosome 5). When you put a deficient copy 1 and a deficient copy 2 together in the same organisms, bad things happen.
To understand what bad things are going on, remember that these plants are diploid organisms, meaning they have two copies of each chromosome. So it's possible that a plant can have one good and one bad copy each of chromosomes 1 and 5; those plants do just fine, because they still manage of have enough good copies of the HPA gene. Other combinations of chromosomes don't do so well: a plant that has two bad copies of chromosome 1, but one good and one bad copy of chromosome 5, still has one functional HPA gene (on the good chromosome 5); it is alive, but not thriving. Plants that have all bad copies of both chromosomes 1 and 5 never make it out of the embryo stage.
In other words, after the HPA gene was accidentally duplicated in an ancestral population of thale cress, different sub-populations started to follow different evolutionary trajectories, by collecting mutations in different copies of the HPA gene. These two sub-populations have barely started down their diverging evolutionary roads, but the distance they've traveled is enough to put up a significant reproductive barrier between these two populations of thale cress. New species develop by these first small genetic steps.
If you've been following along in this series, you're probably sick of speciation genes by now. Tomorrow we'll leave speciation genetics for the time being and talk about some fossils.
In the meantime, go check out the Rugbyologist's reluctant celebration of the Darwin Bicentennial. Read about how he thinks his takedown of the Bicentennial made me cry, but in reality, those were tears accompanying uncontrollable laughter because he's got some hilarious stuff over there, including a third order polynomial model of the demise of creationism.
Join me tomorrow, here at Adaptive Complexity, for day 8 of Show Me The Science Month. Evolution as a science is alive and well. Each day I will blog about a paper related to evolution published in 2009.