Reverse Evolution in Sticklebacks?

A number of people have emailed me a link to this story about “reverse evolution” in sticklebacks residing in Lake Washington. The basic story is that as Lake Washington was cleaned of its pollution sticklebacks went from being low plated to being completely plated. Partially, this was a response to heavier predation caused by increased clarity of the water.

The newspaper story gets some of the basics right, but mangles the rest:

Those ancestors were marine stickleback, a spiny, armor-plated species that invaded freshwater around Puget Sound as the glaciers retreated some 10,000 years ago.
Over time they evolved to shed the bony plating covering their bodies from snout to tail. By the 1960s, only about 6 percent of sticklebacks in Lake Washington were fully plated.
Fast forward to today, though, and about 49 percent of the stickleback sampled for the study were fully plated, the researchers found. And 35 percent were partially armored.
Researchers surmise that is because marine versions of the fish, with armored plates, invaded the lake anew when the fish ladder was put in, mixing their genes with the unplated fish.
So, when the water was cleaned up, the fish were able to tap those genes to zoom backward to their earlier, plated version, and armor themselves against the cut throat trout that suddenly could see — and eat them.

Well, yes and no. Sticklebacks have been extensively studied by evolutionary biologists. As the article points out, the marine form of the stickleback is completely armored. As the marine form invaded freshwater they lost their armor (Nuclear and mtDNA studies rule out the idea that a single population of low armored fish migrated throughout the world. The loss of armor in freshwater fish is caused by two mechanisms, independent mutation and selection acting on already present genetic variation). There has been a certain amount of gene flow back and forth between both forms and this accounts for the presence of low armor alleles among the marine morphs. The presence or absence of armor is actually caused by ectodysplasin alleles (which comes in two alleles, complete and low) and there are actually three morphs, pictured below.
Marine forms are generally homozygous for the complete form, although as mentioned above, due to gene flow with freshwater forms they can be heterozygous (i.e. they have a complete and a low allele). This brings us to the Lake Washington study which is being published in in Current Biology. This is what they found:

In contrast to the prevalence of the low-plated morph in many freshwater environments …, we found a high frequency of completely plated sticklebacks in Lake Washington, an urban freshwater lake in Seattle … In 2005, we found that all three lateral-plate morphs were present, with 49% completely plated morphs, 35% partially plated morphs, and 16% low-plated morphs (Figures 1A and 2C). Although a previous study had also shown that all three morphs were present in Lake Washington in 1968-1969, only 6% were classified as completely plated morphs (Figure 1A) … Instead, the low-plated morph, with a mode of seven plates, was the most common morph until the late 1960s.

So far so good. This demonstrates that the frequency of completely plated sticklebacks has increased at the same time that the lake is being cleaned up and visibility is increasing. So what explains this increase? The Seattle Times article linked to above says:

Researchers surmise that is because marine versions of the fish, with armored plates, invaded the lake anew when the fish ladder was put in, mixing their genes with the unplated fish.

Gene flow was examined in the paper. Sticklebacks were collected from Puget Sound, multiple points Lake Washington and in neighboring streams. The specimens were genotyped with 15 microsatellite markers. Results indicated that sticklebacks sampled from near the ship canal (built in 1917) were similar to marine morphs, while those sampled near streams were close to those from neighboring streams. The question is, can this gene flow account for the amount of change in armor? To examine this question the researchers estimated migration rates based on their data. Long story short, the answer is no. The estimated migration rate was to small to account for the change (you will actually have to read the paper for details).
What about selection then? Based on their estimation of migration rates, the researchers were able to calculate the strength of selection on the completely plated morph. It turns out that their were two different bouts of selection. From 1967-76 the selection coefficient was .58-.72, while from 1976-2005 the selection coefficient was .01-.03. What is the cause of the selection? The main predator on sticklebacks in Lake Washington are cutthroat trout. Cutthroat trout are visual predators and are affected by changes in water transparency and the cleaning of Lake Washington created an eightfold increase in their visual search volume and increased prey encounter rates.
So far, this is a pretty good paper that can, basically, be interpreted in terms of selection favoring one morph over the other – the completely plated morph went from 6% in 1968-69 to 49% in 2005. But then the authors toss us a curve by talking about “reverse” evolution. I find this puzzling. Granted that the marine forms are usually completely plated and freshwater forms loose their armor, the fact remains that completely plated forms did exist in the Lake Washington population. Furthermore, both the completely plated and partially plated forms increased in frequency, while the low plated forms decreased from 1968-69 to 2005. I’m just not seeing how this constitutes a reverse…