Prior posts have noted that selection factors that influence evolution can be highly variable, and sometimes surprising. It is not the simple "survival of the fitest" postulated by Darwin (who in fairness could not have known about many of the selection factors and processes described in these posts). Great example? Let's go with the Transylvanian naked-neck chicken (see http://en.wikipedia.org/wiki/Naked_Neck).
Why naked necks? At university, as they say, perhaps even in high school, we heard that various vultures had naked necks so that blood and body parts would not cake onto feathers. Ostriches and emus? To radiate heat generated from running. But that still begs the question, what is the mechanism that allows a naked neck to be "expressed", as they say.
First, you start with a DNA mutation. You may recall that the "accepted wisdom" once upon a time included this thing called "junk DNA". We and all creatures had lots of genes that allegedly did nothing, we were told; they were junk. No so fast, grasshoper. Turns out a lot (most? all?) of that "junk" does stuff, sometimes very important stuff. In this case, the gene that mutated did not influence a protein, but it did change the "activity" of a nearby gene that encoded a protein known as BMP12 [I know, I know, I am simplifying the protein encoding process; if you want to know more about the process, as such, see "How Proteins Are Made" at http://wps.prenhall.com/esm_krogh_biology_3/17/4439/1136394.cw/index.html].
The change causes more BMP12 to be made. This protein is apparently one of many that control the density of hair on mammals and the density of feathers on birds. Here, BMP12 slows the growth of feathers. Ah, but the story gets more interesting.
Since a genetic mutation that is inherited is present in every cell in the body, why does it only "express" its action in the neck of our now famous Transylvanian naked-neck chicken? Funny you should ask. The researchers found a link with retinoic acid, a derivative of Vitamin A. Retinoic acid regulates the differentiation of tissues in developing embryos. What was discovered was that retinoic acid enhances the the effect of BMP12 so that BMP12 does not merely slow the growth of feathers, it stops the growth of feathers.
Now, of course, you are saying to yourself, but retinoic acid must be present in all cells...and of course you would be right. But (and there is always a "but"), for reasons not yet understood, embryonic neck-skin of birds produces more retinoic acid than the embryonic skin from any other part of the body. [A good question to ask at this point, of course, is whether the extra retinoic acid being produced in the embryonic neck-skin is cause or effect of some other selection mechanism. Great question. No answer, at least not yet.]
Confirmation of this interplay between BMP12 and retinoic acid has been demonstrated with cultured skin samples from ordinary chickens by dosing the samples with BMP12. Naked neck produced, okay, okay, in a cultured skin sample. Good proof of concept, though, as they say.
The report on this process can be found at http://www.nature.com/nature/journal/v471/n7339/full/471413c.html.