...they found that species which relied on a class of chemicals called carotenoids to tint their feathers fared worse when there was more radioactivity around. Intriguingly, that did not apply to birds that used melanin....
Besides acting as pigments, carotenoids are antioxidants that have an important role in protecting DNA from harm. One of the ways that radiation causes harm is by generating molecules that promote oxidation, so a good supply of carotenoids protects against such damage. Using them to make feathers pretty instead of mopping up oxidative molecules thus has a significant cost--as this result shows.
The reason this is interesting is that there is a debate in biology between those who think signals such as flashy feathers are essentially arbitrary and those who think they are signs of underlying health and good genes. Dr Moller's and Dr Mousseau's result shows that the bright reds and yellows of carotenoid-based plumage really do come at a price, and thus indicate underlying health. The unusual circumstances of Chernobyl have exposed that price to human observers, but it will have to be paid all the time, even in places that have not fallen foul of radioactive plumes.
Too much in discussion about sexual selection in the context of evolutionary biology operates so that it appears like a deux ex machina to rescue us. Like genetic drift too often people use sexual selection as a catchall explanation for traits which they can't understand in an adaptive context. But though drift and sexual selection are real evolutionary forces it is important to remember that both are subject to conventional population genetic parameters. Additionally, theories of sexual selection come in varieties, with some being rather more arbitrary than others. Sexual selection maybe stochastic, but the sample space of possibilities does not seem infinite or unconstrained.