Evolution is all about survival of the fittest. Wild animals are locked in an arms race, driven to be the fast enough to chase down a gazelle or strong enough to fight a python, or else so stealthy they escape the hawk’s eye or so poisonous they kill any attacker. But then why does a bird of paradise do things like this:
Or look like this?:
I just read Jerry Coyne’s Why Evolution Is True, which devotes a good chunk of one chapter to the mating behaviors of birds. What sort of natural selection could lead to such diverse and unusual features in animals? These mating dances don’t seem to help obtaining food or avoiding predators.
The idea is that a peacock, for example, has a whacked-out tail because peahens think that’s hot. (A “peacock” is specifically male and a “peahen” female. Together they’re “peafowl”, and if you eat asparagus you will peafowl.) Therefore, the peacocks with the most-whacked-out tails will get laid and have sons with whacked-out tails, too.
Coyne cites a few experiments that agree with this “sexual selection” hypothesis. Peacocks with more eye spots get more sex, and when some of their eye spots are removed, they get less. Red-winged blackbirds get run out of their territory (and lose their harem of fine blackbird honeys) if they can’t sing. Other decorated birds do poorly at attracting a mate if their colors are painted over.
The surprising upshot is that ornamental tails may be detrimental to survival. They’re heavy and awkward. Peacocks suck at flying, at least in part due to their absurd tails. Nonetheless, the ornamentation can be advantageous over all if having a great tail leads to getting great tail. You might have better chances of escaping predators without the giant colorful feathers, but if you live to be 100 and never get it on, you still lose (by genetic standards). In this way, evolution can create adaptations that hurt individuals’ survival odds and presumably harm the species as a whole.
This reasoning is predicated on females liking ornamented tails, and it’s unclear why they should do that. If crazy tails are detrimental to male health, then shouldn’t females like plain males, because they’re fittest? Females who prefer plain tails will have kids with plain tails, and hence their kids will be more likely to survive. So it would appear that females are willfully injuring themselves by deciding that blue eye spots make for a hunky peacock tail.
Coyne presents three hypotheses that may solve this conundrum of sexual selection:
- Females choose males that will be good dads. The ornamentation may be a signal for this.
- Females choose males that have good genes and will sire strong, healthy children. Only strong, healthy peacocks can afford to grow fancy tails, so the tails are a signal.
- Males are exploiting a trait that exists in female psychology for some other reason. For example, the male ornamentation looks like a fruit the female likes, or females like anything novel or decorative. Males caught on to this and females got duped.
In the case of peacocks, there’s some evidence for number 2 – that only males with good genes grow big tails. Coyne cites a study that found that the children of fancy peacocks are in fact healthier.
The other hypotheses are also interesting and have some experimental evidence to support them in other species, but here I want to head in a fourth direction.
Even if ornamented plumage is a signal for a healthy peacock, why should that be? Wouldn’t everyone be better off with a less-ostentatious signal? What’s the real reason that tails grew into giant fans?
Maybe there isn’t one. The tails don’t necessarily have to signal anything, or trick the females. The tails of peacocks and similar adornments in other animals may be there, despite their negative survival effects, because evolving such traits is the expected outcome of a simple iterated game with thousands of players.
Suppose that long ago, before the peacock got so ridiculous-looking, peahens started to develop a slight preference for longer tails in peacocks for an arbitrary reason, perhaps because malnourished peacocks had smaller tails, or even due to genetic drift (i.e. randomly).
Think about two peahens who are pretty much the same, but one prefers longer tails, as is the fashion, and the other prefers short tails. The one who prefers long tails is at a reproductive advantage, but not because long tails are innately better. Her advantage is that if she chooses a long-tail mate, she’ll have long-tail sons, and all the other peahens, who in general have started to prefer long-tail peacocks, will want to get down with her sons. Her sons will have daughters who have long-tail-selecting tendencies.
The selective pressure on an individual peahen’s preferences now comes not from survival fitness considerations, but from the preferences of all the other peahens. It’s a game theory situation where the payoff for one player depends both on that player’s actions, and on the actions of everyone else.
Because each peahen now feels a pressure to prefer long tails, once we get a few generations down the line, more peahens prefer long tails. When that happens, the selective pressure on any one individual peahen to prefer long tails becomes even greater. It’s a runaway, positive-feedback effect.
Eventually peacock tails are three meters wide and have hundreds of brighly-colored eye spots. Although it didn’t make it into Coyne’s book, I found that this idea has been around since the 1979, and is called the sexy son hypothesis.
If the sexy son hypothesis plays a large role in sexual selection, it says that a peacock’s tail is essentially arbitrary – it just happened to be the feature peafowl fixated on. We might then expect that as we look at different species, their sexual selection should exaggerate different features.
That’s true. In other birds, we see sexual selection acting to create long feathers growing out of the head, or strange crests, or bright chest plumage, or even to create strange behaviors like elaborate mating dances or songs.
Australian bowerbirds are sexually selected not for a physical feature, but for the behavior of building huge, colorful, extravagant "bowers" that they don't even use as nests.
The sexy son hypothesis also suggests that the sexually-selected feature should be as extreme as possible, limited either by the physiology of the animal so that it would be impossible to make it any more extreme, or by the point where the fitness disadvantage to males becomes so great that even the extra mating advantage isn’t worth it any more.
What we have, then, is a hypothesis – the germ of a scientific idea. Once we’ve formulated the sexy son hypothesis, we need to expand on two frontiers in order to test it. One one hand, we should try to develop models of how the sexy son hypothesis works and make quantitative predictions. For example, we might predict a positive correlation between the mating effectiveness of a male and that of his male descendents – the heritability of sexiness. If we went further, we might even be able to predict how strong that correlation should be, based on how detrimental ornamentation is to survival odds and how much variability there is in male reproductive success. On the other hand, we should begin experimental studies to test these predictions.
In a 2008 forum article in Behavioral Ecology, Testing the sexy son hypothesis—a research framework for empirical approaches, Huk and Wenkel summarize the research on the sexy son hypothesis:
To sum up, it can be concluded that empirical studies dealing with critical predictions to date only partially support SSH; that is, only studies with rather small direct fitness consequences are compatible with critical SSH predictions. Contrary, the demonstration of compensation of considerable lower direct reproductive success via a heritable genetic effect of male attractiveness, and hence male mating status in sons, is not demonstrated until now. Thus, facultative polygyny in biparental species seems to be best explained by sexual conflict. Approaches derived from quantitative genetic models of mate choice came to similar results (Kirkpatrick and Barton 1997; Charmantier and Sheldon 2006; Hadfield et al. 2006; Qvarnström et al. 2006). Recent studies therefore support the position that inferior direct reproductive success cannot be overcompensated by a “sexy son” effect (e.g., Kirkpatrick 1985). Thus, attractiveness of sexy sons and its resulting fitness advantages seem to be of minor biological effect.
Certainly not a strong avowal, but not damning, either. The jury is still out, so until next time, stay sexy.