The End of Sex Week: Darwin, Sex, and Dada

By Carl Zimmer | July 30, 2010 8:48 am

[This is the last post for Sex Week]

The animal kingdom is filled with wild extravagances, and a lot of them have something to do with sex. Hermit Fiddler crabs wave their claws, swordtail fish flash their swordtails, manakins leap and buzz their wings. Darwin considered these displays so important and so puzzling (“the sight of a feather in a peacock’s tail, whenever I gaze at it, makes me feel sick!” he wrote to a colleague), that he dedicated half of a book to the subject.

Darwin argued that many extravagant displays in male animals were the result of a special kind of evolution he called sexual selection. Females preferred males with certain traits over other males, and so those males had more offspring, which inherited their traits. In recent decades, scientists have documented many cases in which females do indeed prefer males with certain traits over others. As I mentioned in my post on electric fish, for example, female bulldog fishes are more attracted to long electric pulses than short ones.

But why should females have any particular desire? For many scientists, the most compelling explanation is that there’s some meaning in the displays that attract them. One possible meaning is that the male who sports a particularly extreme version of a trait has good genes. As I wrote in my post on the love songs of yeast (sic), the amount of pheromones a yeast cell produces is a reliable clue to the quality of its genes. Other scientists have found similar links between the length of tail feathers and the quality of male birds.

There are other possible meanings, though. Some scientists have argued, for example, that female of some birds species prefer bright feathers on males simply because they’re easy to spot. A female with such a preference will be able to mate more efficiently than a female attract to drab, harder-to-find males. She’ll be less likely to get killed by a predator and can spend her time and energy on more useful things than hunting for a mate. Wham, bam, thank you, sir.

All of these explanations share something in common: sexual signals evolve because they signify something. Many scientists see little evidence to think that sexual displays evolved for no good reason.

And yet in other realms of evolution, biologists have come to accept that some patterns can emerge without the help of selection. Each of us is born with around 70 new mutations in our DNA. In a few cases, a mutation will cause a lethal disease and will not be able to be passed down to the next generation. In a few cases, a mutation will boost a person’s reproductive success and will gradually spread. In both these cases, natural selection is at work. It’s the reason why the harmful mutation is rare, and the beneficial one is widespread. But the spread of mutations is also governed by chance. A mutation that has no benefit or risk at all can spread throughout an entire species thanks to a fortunate roll of the genetic dice.

There was a time when most biologists did not believe that neutral genetic variation even existed. It all had to be the product of selection. Now, however, biologists generally agree that neutral genetic variation is rampant. In fact, netural evolution has taken on an important role in the study of natural selection. Scientists who want to know if a particular stretch of DNA has experienced natural selection must reject the null hypothesis that it is simply the result of neutral evolution. If you can show that neutral evolution couldn’t have produced a particular sequence of DNA, then you can be fairly confident that selection was responsible. Once you do that, you can start to investigate what sort of selection was at work. (For more on the underappreciated role of neutral evolution, check out Larry Moran’s periodic posts on the subject on his blog, Sandwalk.)

The phrase null hypothesis was first coined in the early 1900s by a British mathematician named Ronald Fisher. Fisher did some of the most important work to put Darwin’s theory of evolution on a sound mathematical footing–figuring out how to represent things like natural selection as equations and graphs, rather than just verbal arguments. Fisher showed, for example, how natural selection could proceed through the spread of lots of mutations with tiny effects. But Fisher also developed an idea that’s not so well remembered these days. Like Darwin, he pondered how sexual displays could evolve through female preferences. One idea he came up with is that a display–and a female’s preference for it–could both be completely arbitrary.

It’s fair to say that most evolutionary biologists today don’t find Fisher’s idea very useful. Nevertheless, some important thinkers have embraced and updated it. And in a new review in the journal Evolution, the Yale evolutionary biologist Richard Prum makes a bold case for taking Fisher seriously. Prum argues that it’s quite reasonable to expect sexual signals to be totally arbitrary, signifying nothing deeper about the animals who show them off. In fact, he argues, it should be the null hypothesis for scientists studying sexual displays.

I’ve known Prum for a few years now, having written some articles on his work and having had him talk to my writing class. Many of our conversations have gravitated to this big idea, which he’s been mulling for some time. So it’s good to finally see this argument in print at last.

The idea that something like a courtship dance or a song can evolve with no help from selection is a tricky one to grok. Here’s a simple version of the model. Imagine a population of birds. The males have genetic variation in the size of a red spot on their breast. Some have a bright red spot, and others have a dull one. The females, on the other hand, have genetic variation in their preference for the trait. Some only mate with males with bright red spots, and some will mate with any males. The extravagant males will have more offspring than the plain ones, because they can mate with all the choosy females and some of the non-choosy ones. The plain males only mate with the non-choosy ones. What’s more, the males with the bright red spot and the choosy females combine their genes in their offspring. The population increasingly is made up of males with an extravagant trait and females with a preference for it. And so the extravagant display spreads quickly through the population–even though the trait doesn’t signify anything.

Mark Kirkpatrick of the University of Texas, Russell Lande of Imperial College London, and their colleagues have developed much more sophisticated mathematical models of Fisher’s idea. They have found that even if the variations in male traits and female preferences are subtle, they can still get swept up into all sorts of complex evolutionary changes. Yet many others have been skeptical. One scientist declared that excepting this runaway process as an explanation for sexual selection without a lot of proof was “methodologically wicked.”

One objection was that sexual displays often impose such a big cost that they have to have some major benefit. Yet Kirkpatrick and Lande have shown that female preference and male displays can drive each other’s evolution so hard that males may end up with traits that are a major burden. They’re just so sexy that the males still continue to spread their genes. In fact, Prum argues, this runaway process is much more powerful and flexible than sexual selection based on an honest signal. If sexual displays are just relaying honest information about males, then why are closely related species so different in what they find attractive?

In his review, Prum looks at a series of studies in which scientists tried to figure out the reason that birds have extravagant traits, such as complex songs. He notes how the scientists always go into the studies assuming that there must be some meaning to the trait, so that its evolution can be driven by selection. Sometimes they fail to find that meaning, and when they do, they conclude that there must be another meaning they haven’t found yet. All they’re doing is trying to confirm a foregone conclusion, Prum argues, when they out ought to be trying to reject the null hypothesis.

If they can’t reject the null hypothesis, they should conclude that the best explanation for a bird song or a dance or some other display is that it’s just arbitrary. Prum doesn’t think that every sexual display will turn out to be arbitrary, but he expect that a lot of them will.

“I do not claim that the ‘Emperor has no clothes,'” he writes in his conclusion. “I would predict that the ‘Emperor wears a loincloth.'”

I’d be curious to know what evolutionary biologists who work on sexual selection think of Prum’s manifesto (paging Marlene Zuk). It’s certainly provocative, evoking not just Stephen Jay Gould’s attacks on Dawkins and other adaptationists, but even the Dadaists, who toyed with the arbitrariness of beauty in pieces such as Marcel Duchamp’s Fountain, shown here. And it’s also a good way to end Sex Week, as we leave biologists continuing to argue and marvel over the mysteries of sex.

[Update: Marlene Zuk responds. Definitely worth checking out. This is what makes blogs so cool.]

[Update: Rick Prum weighs in with a lengthy comment. Coolness upon coolness.]

[Image: Wikipedia]

CATEGORIZED UNDER: Evolution, The Tangled Bank

Comments (29)

  1. unemployed

    It has been suggested that these sexual displays are proxies for fitness, since it is not generally feasible for a female to put her prospective mates through a genuinely life-threatening ordeal to determine his suitability. Along these lines, and with a nod to Oliver Sachs’ book “Musicophilia”, it seems to me that our human fascination for rhythm and dance is an extended proxy to evaluate a potential mate’s level of coordination, since our bipedal gait compels us to have a much closer and simpler relationship with the timing of our movements, and since subtle differences in stamina and balance probably correlate well with our ability to survive in the nascent human niche.

    Just a theory in development; I’ve been waiting a while for a suitable thread to share it!

  2. gaddeswarup

    Interesting post. In countries like India many, particularly males, prefer (not only now but several enturies) fair partners. I wonder this led to any thing testable.

  3. MT-LA

    Wow…that’s a long way to go for a justification of putting a urinal at the top of the post!

    I don’t have any insight to offer here, just my thanks. I really enjoyed this series of posts. I’ve gotta say that the bar scene this weekend will be much more interesting as I observe my fellow males…I just wish I had a bright red dot on my breast!

  4. Rick Newman

    It seems to me that the red-dotted hypothesis is flawed, in that it starts with the a priori assumption that some females prefer males with right red dots. Why? To simply assert that this is so because it is a random preference, does not, in my opinion, really account for anything. The hypothesis seems to simply say that traits we can’t explain can propagate, without explaining why they should exist in the first place.

  5. Boy, it’s been too long to remember the source (it might even have been Discover Magazine, but I think it was Natural History) but I remember a really interesting article about attraction in a particular minnow-sized fish from… somewhere in the world. (I feel pretty uncomfortable not having a better citation.)

    Anyway, the naturalists referenced in the article had observed that females of this particular species were more strongly attracted to males with longer tail fins. To test their hypothesis the researchers attached artificially long fins and, sure enough, the females were even more strongly attracted. The effect worked even in males who were otherwise small, less healthy, or otherwise les “fit.” In a particularly interesting variation they attached long tailfins to males of a similarly sized but unrelated species and… the females became just as strongly attracted to them!

    One observation of the researchers was that the females seemed to have extra “capacity” for attraction with the results that over time males with mutations for even longer tails would be more reproductively successful. If they’d been aware of the theories in this post they might have added that in the absence of obvious advantage (i.e. less healthy males, males of other species) the trait indicated only “personal” preference rather than some kind of “anticipatory” signal of fitness.

    Sure, over time the combined preference for longer tails plus mutations that lengthened them could lead to tails so baroque that only the strongest, best-fed, and parasite-free males could compete. At which point it could *become* an “honest signal.” But you couldn’t say it *originated* as a response to pre-existing female “demand” for fitness.

    I love the “Emperor wears a loincloth” point. That’s probably about right. There certainly are sexually-selected characteristics and many of them may be based on female-selection preference for distinguishing fitness. But it makes sense that if selection for actual physical health and fitness was the criteria then the most Occam’s Razor-y measure would most often be actual, overall, not-just-a-tailfin health and fitness.


  6. Preff

    Arbitrariness is a factor, as is female preference for displays of genetic health. Of course it makes sense to the lay person that variations can be caused by various factors. We regular folks have a saying – Stuff happens! The ‘Nature vs Nurture’ debate is a classic example of scientists focusing on one side or the other, when both are so obviously factors. Gould’s arguments were inflexible (but it was an attention-getting schtick!), and he and so many others have subverted the scientific process, which demands open-mindedness. It’s refreshing to hear that Prum doesn’t think every display will have arbitrary causes. But he’s probably wondering if the funding of his research will continue if he doesn’t come up with some satisfyingly black-or-white conclusions!

  7. MT-LA

    Rick Newman: I think you might be misreading the intent of the “red dot” example. The way that I interpreted is: IF some of the females had a genetic predisposition to prefer males with brighter dots, then the trait that gave brighter dots MUST AND WILL propagate because of the females real-but-arbitrary disposition for it.

    The assumption of the female’s disposition was only meant to be an example of how an arbitrary trait might be passed along.

    The point of the debate is one side says that if a female has a predisposition towards males with some trait, then there must be “some meaning in the displays that attract them.” The other side says there is a very real mechanism for how an arbitrary predisposition among (even a minority of) females will, by necessity, drive the males towards a display that will satisfy that predisposition.

    As if there was any doubt about who is in charge out there…

  8. cobra commander

    if there are some females find a red dot male attractive, doesn’t it make sense that roughly the same proportions of females will find the red dot appalling?

  9. Steviepinhead

    Great series, Carl!

    Especially this last entry.

    Though I think you want “accepting” instead of “excepting,” here: One scientist declared that excepting this runaway process as an explanation for sexual selection without a lot of proof was “methodologically wicked.”

  10. Anonym

    “…when they out to be trying to reject the null hypothesis.” Anybody find that ‘gh’ you dropped?

    [CZ: Found, fixed. I’d thank you, but I have no idea who to thank.]

  11. Marlene Zuk

    Thanks for the invite – as it turns out I was one of the reviewers for Rick’s paper (which he knows, so no state secrets revealed). I like it because it places the Fisher process in a new position. Instead of being pitted against so-called “good genes” models, he uses it as Carl says, as a null hypothesis that we need to falsify if we want to demonstrate that mate choice is adaptive.

    Fisherian runaway selection has always been hard to understand, partly because we humans like to think of female animals as getting something useful out of mating with males that have goofy traits like peacock tails, and partly because Fisher himself is really hard to understand (needed to take writing lessons from Carl in a big way). But I think figleaf is on to something in making the distinction between the origin and maintenance of long tail fins in fish (this may refer to some of Alex Basolo’s work on swordtails). When you look at an elaborate trait like a long tail or red spot, and ask, “What’s going on with that?” you are really asking several questions. First, how did it get there? Second, why a red spot and not a green carbuncle? And third, what keeps it there, i.e. what evolutionary process maintains it in the population?

    Fisher’s process can be going on with the first or the third questions. We don’t know much about the second, although some interesting work on the swordtails and other related fish suggests that females may have a sensory bias for certain traits; they may gravitate to red spots because their eyes are more sensitive to the red fruit they eat, for example. But it’s perfectly possible for a trait to arise and get more common in a population through runaway selection, and then to be maintained because the trait becomes linked to genes for parasite resistance, say (my own favorite idea), and hence indicates a higher quality male to the female who prefers the red spotted males. So both Fisher and the “good genes” models work at the same time.

    The other major objection to some of Kirkpatrick and Lande’s models was that female choice itself is often costly, so that females need something to make up for the time and trouble they take to seek out and mate with the red-spotted males. The models mostly assume that the cost is zero, but we know that’s not the case for a lot of species, where females travel long distances and evaluate lots of suitors before choosing one.

    Finally, I really hate to be picky, but it’s fiddler crabs, not hermit crabs, that wave their claws! One of my graduate students is working on signals and parasites in fiddlers and I want the record set straight!

    [CZ: Fiddlers it is. Thanks for the comment and the correction.]

  12. @Marlene Zuk: Thank you, thank you! I’ve been looking for the citation on and off for quite a few years. Looks like the article I read was based on “Basolo, AL 1990 Female preference for male sword length in the green swordtail. Animal Behaviour 40, 332-338” or one of a couple related papers from the same time. After almost 20 years I’d muddled some of the details but correctly remembered that females were predisposed to prefer long tails.

    I like the idea of traits such as spots becoming bound closely to critical genes. When I was in college just before gene sequencing became at all practical my microbiology professor said that was one hypothesis for why there’s almost no color variation in crows and ravens — that the color itself wouldn’t be critical but mutations anywhere in the vicinity would be invariably fatal.

    So if crows and pigment then why not, say, robins and red breasts? Genes being genes, though, and mutation being generally fairly random, it ought to something close to a coin toss whether the gene linked to parasite resistance was the gene for color in males or color preference in females.

    And… hmm hadn’t thought about this before but… which linked trait do you think would be most likely to be directly conserved — parasite resistance in prospective males or in predisposed females?


  13. Michael

    I may have misread you, but it seems strange to conflate female choice with neutral evolution. Even if the reason for female choice is ultimately arbitrary there is still very real selection guiding the evolution of both the trait and preference for the trait.

    [CZ: Michael–Prum’s paper draws a parallel between neutral evolution and Fisher’s model of sexual selection as two null hypotheses against which adaptive scenarios must be tested. Female preference evolves in this scenario not because females get a benefit such as good genes, but only because their sons will also be more attractive.]

  14. So maybe this is totally obvious (which perhaps shouldn’t be surprizing given its source) but our perspective on sexual displays is one not extensively considered in the above discussion: that they signify gender. One can argue as to whether the specific form of the display more broadly signifies fitness, but because sexual behaviors are dimorphic it is clear that signification of gender provides an evolutionary advantage. In other words, red spots don’t just tell females which are the best males; they tell females which animals are males and not females. Those spots also tell males about the presence of other males. A side note is that we neuroethologists are more and more observing systems in which animals have the capacity to exhibit the sexual behaviors of both genders (see the high baseline wild-type female mounting rates in Catherine Dulac’s work on the TrpC2 mice, or the Giro Miesenbock Flypod data for example) – these observations may make a dialogue about who is which gender in a social encounter even more important.

  15. Michael


    But having more attractive sons is also a benefit, which was my point; Fisherian indicators are acted on by selection just as viability indicators. However if the comparison was just as two null hypotheses then that makes sense, I suppose I better look at the paper.

  16. Carl: Hi! You may remember me as the creator of the Facebook application Evarium. I’ve also done some actual research on sexual selection (e.g., Evolution 58:1414-1423). It’s worth noting that Fisher’s process lurks in the background even when other things are going on (“good genes”, costs of mating, etc.). Preferences of one sex for traits of the other lead to associations (linkage disequilibrium) between genes for the preferences and genes for the traits, hence the direct selection for the traits due to the preferences leads to indirect selection for the preferences via the associations – this is the essence of Fisher’s insight. When you look at population- and quantitative-genetical models of sexual selection, you always see terms representing Fisher’s process, just as in Kirkpatrick and Lande’s pioneering formulations. So Fisherian sexual selection is a null model in somewhat the same sense as things like the simple harmonic oscillator in physics.

  17. He did mean to write want “accepting” instead of “excepting,”, but I think that everyone got the point. Great read.

  18. It’s great to have chance to read and respond to this exciting discussion! Thanks to Carl, Marlene, and other commenters for your interest in the topic!

    So, what do we call the null model in sexual selection. The verbal idea first came from Fisher, but the idea was turned into real explicitly mathematical model of evolution by Lande and Kirkpatrick in the early 1980s. Here, I will call it the Fisher-Lande-Kirkpatrick null model, or FLK. In the paper it was just LK.

    One of main difficulties people havfe with the idea of the FLK mechanism as a null model in sexual selection is, “How could a null process so complicated and open ended and still be the null? Don’t null models imply simplicity? Isn’t the Fisher process anything but simple?” I think the key is that the Fisher-Lande-Kirkpatrick mechanism is the complete description of the intrinsic consequences of the existence of genetic variation for trait and preferences with nothing else going on. If you add natural selection on male display trait, it changes the details of the equilibrium, but not the general process. So, natural selection on trait is okay, and within the null model.

    The FLK mechanism is really about the evolutionary consequences of the existence of mating preferences. It examines the results of genetic variation selecting on other genetic variation is absence of any reference to the outside environment. So, that’s why I don’t (and Darwin didin’t, I think) call this natural selection. Natural selection requires some correlation to the outer world. In the FLK null, fitness is entirely defined by the frequency of corresonding preferences in the population. This is why much of intersexual selection evolves so dynamically like art or fashion, which have no function except in the minds of observers.

    This null is complicated because the consequences of the simplist possible process in which genetic variation selects on other genetic variation in absence of other forces is intrinsically complicated. This is the core of Fisher’s observation.

    I don’t think it is an accident that Fisher proposed this idea just 8 years after Hardy articulated the Hardy-Weinberg Law (the only Law in Biology?) which is the most basic null model in genetics. It is also not an accident that Fisher coined the term “null hypothesis,” etc. He was interested in looking at the simplest mechanisms that could describe evolutionary dynamics.

    Behavioral ecologists have a tradition of hating Fisher’s idea as a mathematical notion of no consequence to nature. Now, I have spent enough time in tropical forests of several continents watching and describing polynyous lek behavior to earn my stripes as a behavioral biologist. And Fisher’s idea always seemed like an excellent match to Nature as I saw it. Why this huge disconnect?

    It come down, as Marlene Zuk points out, to the costs of mate choice, or whether female preferences are under natural selection to get better mates. It is clear that mate choice can be costly in general, either directly in terms of the costs of searching for a mate or the sexual tranmitted diseases that could be acquired from the wrong mate, or indirectly in terms of the genetic quality or genetic compatibility of a mate. But natural selection doesn’t just require general costs, it requires differenital costs. That is, a preference for one display has to be correlated with higher fitness than a preference for another display. For example, the male Yellow Warbler sings –sweet-sweet-sweet-sweeter-than-sweet, whereas the closely related Chesnut-sided Warbler sings –sweet-sweet-sweet-I’ll-switch-you. Imagine now that in order for such differences to evolve as honest advertisements of quality, female Yellow Warblers have to have higher fitness if they prefer the former, whereas female Chesnut-sided Warblers have to have higher fitness because the prefer the latter. Both songs have lots of features– like complexity, and requisite skill– that might generally be hypothesized to indicate condition or fitness. But that doesn’t explain the differences. For the hypothesis to work, the specific features of one song must yield higher fitness for females preferring it, and vice versa. What would those differences even be? To me the same questions apply to the specifics of the myriads of spots, dots, and stripes in bird plumage or butterfly wings.

    Has anyone every found a case to support this kind of differential success of heritable variations in mating preference? Not really. Turns out that biologists most often use the concept of general costs to assume the existence of differenital costs. This main assumption is rarely tested. So, the whole of honest signaling literature rests on the concept of general costs to mate choice rather than on evidence of differential costs of different preferences.

    The lack of evidence for natural selection on preference is a big problem given that the null hypothesis can easily explain these warbler songs, and just about any other display trait without it. Many biologists will see the ease of success of the null as a problem with the null. To me, this view comes from a failure to grasp the complexity of the evolutionary process in which fitness depends only on what’s going on in the brains of other individuals. Biologists will also counter that asking for evidence of natural selection on preferences is asking too much. I would counter by quoting from a classic Onion article, “Science is hard.” It may feel easier to merely confirm our favorite impressions about how nature evolves, but it’s not the same as doing science.

    The comments by Figleaf raise a second big issue which is that the Fisher-Lande-Kirkpatrick mechanism is not about the origin of mating preferences but the evolutionary consequences of the genetic variation in mating preference. Fisher did make some comments on the question, but that was just to set up the issue he was interested in- the coevolutionary feedback between display and preference. To me, this is not a problem for the model. The Hardy-Weinberg doesn’t address the origin of genetic variation either, and is not routinely criticized for it. Every evolutionary model has its initial assumptions and we are used to working with them. In my research, I am really interested in the origin of evolutionary novelties (like feathers). But the evolution of mating preferences, like the evolution of sex, is just not a question that needs to be addressed in every sexual selection study. For example, if I am working on the breeding biology of the Chesnut-sided Warbler, no one will ask, “But how does your study explain the origin of sex in this species?” But, for some reason, it is frequent to ask the question, “If you think this display trait is arbitrary, how do you explain the origin of mating preference in this species?” Why is that? I think it’s because most biologists haven’t realized the origin of preferences and the evolutionary consequences of preferences are different fields of research.

    In the paper, I use phylogenetic evidence to argue that the ubiquitous mating preferences of birds originated before the break up of Gondwanaland. They are REALLY old. For example, they are certainly older than distinguishable, typical ZZ-ZW avian sex chromosomes, which are absent in ratites (like ostrich) and tinamous which have obvious mating preferences. So, if mating preferences are older than many details of sex determination itself in birds, why do researchers expect research on sexual selection in a single species to address this point?

    I thin it is because, as with sex itself, natural selection is required to explain the origin of mating preferences. Since most sexual selection researchers are assuming natural selection on mating preferences, they see there work as somehow addressing the more fundamental question of the origin of preferences themselves. But just like the origin of sex, I think the origin of mating preferences should be a separate literature of little overlap with questions in current biodiversity, addressed in separate experimental and theoretical literature. Like the evolution of sex, it actually might best be addressed experimentally in microbes with chemical displays and chemosensory preferences.



  19. Isn’t it fairly well established by now that the histocompatibility complex (and specifically, complementary HLA genes) play a significant role in sexual attraction in humans (and possibly other vertebrates?; and not solely through olfaction)? I would think this might be a more productive line of study in birds as well, and at a more fundamental level (has anything been done???), than the all-too-easy, after-the-fact theorizing about highly-subjective visual or physical cues, which humans may not perceive/interpret the same as birds do anyway. To the extent that those visual/physical cues can be directly linked to molecular immune-function genetics (of male & female) seems like we’d be on firmer ground.

  20. The other Jim

    To Shecky Riemann,

    I don’t think the MHC story is the slam-dunk that many seem to believe it is. There are still credible doubts as to whether this is a real mechanism.

    ex. Derti A, Cenik C, Kraft P, Roth FP (2010) Absence of Evidence for MHC–Dependent Mate Selection within HapMap Populations. PLoS Genet 6(4): e1000925.

    AND – just to concur that complicated nulls are probably quite common in all of Biology. When you look at a species, it is an evolved system. Assuming simple random distribution-type nulls is not appropriate. For example;

    Lusk RW, Eisen MB (2010) Evolutionary Mirages: Selection on Binding Site Composition Creates the Illusion of Conserved Grammars in Drosophila Enhancers. PLoS Genet 6(1): e1000829. doi:10.1371/journal.pgen.1000829

    (Both articles are free via Open Access).

  21. GMBurns

    I have what I see as a strong objection to what I call the ‘pointless preference hypothesis’, which I have run into before. It forgets that the females of the species are also subject to life’s perils, be they from predation, failure to find sustenance, or inferior reproductivity, and so those who waste whatever ability to enforce ‘preference’ they have on something truly pointless will tend to do less well than those whose preference actually does signify something useful, even if what the preference is tied to may be hard to discover (I am still studying the Claudia Schiffer/David Copperfield case, with him really having magic being the null hypothesis).
    In the meantime I remind everyone, females are under every bit as much evolutionary pressure as males, and arguably more than males are because females are also usually more heavily burdened by offspring. If a given trait is truly useless, the poor chooser will eventually be replaced in the gene pool by females who go after something more useful. It feels odd to say it, but it strikes me that the fallacy behind this hypothesis may be especially difficult to see for those of us relegated to geekdom…to us female preference is likely to seem a deeply mysterious thing, and surely the poor goddesses must be irrational herd animals somehow gifted with an ability to promise or deny us all happiness with no need to worry about survival themselves.

    Fat chance.

  22. amphiox

    if there are some females find a red dot male attractive, doesn’t it make sense that roughly the same proportions of females will find the red dot appalling?

    It seems to me that such a scenario may result in the population being split in twain: males with red dots and females that prefer red dots will mate more and more with each other, producing offspring that inherit both the red dot and the preference for it, while males without red dots and females that are repulsed by the red dot will mate more and more with each other, producing offspring that do not have red dots and hate red dots.

    Taken to its logical extreme, we could get sympatric speciation into a red dot/red dot loving species, and a dotless/red dot hating species which never mate with one another thanks to the preference/aversion for red dotted males.

    If a given trait is truly useless, the poor chooser will eventually be replaced in the gene pool by females who go after something more useful.

    I think the likelihood of this happening depends on whether the initial variation in preference actually includes preference for useful traits, and how useful those traits are. If for example the useful trait is only slightly useful, and the useless trait truly neutral (no detriment), then genetic drift could by chance alone give the preference for the neutral trait a slight upper hand over the preference for the useful trait. As I understand the argument, once that upper hand is achieved, a runaway, positive feedback loop occurs in favor of the arbitrary trait preference that rapidly changes the trait from being neutral to being highly, highly favorable – favorable in that it is attractive to the majority of the opposite sex! And so favorable that it outweighs the advantage of preferring the useful trait, which is now in the minority and becomes saddled with the disadvantage that those preferring it will have a harder time finding rarer mates.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

The Loom

A blog about life, past and future. Written by DISCOVER contributing editor and columnist Carl Zimmer.

About Carl Zimmer

Carl Zimmer writes about science regularly for The New York Times and magazines such as DISCOVER, which also hosts his blog, The LoomHe is the author of 12 books, the most recent of which is Science Ink: Tattoos of the Science Obsessed.


See More

Collapse bottom bar