More Recent Comments

Wednesday, January 14, 2015

How did the zebra get its stripes? (again)

The National Geographic has just posted an article on Why Do Zebras Have Stripes? New Study Makes Temperature Connection. Here's part of it ...
A leopard may not be able to change its spots, but some zebras change their stripes. Zebras in warmer places have more stripes, a new study shows, which might help answer an age-old question: Why stripes?

The answer probably comes down to keeping zebras cool and fending off disease-causing insects that are more common in hotter climates, researchers reported Tuesday in the journal Royal Society Open Science.

All three species of zebra have bold black and white stripes that stand out among more drab-looking African grazers, like buffalo and antelope, especially against a plain savanna background. And standing out would seem to make a zebra more likely to become a lion's lunch.

This "stripe riddle" has puzzled scientists, including Darwin, for over a century. There are five main hypotheses for why zebras have the stripes: to repel insects, to provide camouflage through some optical illusion, to confuse predators, to reduce body temperature, or to help the animals recognize each other.
There's a fifth possibility: maybe there's no reason at all and stripes are just an evolutionary accident.

The last time I addressed this issue was in 2012 [How Did the Zebra Get Its Stripes?]. At that time I quoted the famous Spandrel's paper where Gould and Lewontin wrote ...
... the rejection of one adaptive story usually leads to its replacement by another, rather than to a suspicion that a different kind of explanation might be required. Since the range of adaptive stories is as wide as our minds are fertile, new stones can always be postulated. And if a story is not immediately available, one can always plead temporary ignorance and trust that it will be forthcoming ....
That's exactly what's happening in the National Geographic article. After almost 100 years of speculation, nobody has come up with a good adaptive explanation of zebra stripes. They never consider the possibility that there may NOT be an adaptive explanation.

Stephen Jay Gould wrote about this problem more than 30 years ago [How the Zebra Gets Its Stripes]. He pointed out back then that stripes are almost certainly due to small changes in just a few genes that alter the timing of differentiation in early embryology. He rails against adaptationist thinking then says ...
For many reasons, ranging from probable neutrality of much genetic variation to the nonadaptive nature of many evolutionary trends, this strict construction [vulgar Darinwism] is breaking down, and themes of unity are receiving attention.... One old and promising theme emphasizes the correlate effects of changes in the timing of events in embryonic development. A small change in timing, perhaps the result of a minor genetic modification, may have profound effects on a suite of adult characters if the change occurs early in embryology and its effects accumulate thereafter.
The point is that the prominence of stripes on zebras may be due to a relatively minor mutation and may be nonadaptive. That's a view that should at least be considered even if you don't think it's correct.

Gould was being very optimistic when he suggested that the old ways were breaking down.

Gould, S.J., and Lewontin, R.C. (1979) The Spandrels of San Marco and the Panglossian Paradigm: A Critique of the Adaptationist Programme. Proc. R. Soc. Lon. B Biol. Sci. 205: 581-598. [PubMed] [doi: 10.1098/rspb.1979.0086]


  1. The last time I addressed this issue was in 2012...

    Not quite the last time. It came up again last April (and provoked some interesting discussion):

  2. This is an amazingly poor hook on which to hitch a lesson on genetic drift. In the midst of a sea of predators and insect parasites, a property as eye-catching as a zebra's stripes is almost certainly not neutral enough to be common due to drift. It boggles my mind that anyone would suggest this. The strength on selection of coat color appears to be strong in most animals, either for camouflage or as signals. While some aspects (especially choice of signal color) might be due to drift, the major themes are clearly not. They do not show the amount of random variation within species that one would expect if they were truly neutral. It's just like in genes--strongly conserved phenotypes are unlikely to be due to drift.

    1. It boggles my mind that anyone would suggest this.

      That's exactly why I posted it. Your mind needs to be boggled.

    2. The strength on selection of coat color appears to be strong in most animals, either for camouflage or as signals.

      And then there are wolves and foxes, where it seems to be a side effect of selection for tameness, i.e., a spandrel (at least in that context).

    3. @judmarc - Aha, but can you be sure? 'Just-ain't-so' stories have the same difficulty as 'just-so' stories - speculative assessments of the s value of traits and their underlying causal relationships with survival, plus the confounding issues of epistasis and hitchhiking. Presumably there was no control against (possibly unconscious) bias by the breeder.

    4. ... or the relaxation of selection for the wild-type coat colour in a domesticated setting.

    5. There are studies of the strength of natural selection for coat color matching substrate in rodents. It is strong. As far as I know, zebras have (or had) fairly large populations, meaning even small values of s will be effective. And zebras show little intra-specific variation in color, again contradicting the neutral argument. Now, if you said that the placement and exact number of spots on baby deer or baby tapir were determined by drift, I'd not be "boggled" by such a claim. But predation and parasites are important, and an animal's color pattern definitely has some interaction with both of those factors.

      I think only someone who has not experienced the strikingly fine-tuned coat color of most mammals in nature would suggest that a zebra's dramatic coat color is close enough to neutral to be susceptible to fixation by drift. Color me boggled. LJ

    6. By the way I am not a skeptic about drift, nor an "adaptationist". My main publication in genetics is partly about quantifying the expected amount of genetic differentiation between subpopulations due to drift. I spend a lot of effort to quantify the role of drift in speciation. But just as some people are ignorant about drift, some people who haven't spent much time observing nature closely might be too quick to jump on the drift bandwagon. In most organisms in nature, color is very finely tuned to the environment. The bulk of the evidence suggests adaptation as the reasonable first guess when looking for explanations of skin color in diurnal animals that have strong interactions with predators and parasites and each other.
      L Jost

    7. The probable role of adaptation in coat colour is particularly striking in the case of white coloration in Arctic environments. There are too many different species converging on the same result for this to be seriously considered non-adaptive. Arctic foxes, ermine, ptarmigan and hares have an almost complete switch between winter white and summer drab. Caribou go half-and-half, polar bears stick to white all year, certain seals are white as young and darken later; beluga whales go the opposite way. Non-adaptive forces can't be ruled out explicitly in every case, but the correlation is undeniably strong, while the variations in the role of whiteness (including those Arctic species lacking it) do appear to bear a significant relationship to life history in every case.

      The causal role of whiteness may actually vary - camouflage and thermal efficiency seem to be the main contenders, though they aren't mutually exclusive. But what doesn't seem at all likely is 'random chance'.

    8. @Allan Miller - Now that you mentioned hitchhiking, that may be what was going on in the Soviet/Russian fox experiments. Since I have just seen a little bit of television on these, I don't know what controls were in place re human bias as to coat color (as well as tail conformation, which also appeared to change when tameness was selected for).

    9. White cats tend to be deaf. ( Is there any serious argument that this is an adaptation?

    10. Unknown says,

      In most organisms in nature, color is very finely tuned to the environment.

      There are enough exceptions to this rule that you would be wrong to assume that it applies to every species you encounter.

    11. White cats tend to be deaf. ( Is there any serious argument that this is an adaptation?

      Which - being white or being deaf? This may be a pleiotropy. Not sure of your point, though. White fur is not argued a 'universal adaptation', particularly indoors! Polymorphy, and the individuals being domesticated and highly inbred (compared to typical wild populations) also tend to argue against adaptation.

    12. The probable role of adaptation in coat colour is particularly striking in the case of white coloration in Arctic environments. There are too many different species converging on the same result for this to be seriously considered non-adaptive.
      Yes, multiple species converging on the same phenotype in the same environment is strong evidence in favour of adaptation. Such evidence seems to be lacking in the striped Zebra case.

    13. The point is pleiotropy is kind of the rule in reality and not the exception that it is generally presented as. Which kind of complicates a lot of adaptationist arguments which act as if genes encoded unique Mendelian traits rather than encoding molecules affecting many different traits. Even if a gene was selected for on a basis of a phenotype, how do you know the phenotype you are looking at (eg. zebra stripes) is the phenotype giving the selective advantage? If coat color is linked to hearing in cats, why not in other animals?

    14. I don't think pleiotropy causes a particular problem for adaptationism. The s value of a stretch of DNA is an integration of all its selective effects, naturally mediated via phenotype. Even a single-effect gene may have both positive and negative results at different times - eg something that makes one fluffy is great in winter, less so in summer. So we have a spectrum of phenotypic interactions, from epistasis to pleiotropy, and a temporal dimension, but it all resolves to the net effect of 'DNA-stretch x' upon its own survival/replication.

      It is certainly possible that one is mistaken about the causal link. If 'deaf+white' has a positive s value, one might be mistaken if one thought that it was deafness that gave the sequence the edge. But, regardless, this is adaptation, because s (in this example) is positive. We were just wrong about the cause.

      This does somewhat play back to Simon Gunkel's contention that separation of the causal elements in an allele's spread, when that is essentially due to a stochastic sampling process, is artificial.

      We are naturally interested in causality, and drift acting alone is certainly a cause. Hitchhiking is another. But my own contention would be that phenotypically visible fixed traits are on balance more likely to be due to adaptation than to drift, because even a tiny difference in the exponent (s) can make a massive arithmetic difference to fixation time and probability. One would need a substantial excess of mutations in the neutral zone to counter this tendency, granted that some are bound to get through. While this excess probably occurs in junk and silent sites, I am less convinced about the phenotypically visible portion.

    15. Don Cates, Such evidence seems to be lacking in the striped Zebra case.

      That's true enough. But one would (naively, I admit) expect a cost to being so visible, and hence an offsetting benefit. Far from being camouflage, zebra stripes make them stick out like a sore thumb. But perhaps this is less of an issue for a herd animal. Musk oxen, for comparison, are an exception to the 'white Arctic' rule, and caribou partially so.

      Nonetheless, my main point is that the burden of proof is not upon the adaptationists alone. Stripe genes have a net s taking some value. One can call s=zero the null hypothesis, but it remains a positive claim about s.

    16. Allan Miller says,

      my main point is that the burden of proof is not upon the adaptationists alone...

      Actually it is, and that's why so much effort has been spent on trying to find out if zebra stripes confer a benefit. However, my main point was that the article didn't even mention the possibility that there might not be an adaptive explanation. Surely you will agree that this is not good science?

    17. Actually it is

      So it must be shown that s<>0, and why, otherwise s can be assumed to be 0? Surely the sign of s, or lack of one, must remain an open question? Certainly I take the point that adaptation is very much a default assumption for significant traits, while 'none' is a viable possibility.

      I believe Larison is looking for the genes, which may shed more light on the question.

  3. Has any one ever actually bothered trying to find out which genes actually determine this pattern and look for evidence of selection?

    1. From a 2005 paper by Leonard et al an opposite theory is proposed, why did Quagga lose their stripes compared to zebra's.
      They conclude:
      "Existing plains zebras show a geographical gradient in coloration with progressive reduction in striping from north to south, which has been explained as an adaptation to open country and for which the quagga represented the extreme limit of the trend".

      This doesn't solve the 'why do zebra's have stripes' question, which indeed could be just a fluke of nature. But, there seems to be a trend of progressive reduction in striping from north to south, which might be selection at work.

  4. Gould did ask the same question about the giraffe's neck.

  5. I had only read the article on the stripes just before coming to Sandwalk today, and it seemed a stretch to suggest that microcurrents of air would provide added cooling and drive evolution. Nice review

  6. Dr. Moran,
    In this, and the previous post you have held S. J. Gould in high regard. At least we have found something in common. Gould is definitely my favorite evolutionary biologist -- mostly because he was quick to question the box, quick to suggest that mainstream Darwinian theorists were wrong. I wish I had the available time to read much more of what Gould had to say.

  7. They should ask creationists.
    lets think about this.
    Horses are only post flood adaptions from some earlier thing.
    originally they would of lived in a very rich predator society. nOt the poverty of Africa for the last thousands of years.
    Horses are creatures that powerfully escape by running. Thats the point. They must outrun. In those days horses , exclusively, moved in herds.
    The colours are not to hide them from being sighted as a herd. Rather, like white tailed deer and many more, the colour confuses the predator up close. They can't segregate to pinpoint. or rather the margins help protect the zebra.
    the stripes really are for a good reason. If they were a accident surely selection would of changed them a little and surely surely more then that. Is selection a myth/ Did no more mutations occur in zebras?
    I see the stripes as coming from the same mechanism as all colours in creatures for survival.
    What that is IS another question.

    1. Byers is much more Darwinist than Larry.

    2. Robert Byers, please understand that in these parts the term "creationist" is used in a much broader sense than that of young earthers. Your argument only addresses the young earth perspective.

  8. It seems to me that a corollary of the claim that stripes aren't under selection would be the existence of developmental constraints that prevent stripes from being lost easily. This would be supported by the observation that loss of stripes is exceedingly rare or nonexistent.
    Ed's comment above certainly supports adaptation. Its too bad the loci determining stripes are not known - one could see if there have been selective sweeps in the quagga

    1. Yes I've been thinking the same thing. If the stripes themselves are not adaptive, there must still be some underlying reason for their persistence. (Of course here there's a question about how long the Zebra and it's ancestors have had these stripes, and to what extend they have changed over time. If neither the stripes, nor the underlying genetic loci are under any selection, you'd expect them to eventually disappear due to drift.)

      On the other hand even if the stripes aren't adaptive, they might still result from loci that are maintained by purifying selection, because these in turn do contribute to other traits which are adaptive.

    2. There are also phylogenetic questions concerning the clade that contains zebras and asses:

      It seems that stripes were present in the MRCA but have disappeared more than once in different lineages, or perhaps evolved several times independently. The formes seems a priori more likely.

    3. Agreed. It would go even further and say it would seem rather absurd that you could select for the emergence of stripes unless it was some kind of mutation that happens very frequently.
      It also seems a simpler explanation for the diverse "types" of stripes that they share a common ancestor and have simply been diverging through drift since then, than to postulate unique and small variations in selective pressures for all of them.

    4. Why would you expect the stripes to disappear due to drift if they are neutral? If a single mutation could eliminate the stripes, how often would it occur in the population and what is the probability it would drift to fixation?

    5. It is possible to select against stripes - see 'project quagga'. Therefore, to this degree at least, the necessary genetic variation already exists. p(fixation) depends on the population size and the number of different 'stripe-diminishing' alleles that actually exist - it's probable there are a number of ways of causing the same result, though probably fewer with stripe-reduction as their sole effect.

  9. I was intrigued at Piotr's suggestion that zebras are polyphyletic which would suggest to me convergent evolution hinting (to my mind at least) greater support for adaptation than Neutrality. (forgive my bias)

    However, it would appear according to a variety of sources that zebras are in fact monophyletic

    No matter, I am reminded of how the military co-opted biological insight during WW II.

    From wikipedia:

    The British zoologist John Graham Kerr, who first applied dazzle camouflage to British warships in WWI, outlined the principle in a letter to Winston Churchill in 1914 explaining that disruptive camouflage sought to confuse, not to conceal, "It is essential to break up the regularity of outline and this can be easily effected by strongly contrasting shades ... a giraffe or zebra or jaguar looks extraordinarily conspicuous in a museum but in nature, especially when moving, is wonderfully difficult to pick up."[1]

    While dazzle did not conceal a ship, it made it difficult for the enemy to estimate its type, size, speed, and heading. The idea was to disrupt the visual rangefinders used for naval artillery. Its purpose was confusion rather than concealment.[2] An observer would find it difficult to know exactly whether the stern or the bow was in view; and it would be equally difficult to estimate whether the observed vessel was moving towards or away from the observer's position.[3]

    So are zebra stripes as camouflage provide an advantage? The answer is an unequivocal yes.

    Unless I am missing something: "score 1" for the "adaptationists".

    1. From the same wiki article: "Dazzle was adopted by the British Admiralty and the U.S. Navy with little evaluation. Each ship's dazzle pattern was unique to avoid making classes of ships instantly recognisable to the enemy. The result was that a profusion of dazzle schemes were tried, and the evidence for their success was at best mixed."
      So before giving the 'adaptationists' any points I would like to see some evidence of differential reproductive success.

    2. @ Don Cates

      " So before giving the 'adaptationists' any points I would like to see some evidence of differential reproductive success.

      Good point - if dazzle camouflage interferes with estimates of prey type, size, speed, and heading, google-whacking should come up with some hard data.

      This is all I could come up with:

      But let's say the military tested that dazzle camouflage is effective in military settings, can we not extrapolate back to the zebra?

    3. Nothing new in that link. Did you not notice " the evidence for their success was at best mixed." I quoted from your other link? Even if it did work, it would be difficult to extrapolate back to zebras since they used a different pattern on different ships so that the 'dazzle' wouldn't be a recognition pattern.