Did birds evolve larger beaks to eat larger invasive snails? (The NYT gets it wrong)

When an animal population changes over time for some trait, there are several possible causes. For instance, the animal could be responding nongenetically to environmental changes, just as a cat grows a longer coat in winter (this is within a single generation, but could also apply if cats move to colder climes). This is called “phenotypic plasticity”. Such changes aren’t considered evolutionary change because there are no changes in the frequencies of different gene forms, just in the expression of genes caused by the environment. Evolution, after all, is usually construed as genetic change over time.

Alternatively, the changes over time could reflect genetic change and thus instantiate genuine evolution. This has been seen many times—even during a human lifetime. The most famous of these studies was Peter and Rosemary Grant’s work on the increase in beak size in a Galápagos finch when there was a drought in the late 1970s, but there are other studies as well (e.g., industrial melanism in the peppered moth and many other insects), the changes in mouthpart size in the soapberry bug (coincident with new, invasive plant species) described on pp. 134-135 of my book Why Evolution is True, and, of course, the many cases of response to human changes in the environment: DDT and organophosphate resistance in insects, antibiotic resistance in bacteria, herbicide resistance in crops, and so on.

Finally, there could be an interaction between or a combination of environmental plasticity and genetic response, so that both factors could cause a population to change over time.

The point is this: if we see a population at time A, and then it’s changed in some feature at a later time B, with the difference between A and B being at least one generation, we can’t automatically assume that the population has evolved. It could be simply a change in the trait due to effects of the environment on development. This mistake is not uncommon: a paper was once published in Science showing that lizards transplanted to tiny islands on which they had to climb trees had a big increase in limb length over a couple of generations. The paper was published in Science because this was taken to be a case of evolution in real time, like the Grants’ work (their work is not in question). When I read it, though, I saw they hadn’t done any tests to show the change was genetic. And, sure enough, it wasn’t: the authors later showed you could effect the same changes just by letting the lizards crawl around on artificial trees. Another example is the well known difference in height between North and South Koreans (3-8 cm., or 1.2-3.1 inches)—surely due to differences in nutrition and not evolutionary change since the 1950s!  For previous posts by Greg and me on this issue, see here and here.

 A new paper in Nature Ecology & Evolution by Christopher Catteau et al. (reference and free access below, free pdf here) reports changes in beak, body, and tarsus size in the snail kite (Rostrhamus socialbilis subspecies plumbeus) in Florida, a change that occurred after the area was invaded by a larger prey item: snail. Bigger birds can eat the larger invasive snails, so there could be natural selection favoring bigger birds with bigger beaks, for they would get more nutrition and possibly produce more offspring. But it’s also possible that the change (observed over 1-1.4 generations) was nongenetic: the new source of food could cause the birds to simply grow bigger, making their bodies larger, and that includes the tarsus and beak. Finally, the changes in size could be due to both factors.

The authors conclude that the main effect here is not evolution but phenotypic plasticity; as they say:

The increase in sizes of morphological traits was probably due to phenotypic plasticity based on three lines of evidence.

You can read that evidence itself; it’s a bit genetic-y and arcane for the general reader, but they’re right.

However, you wouldn’t know that conclusion from the paper’s blurb article (from the “Trilobites” column) in the New York Times (click on screenshot below):

First of all, the title is wrong: we are not at all sure that the birds evolved bigger beaks. In fact, the genetic evolution was tiny at best, and the authors of the paper admit that the bulk of the bird’s response in beak size was due to phenotypic plasticity. The author, Douglas Quenqua, alludes obliquely to a nongenetic cause in the passage below, but he still says that the body size change is an “evolutionary trick,” which clearly implies “evolutionary change” rather than a plastic response. After all, when your cat grows longer fur in a colder year than in a warmer year, it’s not pulling off an “evolutionary trick”. From Quenqua’s piece:

Dr. Fletcher and his colleagues analyzed 11 years of morphological data they had collected on the birds. Because snail kites can live to the relatively old age of 8, that time period represented fewer than two generations for the birds. Nonetheless, the researchers found that beak and body sizes had grown substantially (about 8 percent on average, and up to 12 percent) in the years since the invasion.

Exactly how the birds are pulling off this evolutionary trick is not clear, but natural selection does appear to play a part. Young snail kites with larger bills were more likely to survive their first year than snail kites with smaller bills, presumably because the large-billed birds were better able to eat the invasive snails.

. . . Young birds eating the invasive snails, which are two to five times larger than the native ones, were also growing faster than birds weaned on the smaller ones, which may account for the increase in overall body size.

But the researchers found suggestions of a genetic component to the changes, as well. By tracking the birds’ pedigrees, they found that large-beaked parents gave birth to large-beaked offspring, setting the stage for large-scale evolutionary change.

Here the NYT is simply guilty of not only overhyping its conclusions, but misleading readers. Quenqua absolutely fails to make a clear distinction between the two interacting causes of change in a trait over time.

Now a bit about the paper, but I’ll be brief. Here are the highlights:

  • In 2004 the large apple snail Pomacea maculata invaded a Florida lake and spread rapidly through the range of the snail kite (which eats almost exclusively snails and feeds them to their young). The big snail replaced the smaller native snail, P. paludosa. Here’s the difference in size:

From paper: The exotic apple snail (P. maculata; right) is a novel prey for snail kites, because it is much larger than the native congener (P. paludosa; left), leading to implications for foraging and demography.

  • The invasion of the larger snail led to a rapid increase in the population of the snail kite (from 700 to 2,000, still about half of its high point), almost certainly because there was a lot more food around. So in this case an invasion—usually considered undesirable by biologists–may have saved an endangered species.
  • Young with bigger beaks were more likely to survive their first year (this was basically a one-generation study), so there was a form of natural selection for large beaks. But for that selection (differential survival) to cause evolution of beak size there has to be genetic variation among individuals in beak size.
  • The problem was that the study showed very little genetic variation for beak size. Estimates of “heritability” (the proportion of variation in beak size among individuals that had a selectable genetic component) was rather small, and in some cases indistinguishable from zero.
  • Moreover, these heritability measurements appear to be based on the correlation between parents and offspring in beak size. Now that could be a valid estimate of heritability if there is no environmental factor causing a correlation, so that we can assume that the correlation reflects only genes passed to offspring from parents.  But is that a good assumption? Not in many species: for example, the traits that have the highest “heritabilities” in our species include wealth and religion, but these don’t have major genetic components—the correlation is an environmental one because parents give their religion and their dosh to their kids.  What about the snails? As far as I can see, there’s a possibility for such an environmental correlation here, too. If adult kites with bigger beaks manage to get more of the big invasive snail, and thus feed more of them to their young, their young could grow bigger because of food, not because (or just because) of their genes. That would lead to a spurious genetic correlation between parents and offspring, overestimating the genetic component of the trait’s variation—and thus its liability to respond to selection.
  • The offspring did get bigger in body size, beak, and tarsus, and beak size got bigger even taking into account its correlation with body size. So there clearly was change between generations.

The authors, as I noted, emphasize the environmental component as making the kites bigger within a generation, downplaying the genetics because of their heritability data. I gave one quote before, and here’s another from the paper (my emphasis):

The breeder’s equation and breeding values are often used to detect evolutionary change. By contrasting these methods to observed phenotypic change, we found that phenotypic plasticity, not evolution, is probably responsible for most of the morphological changes in the snail kite at this point in time.

We can’t rule out that some of the change in bill, tarsus, and body size was genetic, and thus true evolutionary change within a generation, but at best that change is trivial compared to the change due to plasticity. The authors are much more careful in their paper than is Douglas Quenqua in his column. I’m not sure whether Quenqua has any training in biology (his LinkedIn profile shows a bachelor’s degree from Adelphi University), but it doesn’t matter. What he needed to do in his piece was understand evolutionary biology, and you can do that without a Ph.D. in the field. And he failed in his attempt to communicate the main findings of the paper: the hegemony in this case of plasticity over genetic change in an important trait change. In fact, he not only failed to communicate it—he got it backwards.

The New York Times really needs to do better in its science reporting. The author of this piece should absorb the lesson I emphasized in caps in 2011:


h/t: Diana MacPherson, Greg Mayer


Cattau, C. E., R. J. Fletcher Jr, R. T. Kimball, C. W. Miller, and W. M. Kitchens. 2017. Rapid morphological change of a top predator with the invasion of a novel prey. Nature Ecology & Evolution. Published online 27 November 2017; doi:10.1038/s41559-017-0378-1 (pdf here).


  1. GBJames
    Posted November 30, 2017 at 11:41 am | Permalink


  2. Posted November 30, 2017 at 11:41 am | Permalink

    Is a species capacity for phenotypic plasticity itself a selectable trait?

    • Posted November 30, 2017 at 11:43 am | Permalink

      Certainly, and only in that sense can the changes in body size in the kites be said to reflect “evolution”. But that’s not what the NYT article implies.

  3. Posted November 30, 2017 at 11:47 am | Permalink

    This article is refreshingly easier for a non-biologist like me to read, understand, and appreciate, than some of your other more technical articles. I enjoyed reading it!

    • Diana MacPherson
      Posted November 30, 2017 at 4:40 pm | Permalink

      And glad Jerry corrected the NYt article that got it so wrong.

  4. Frank
    Posted November 30, 2017 at 11:50 am | Permalink

    Given the high frequency with which scientists are misquoted, it is possible that the author used the phrase “evolutionary trick” to refer to the capacity for (or degree of) phenotypic plasticity itself. Since the DEGREE of plasticity itself can evolve (usually portrayed as selection on the different reaction norms of different genotypes), maybe the author meant that the phenotypic response (even if it is purely phenotypic) is the result of the birds evolved ability to respond to environmental change – as in, they have a built-in “trick” up their sleeve? Who knows. Of course, none of this excuses the sloppy writing in claiming recent evolutionary change.

  5. ThyroidPlanet
    Posted November 30, 2017 at 12:08 pm | Permalink

    In case anyone’s memory was, like mine, jogged by hearing about lizards on islands :

    Lizards of Pod Mrcaru are an example of evolution “Before Our Very Eyes” (chapter 5 ‘stitle) in Richard Dawkins’ book “The Greatest Show on Earth”.

    Original literature citations are :

    Herrel, et. al. 2004 J. Evol. Biol. 17, 974-84
    Herrel, et. al. 2008 PNAS 105 4792-5

    • Austin Johnson
      Posted November 30, 2017 at 1:32 pm | Permalink

      Is this what is being referred to in the post? I know there’s been work done on Anolis lizards and limb length, so I wasn’t sure.

  6. Randall Schenck
    Posted November 30, 2017 at 12:11 pm | Permalink

    I wonder if it is not a symptom of problems in the newspaper business, maybe they cannot afford to staff the science journalist the way it should?

  7. Craw
    Posted November 30, 2017 at 12:18 pm | Permalink

    I will read this. Let me record my thanks for this.

  8. Liz
    Posted November 30, 2017 at 1:19 pm | Permalink

    This was very informative. I had to read it twice. It made me wonder how the difference in the genes is measured. Or how this: “changes in the frequencies of different gene forms” is measured.

    • Posted November 30, 2017 at 1:41 pm | Permalink

      Well, they didn’t actually measure genetic change; that is, they didn’t do any sequencing of DNA. They modeled it. I am not familiar with the techniques so can’t talk sensibly about them but if you scroll down to the methods section, you can see the part called “quantitative genetic analysis” for a description of the models they employed.

      • Posted November 30, 2017 at 3:45 pm | Permalink

        Yes, I saw that of course, but what isn’t clear is how they measured heritability and additive genetic variance, which are statistics that need to be entered into these formulae. The only way I can see is through parent/offspring correlations, which has the problems that I outlined above.

  9. JohnnieCanuck
    Posted November 30, 2017 at 11:48 pm | Permalink

    Cute tpyo, “If adult snails with bigger beaks […]”.

  10. Jim Ritchey
    Posted December 1, 2017 at 10:08 am | Permalink

    Would the Flynn Effect be an example of phenotypic plasticity?

  11. Posted January 24, 2018 at 4:38 pm | Permalink

    I am glad that I finally know how non-hereditary variation is called in English! (My books give it as “modificational variation” vs. “genotypic variation”.)

%d bloggers like this: