Niche construction: does it represent a “vastly neglected phenomenon” in evolutionary thought?

February 22, 2017 • 1:15 pm

“Niche construction” is a new term in evolutionary biology—indeed, a buzzword—although the idea has been around under other rubrics for years. It is the idea that the niche of an organism is not something static, imposed by its environment, but that the organism, as it evolves in behavior, morphology, and physiology, can change its environment in a way that changes how natural selection operates. In this way, its proponents say, there is feedback between the organism and its environment, mediated by the evolutionary process of natural selection.

The classic example is the beaver.  The ancestors of this creature presumably evolved to cut down trees and build dams and lodges, eventually creating lake environments that they did not have before. And living in that lake and in the lodge will affect what evolutionary changes are useful to the beaver, that is, how it will evolve (presumably to cope with a more lake-y existence).


But you can think of many, many examples. Any animal that builds a nest, a tunnel, or a hole to live in changes its environment, though not necessarily in a way that would affect future selection. Humans, through evolution, got big brains that allowed them to not only eat many more sweets and fats (thus creating selective pressure for genes protecting against diabetes and heart disease), but also invented medicine, therefore relaxing many of the forms of selection that previously killed us. The list goes on and on, and is uncontroversial. (I should add that some aspects of the environment are unlikely to be altered by the evolution of its inhabitants. The hydrodynamic properties of water don’t change when a fish evolves, and the color of snow doesn’t change when a polar bear evolves a white pelage.)

What is new is that a group of “new wave” evolutionists, most notably Kevin Laland at the University of St Andrews, claim that niche construction (henceforth “NC”) is an unrecognized factor in evolution, a very important one, and is an evolutionary process like natural selection.  This is part of the “extended evolutionary synthesis” (EES) that is, in my view, largely misguided, but is also funded to the tune of $11 million by the John Templeton Foundation—a grant given to about 50 investigators headed by Laland. I’ve written before about the various issues subsumed by the EES, including NC, and so won’t reprise my criticisms here. Instead, I’ll direct you to a a new manuscript on NC by Manan Gupta et al. (it’s on the preprint server bioRχiv, where you can download the pdf file; reference and link below).

The manuscript (I don’t know where it will be or has been submitted) makes several points about NC, and is critical of its use as a buzzword and of the claim that it’s an area that has been sorely neglected by evolutionists until now. Here are the paper’s main points:

a.) Basically, every evolutionary change can be seen as NC. Every evolutionary change in an organism ultimately came about because it improved reproduction (and sometimes survival), and those changes almost always alter the environment. Even an improvement in the ability to eat, for example, depletes resources faster, and thus changes the environment. Coevolution, in which members of different species affect each other’s evolution (parasites and hosts, predators and prey, pollinators and flowers, etc.), are well known: each change in one species affects the other species, and that change feeds back on the first species.

Darwin’s orchid, for instance, has a long nectar spur, forcing the moth pollinator to get really close to stick its tongue in for nectar, allowing the pollen to stick to the moth’s head. (This is adaptive for the flower, which “needs” to spread its genes.) That, in turn, drives an evolutionary increase in moth tongue length (to get the nectar more easily), and that drives an increase in the length of the nectar spur to make getting the nectar more difficult, creating continuing reciprocal selection that has led to the evolution of moths with long tongues and an orchid with a long tube (see below). Each organism’s evolution changes its “biotic” environment, thus affecting its future evolution, and that is NC:



Gupta et al. also cite experiments in flies showing that crowding vials with larvae led to a change in natural seletion, leading to the evolution of two types of larvae: those developing fast to get to the pupal stage first, and those developing later that could tolerate larval excrement better.  In humans, our evolved big brains led us to domesticate animals for milk, and that, in turn, gave a substantial reproductive advantage to those individuals who, unlike their ancestors, were able to get nutrients from the milk’s lactose by keeping the previously-inactivated enzyme lactase (not useful after weaning in ancestors) turned on after childhood. Our “pastoral” habits of rearing animals led, in those areas where humans kept milk animals, to an evolution of lactase genes that were permanently activated.  That, too, can be seen as a case of NC.

b.) NC has already been incorporated into evolutionary biology, not just in the examples above, but in theoretical work. NC can be modeled by a simple two-gene situation in which evolution at one gene affects evolution at the other, and there are many such models. Even if they don’t explicitly evoke environmental change as the result of evolution at one gene, that is what could be the result, and that environmental change could impose selection on the second locus. The added feature with NC is that there may be a time lag between the changes at gene 1 and at gene 2, as the environmental modifications produced by gene 1 may take some time to exert selection on gene 2. The condition whereby the “fitnesses” of variants at one gene depend on the fitnesses of variants at another is called epistasis. And epistasis has long been a feature of models in evolutionary genetics.

c.) NC is not a “process” like natural selection. Gupta et al. assert, and I agree, that NC is simply a result of natural selection that itself can constitute a selective pressure, but the main process itself remains natural selection. As their paper says:

NC affects the way in which selection acts. Its role is thus of a modifier which affects how a certain category of evolutionary process acts in a given instantiation, whereas selection has a very different logical or epistemic status as a specific category of process.

d.) The proponents of NC as a novel, unrecognized evolutionary “process” tend to make the same arguments over and over again in different papers. I agree; one sees a surfeit of assertions and a deficit of examples—at least examples showing that NC has been neglected.  This is also true for other claims about the EES, including the importance of “Lamarckian” evolution produced by environmentally-induced methylation of DNA, the primacy of evolutionary plasticity in evolution, the self-organizing properties of organisms, and so on. As Clara Peller said, “Where’s the beef?”

e.) The relentless pushing of NC as a neglected but important aspect of evolution reflects in part the careerism of investigators. I agree again. Evolutionists are not ambition-free, and we are human. In science, you make a name for yourself not by confirming what somebody else already found, but by suggesting and pushing new paradigms. This is especially true of evolution, a field in which new paradigms are rare because Darwin got so much right. That’s why “Darwin was wrong” claims make headlines in the popular science press. (The “neutral theory” of evolution, which assumes that many genetic variants have no differential effect on fitness, was a truly new paradigm.)  It’s not seemly to say this, but I don’t see why not. Regardless of someone’s motivations, their scientific ideas are always judged against nature. One can then ask informally, “Well, if NC has long been recognized under other names, and isn’t really a new process, why is it being touted as a really important unrecognized aspect of evolution?” I think the answer has to involve big ambitions.

But the authors push this part too hard, saying that that careerism is “an instantiation of academic niche construction.” That’s pretty funny, and partly true, but is not going to win over the reviewers of the paper, for one can criticize ideas in a manuscript without assessing the psychology of one’s opponents. I suspect this part will work against the acceptance of Gupta et al.’s paper. And Gupta et al. say this in rather harsh ways, as in the paper’s abstract (my emphasis):

In recent years, fairly far-reaching claims have been repeatedly made about how niche construction, the modification by organisms of their environment, and that of other organisms,  represents a vastly neglected phenomenon in ecological and evolutionary thought. The proponents of this view claim that the niche construction perspective greatly expands the scope of standard evolutionary theory and that niche construction deserves to be treated as a significant evolutionary process in its own right, almost at par with natural selection. Claims have also been advanced about how niche construction theory represents a substantial extension to, and re-orientation of, standard evolutionary theory, which is criticized as being narrowly gene-centric and ignoring the rich complexity and reciprocity of organism-environment interactions. We  examine these claims in some detail and show that they do not stand up to scrutiny. We suggest that the manner in which niche construction theory is sought to be pushed in the literature is better viewed as an exercise in academic niche construction whereby, through incessant repetition of largely untenable claims, and the deployment of rhetorically appealing but logically dubious analogies, a receptive climate for a certain sub-discipline is sought to be manufactured within the scientific community. We see this as an unfortunate, but perhaps inevitable, nascent post-truth tendency within science.

That last bit, especially the “post-truth” characterization, is unnecessary and a bit mean. Were I a reviewer of this paper, I’d recommend acceptance, but after some revision, including removal (or toning down) of the psychologizing!


Gupta, M., N. G. Prasad, S. Dey, A. Joshi, and T. N. C. Vidya. 2017. Niche construction in evolutionary theory: the construction of an academic niche? bioRxiv.

55 thoughts on “Niche construction: does it represent a “vastly neglected phenomenon” in evolutionary thought?

  1. I might be missing something obvious…but what, if anything significant, is the difference between “niche construction” and Richard’s extended phenotype?



        1. The Baldwin effect is a special case where organisms learn to live in their environment, thereby effecting their evolution. I suppose it also means there is selection for learning in species that can do that.
          Richard’s extended phenotype is like NC, only his emphasis is very gene-centric. That is, genes act through their proteins to effect an organisms’ phenotype, and that phenotype effects the environment that the organism lives in.

          1. I am a little lost. I would have thought that all organisms ‘learn’ to live in their environment, else there would be no ancestor’s tale for their progeny. Can you clarify?

            1. By learning, I took it to mean that they modify behaviors based on experience. So this characteristic would only apply to organisms that have a brain.

      1. Me too. The dam is part of the beaver’s extended phenotype, and part of ours too.

        It seems quite likely that literacy will have an effect — up or down! — on reproduction. I don’t think this is a new idea.

    1. I remember Dawkins writing a paper on just that in the early 80s or 90s, about niche construction compared to the extended phenotype, I think you can find it for free online by googling those words and Dawkins’ name.

    2. That would have been my comment as well. Dawkins clearly discussed extensions and feedback. I think, ever 20 or 30 years ideas leap from the past under a new label and dance around excitedly for a while. Only people who are old enough to remember and willing to throw a bucket of cold water can get folks to curb their enthusiasm and return to sobriety.

    3. Along with the rest of you, I saw the picture of the beaver, read some text and thought “extended phenotype”. I need someone who understands these things better than I to explain the difference. Jerry, can you draw a line here, or are the terms synonymous?

      1. The moment I read “feedback between the organism and its environment”, I already thought of Dawkins’ EP conjecture. The beaver is just the icing on the cake.

    4. Based on “Ecosystem engineering, experiment, and evolution (2011)” paper by Trevor Pearce, only some results of niche construction are
      extended phenotypes:

      “However, despite Dawkins’s claim that niche construction is the special case,
      there are many results of niche construction that cannot be seen as extended
      phenotypes. For something to count as an extended phenotype it has to be an
      adaptation ‘‘for the benefit of replicators,’’ i.e., there must be a causal connection between replicator-variants and phenotype-variants that advantages those replicators associated with certain phenotypes (Dawkins 2004, 379). Since in many cases of niche construction this is not explicitly assessed (and such assessment is difficult at the best of times), the result of niche construction, even when positive, is usually only a putative extended phenotype.”

      I recommend this article to anyone interested in the relationship between ecosystem engineering, niche construction, and extended phenotype concepts. Basically niche construction is nested concept within ecosystem engineering defined as modification of the physical environment by organisms. Based on Google Scholar citations, ecosystem engineering is much more influential concept than niche construction.

    5. I was also reminded of Dawkins’ book—just as mind-blowing as The Selfish Gene. I can almost hear Dawkins now: “Genes for building dams…”

      And I think PCC hit the nail on the head about careerism and the need to push paradigms. But that invites the question: where do evolutionary biologists really have to push? Cosmologists still don’t know what dark energy or dark matter are, and astronomers have tons of stuff still to figure out (I can’t wait for the James Webb Space Telescope to go up!). But, post-Darwin and the synthesizers of the mid-20th century, which Dawkins capped, what do evolutionary biologists have to do besides fill in the details? Not that those details aren’t fascinating, but… it seems kind of a bummer.

  2. So what is the Templeton Foundation’s interest in all this? I don’t see where any of this opens the door to “the hand of God” guiding anything (and it seems to me that the Templeton Foundation is only really interested in God).

      1. If so, it’s a spectacular own goal, as they’re just demonstrating themselves to be utterly ignorant of the seminal technical work of the most famous living evolutionary biologist who, oh-by-the-way, just happens to also be the best-known critic of theism.

        What’s next? Templeton funding “groundbreaking” and “revolutionary” research into so-called “dark superstars” they theorize might form in the presence of sufficient densities of matter and which call into question the very foundations of modern cosmology?



        1. Ben, have you been watching TV again? Naughty. I caught the dark superstars bit on a recent episode of “How the Universe Works” on The Science Channel. Perhaps you have a different source?

          As you say: “Cheers”.

      1. This is why I have mixed thoughts about neutral evolution, and regard it as mostly a “background noise” creating methodological hurdles to studies of positively selected beneficial mutations. (I am talking about anagenesis; neutral evolution may be very important for speciation.)

    1. I think it is to make the JTF look like they are sincere in their desire to broaden the scope of evolutionary studies. Into this bigger tent they can then slip in some numinous bits.

    2. It seems to me that the language implies direction, intent, so teleology to TF ears, and if evolution means “more lakes” to a beaver, a TF might see “more spirituality” as part of the.a human CN.

    3. All the above; and also that anything that seems to undermine the gene-centred view of evolution opens a crack in which to insert whatever version of supernature you prefer.

    4. NC and the Extended Synthesis in general tries to imply that organisms have agency in guiding their own evolution. I think that appeals to the Templeton foundation.

  3. That a population of organisms affects its environment and that its environment produces selection pressures on the population of organisms seems damn obvious to me. So damn obvious that I don’t find it plausible that modern biology hasn’t realized that until just now.

  4. Academic niche construction is of course a well known phenomenon in the “soft” sciences, where it is of course perfectly viable since it deals with the complexity of human thought.

    I do wonder if there are similar examples in physics or chemistry.

    1. Probably much less, since the impact on ‘spirit and soul’ is much less direct. Which means less funding from JTF and the like, I’d think.

  5. The recent article in National Geo. on Booze, over the last 9000 years would be NC as well? The monkeys came down from the trees to get the fruit that fell to the ground. Fermented just a bit but genetic change allowed breaking down methanol at 40 times the previous rate. Early humans grew more crops, not to eat but to make booze and the beat goes on.

  6. Didn’t Darwin himself dedicate a lot of time to research how earthworms pretty much create fertile soil? (And that in turn has evolutionary implications for other species). How is this news?

    If I understand correctly, the idea is that individuals aren’t being passively modeled by the environment like sculptures, but rather they interact and that everybody is part of everybody else’s environment. Alright this is interesting, but hardly revolutionary. I expect it’s pretty well established at the real science level.

    1. That is true. The concept was first formalized in 1994. by Jones el al., in “Organisms as ecosystem engineers”. As of today, that paper has 4280 citations on Google Scholar, so yeah, it is pretty well established. Niche construction concept developed independently few years later and it is very similar to the ecosystem engineering concept. Basically, they are competing concepts to explain same/similar phenomena in nature, and so far ecosystem engineering is much more established. So maybe there lies the motive behind such a push of niche construction as something new or neglected by mainstream theory?

  7. Another good example of this is the Tephritid fly Eurosta solidaginis. Adults lay eggs on the buds of goldenrod plants. The larvae burrow into the stem and stimulate the growth of a gall that both protects the developing larva and provides it with a constant supply of food (niche construction).

    Smaller galls are more easily penetrated by parasitoid wasps which prey on the larvae. Larger galls are more easily located by larger predators like chickadees that peck into the gall and eat the larva.

    An interesting article (DOI: 10.2307/2409548) treated gall size as a phenotypic trait and showed that this was a heritable trait using families of flies. They also showed that the local selection pressure on gall size was an interaction between the frequency distribution of gall sizes at a location with the relative impact of the two types of predation which exert selection pressure in opposite directions.

    This was published over 25 years ago.

  8. I was being taught this years ago during undergrad that organisms can develop their niche as well as the environment affecting the organism.

    Sphagnum creates an environment where it can thrive at the expense of other organisms and changes the environment to bog.

      1. I can understand his excitement. It fills the aqueous zone below with gradually degrading sponginess in an acid broth. What’s not to like? You can bounce on it!

  9. Once again, thank you, Jerry for another marvelous exposition of what does, and what doesn’t, constitute an important improvement to the Modern Synthesis. It would be interesting if you published a book which was your Collected Takedowns. We have more than enough hype — you are filling an essential role by doing the hard work of unhyping.

  10. Vavilovian Mimicry is a perfect example of humans inadvertently selecting weeds that look like crops. I agree with Cyne 1,000% This seems like nothing new. buzzword seems like a good way of describing it

  11. Thank you for this excellent post and analysis. I collect these because I have a couple of friends who are theistic evolutionists and are constantly letting me know how the modern synthesis is “outdated”. It’s nice to throw takedowns like this back their way.

    Keep the science posts coming!

  12. “… one can criticize ideas in a manuscript without assessing the psychology of one’s opponents. ”

    I will disagree. The topic of the paper is not just the phenomenon, also but how it is being sold to the public. Is this just a dispute between experts in the field that has yet to be resolved? Or is it really the case that, as one side claims, the other side is motivated by outside factors?

    Outsiders generally don’t have the background to deeply evaluate the arguments on each side, so in part rely upon proxies such as the perceived authority of the participants, as well as any signs of potential bias. A naive observer might just see a dispute between two groups of knowledgeable smart people, and withhold taking sides until eventually one of competing groups yields to the other’s position (or one side slinks off into obscurity). In a situation where one side is claiming that it is a completely settled issue, potential sources of bias in the other group can help an outsider understand possible reasons why there is still a dispute.

  13. Hi Jerry,
    I’m glad you found our article interesting enough to write about. We did get comments from reviewers to tone down the language as the original draft was an order of magnitude snarkier (We’ve been debating EES among ourselves for a few years now, so, I guess the sense of incredulity got built up). We’ve submitted this (after review) to the Journal of Genetics’s special issue on Foundations of Biology. The issue should come out around June.

Leave a Reply