Much Ado About Niches

Adrian Currie writes...

I’ve found talking about ‘niches’ helpful for conceptualizing adaptive patterns on life’s macro-level. In particular, why we see similarities cropping up between different organisms all over life’s history. Thinking about niches allows us to see how similarities between quite different lineages might be due to them being adapted to the same, or similar, ways of life. This is related to a way of thinking about lineages, that they are what David Hull called ‘genealogical actors in ecological roles’. However, as we’ll see, taking niche-talk at face value might be problematic. In this post I’m going to have a bash at explaining niches, what might be problematic about them, and then try and resolve that by considering how niches are used in paleontology. It also gives me an excuse to babble on about some of the weird critters from Aotearoa (New Zealand).

The Kiwi of the North

Here’s a kiwi:

Only matched in cuteness by their bad-temper...

Only matched in cuteness by their bad-temper...

Cute, right? Weird too. Obviously kiwi are birds – they’ve got feathers, a bill, lay (whopping) eggs, and are part of the Aves family. But they’re weird birds. They come out at dusk and root about for insects. They live in burrows which they dig with their powerful, massive legs. They have an exceptional sense of smell. Their feathers form a soft down. Their main defence is a stout kick (again, those are big legs!). Their eye-sight is profoundly sub-par. They’re also typically pretty grumpy, lonely critters: they’re usually very territorial. Compare that to a normal bird:

This was literally the first google-image hit for 'normal bird'.

This was literally the first google-image hit for 'normal bird'.

She is pretty social, has lame—lame—olfactory prowess, but can see really well. She’s diurnal, certainly doesn’t live in a burrow, and those legs couldn’t do much kicking nor burrowing. Oh, and I guess she can fly while kiwi can’t, but that seemed too obvious to mention. But there are a bunch of animals who the kiwi does seem similar to…

A badger!

A badger!

That’s right: Badgers. Nocturnal, insectivorous, burrowing, nice soft fur, amazing nose, rubbish eyes—a bit of a loner, a bit irritable. They’re the kiwi of the north.

What’s going on here? It’s tempting to think that badgers and kiwi have similar traits because they make their livings in similar ways. That is, they occupy the same niche—that of the nocturnal, forest-floor insectivore—and evolution has shaped them to fit that niche. Hence their convergent similarities. The thought is that there is some ecological and environmental ‘space’ which both badgers and kiwis evolved to occupy. Although they are profoundly different genealogical actors, they play the same ecological role—and thus their evolution has converged.

If such niches exist, this explains certain patterns in the fossil record, particularly those involving various kinds of convergences, such as that between kiwi and badger. But it also sheds light on phenomena such as island dwarfism and gigantism, as well as adaptive radiations. It also has the potential to explain the ‘ecological replacement’ that seems to happen after mass-extinctions (swapping sauropods and ceratopsids for bovines, say).

Misplaced Concreteness

So, what’s wrong with niches? In his recent book, Dennis Walsh writes of them:

The niche is an apposite metaphor for the nature of organism/environment relations, as conceived by the adaptationist programme. A (nonmetaphorical) niche is a space, or a recess, into which something (say, a statue) might fit. Like a real niche, an evolutionary niche is a set of properties of an environment, to which an organismal form may fit. (169)

On this kind of picture, prior to the kiwi’s arrival, Aotearoa had an empty niche space which wanted filling.

On this kind of view, niches are non-relational properties of environments which are independent of organisms. Again quoting Walsh,

[on such views] The features of the evolutionary niche can be specified independently of the organisms that occupy them (169).

That is, I can specify an ecological role without worrying too much about the genealogical actor. Niches appear to be concrete objects: they exist within environments. And they are in some sense independent of the organisms to which they’re shaped: the kiwi’s ancestors were adapted to that pre-existing niche. Dick Lewontin ties these independent, non-relational notions of niche to adaptationism:

To make the metaphor of adaptation work, environments or ecological niches must exist before the organisms that fill them. Otherwise environments couldn’t cause organisms to fill those niches. The history of life is then the history of coming into being of new forms that fit more closely into these pre-existing niches. (Lewontin 2001: 63)

Unsurprisingly, both Lewontin, Walsh, and others, think that this kind of thinking is just a mistake. Environments and organisms are dynamic: they shape one another. Natural selection doesn’t simply mould a lineage to fit a foreordained, independent environment. First, a critters’ ancestry makes a difference to how it may be shaped. Kiwi and badgers might be pretty similar in the ways I specified, but they are, of course, profoundly different. You could put a bird in a badger’s environment for as long as you want, but you won’t literally get it to evolve into a badger (I suspect it’s even harder the other way around). Second, environments are mostly constituted of other living things, and these interact with each other over ecological and evolutionary time. It is not as if the various bugs the kiwi prey upon would not themselves have been effected by the kiwi’s arrival. They would evolve too.

So, there’s something a bit weird about, let’s call it, concreteness about niches. This is a metaphysical view – it tells us what niches are – and it claims they’re (1) non-relational, concrete objects which are (2) part of environments, and (3) exist independently of (or prior to) organisms. We shouldn’t like this view because 1 and 3 are likely false: niches appear to emerge from the interaction between organisms and environments, and so are relational, and are not independent of organisms.

I just spent a week with the wonderful people at Egenis in Exeter. Some of the folk there are interested in ‘process ontology’, and so have been reading Alfred North Whitehead (they’re much braver than I!). They introduced me to a little nugget of his which I suspect helps clarify what’s potentially problematic about niches. According to Whitehead, we often commit the fallacy of misplaced concreteness: taking an abstract concept to be a concrete object. This often happens when we start measuring complex things. Consider gross domestic product – this is a measure intended to estimate the total number of goods and services a country provides in a year. It’s an aggregate measure at best: an abstraction. And yet, after years of economists discussing GDP, its movements, what forces affect it, and so forth, one might start to think of it as a real, concrete, object. This would be to commit Whitehead’s fallacy.

I suspect people who are worried about niches are worried that something similar is happing here. What was a useful metaphor for organism-environment interactions became reified, concretized: we started to mistake the abstract for the concrete.

Enter the Moa

There are three things about the discussion in the last section that leave me feeling uneasy. First of all, I’m generally suspicious of distinctions between ‘abstract’ and ‘concrete’ properties: I’m inclined to think that whether we should consider a property to be abstract or concrete is extremely sensitive to how we describe things, and that the two fairly casually shade into one another in most circumstances. But let’s leave that to one side (if you want to know why I think something like this, read up on homunctionalism…), and focus instead on the next two worries.

Second, worries about niches seem to have focused on (and here’s some hip philsci lingo) the products of science rather than its practice: adaptationist appeals to niches in explanation are taken to reflect a view on what niches are. And of course the way scientists talk about niches often matters, but I’m more interested in how notions like ‘niche’ shape how they reason, the evidence they're interested in, and the investigations they pursue.

Third, much of the niche-talk comes from neontology, that is, the biological science of currently living critters. If niches are the sort of thing which lineages adapt to over evolutionary time, then we need a broader temporal scope than that. A single time-slice of biological history can only provide limited information about macroevolution. As such, it’s not obvious to me that an evolutionary notion of ‘niche’ will feed into neontological practice.

So, putting both worries together, I want to ask what role niches play in the practice of paleobiology. As we’ll see, at least given the examples I’ll look at, niches play an important role in framing questions about macro-evolutionary biotic turnover, but that taking the view metaphysically seriously is not supported by what paleobiologists do. That is, they are not over-concretizing.

To make my argument, I want to switch to another weird New Zealand bird.

Here’s a representation of a fair chunk of Aotearoa’s contemporary ecological relationships:

Some Sheep. Romney sheep, to be exact.

Some Sheep. Romney sheep, to be exact.

Sheep and grass. However, humans have only lived on those islands for around 1500 years at the absolute maximum, and before we turned up with our repertoire of mammalian compadres things were, well, a bit different:

Aotearoa circa 500 A.D.

Aotearoa circa 500 A.D.

The picture above represents a predator-prey relationship between Haast’s Eagle and some of the largest moa. Haast’s Eagle is in the running for being the biggest raptor ever. Her wingspan at maximum is around 3 meters, which isn’t too impressive (the biggest extant eagles manage this), but her body mass is around 40% larger than the biggest living eagles. Anyway, these chunky eagles appear to have preyed on moa: flightless browsing birds which managed around 3.6 meters in height (when fully erect, although their stance was typically more horizontal).

Richard Owen inferred the existence of a moa from a single leg in 1839. Here he is an older man with the complete skeleton of Dinornis robustus, vindicated. Vindicated, but it's worth noting that moa totally wouldn't stand like that.

Richard Owen inferred the existence of a moa from a single leg in 1839. Here he is an older man with the complete skeleton of Dinornis robustus, vindicated. Vindicated, but it's worth noting that moa totally wouldn't stand like that.

There were 13 species of Moa, and there was significant overlap in browsing range (although there was clear niche-separation due to different browsing heights and bill specifications).

So, two thousand years ago, Aotearoa’s flora was largely munched upon by giant bird fauna. Nowadays, the flora has to deal with sheep, cattle and other mammalian munchers. How has this changed the plant life and ecology of the islands? Someone might think: not much. After all, after the loss of moa, Aotearoa housed an empty niche space. In came new herbivores to fill it. At least, we might think so given the concrete view of niches. Recall that on the view organisms are shaped to their environments; and so, as sheep, goat, deer and their brethren arrived, they would have adapted to the niches they found there.

So, what happened? There is a delightful, but very dated study by Atkinson & Greenwood which uses the idea of ecological equivalence to examine just what happened when moa were replaced by mammals (it’s from 1989, so very dated, but I’ll let you know when I’ve looked into more modern stuff…). 

Discussions of the effects of introduced browsing mammals on New Zealand plants and vegetation have sometimes included the suggestion that they are not fundamentally different from those formerly exterted by moas. This suggestion is treated here as a null hypothesis and tested by some comparisons between moas and mammals (67).

Basically, the study looks carefully at how different plants adapted to the moa, and how the moa adapted to them. They argue that most of the time sheep and moa are not ecologically equivalent. The paper’s very detailed, but I’ll just give you one example. Some of Aotearoa’s indigenous plants display a very odd characteristic: their lower growth takes on the appearance of dead twigs. Consider the leafless pohuehue, or Muehlenbecki ephedoides:

Plants can play dead too. Nature is weird.

Plants can play dead too. Nature is weird.

What could explain the trait? One thing to note: playing dead is no defence against sheep and goats, who happily munch right through them. Atkinson & Greenwood’s suggestion is that the ‘dead-leaves’ trait is an adaptation to moa-grazing. As (non-kiwi!) birds, moa are likely to have identified food visually, making them susceptible to this kind of trickery on the leafless poehuehue’s part. Assuming that’s right, we have an example of co-evolution: the moa and the leafless pohuehue have evolved together, and in virtue of the differences between moa and mammals, the replacement of one by the other makes a large difference. Where fooling the moa's sight worked well, under the new regime those dastardly mammals just smell the difference. In fact, in many ways the general point that mammals didn't just replace moa should be obvious:

It is unlikely that moas bit and stripped bark from stems, in the way that deer and goats bark pseudopanax spp. and some podocarps. It is also hard to imagine them ploughing the soil and destroying roots in the way that pigs do even though they may have scratched and torn at the ground with their feet. Moas were presumably not capable of climbing trees, so there would be no parallel with the arboreal feeding of brush-tailed possums. (89).

Atkinson & Greenwood pull together many examples like this, to argue that even if moas and mammals have acted in *similar* ecological roles, they certainly do not play them equivalently. Indeed, their paper is only “a first approximation towards a more precise comparison” (92), which would require further tests of various hypotheses, and the inclusion of a range of other features.

In short, though, the empirical story seems clear: there was no independent, concrete niche vacated by moa and then occupied by mammals. What, then, should we say regarding concreteness about niches?

Niche-talk & Metaphysical Commitments

On the face of it, Atkinson & Greenwood’s study is bad news for the view that niches are non-relational and exist independently of organisms. Aotearoa’s plants and moa co-evolved together, and the removal of one half of the partnership, and the addition of a new set of players, looks nothing like ecological equivalence. However, it's worth pointing out that their study started from niche-concreteness, and then built towards a richer, localized picture of the actual relationships between these organisms. It strikes me that it is not that there are no niches here, but rather than the niches depend crucially on environment-organism coupling (that is, they are relational, and not independent of organisms).

I’m inclined to think that this tells us something important about the role of niches in some kinds of scientific inquiry. It’s not that scientists assume a kind of concreteness about niches, but rather that the concept helps shape their investigation. It was the notion of niches, after all, which lead Atkinson & Greenwood to contrast mammalian and moa herbivory in the ways that they did. By starting with a coarse-grained comparison, we gain a vantage point to reach the relevant level of description which allows us to understand the differences and comparisons between different lineages.

If that story is representative, then, it seems as if worrying about ‘niche-talk’ is somewhat misplaced: it’s not a metaphysical claim that scientists problematically commit to, rather it’s a starting point for understanding the ecological and evolutionary relationships between different organisms. Thinking about niches does involve considering genealogical actors in ecological roles. However, that doesn’t mean that the genealogical aspects – the traits making moas a very different browsing prospect to mammals – aren’t an essential part of the story. Further, that doesn’t mean that the roles can’t themselves be ultimately delineated in fine-grained, relational ways.

But what does this mean for my badgers and kiwis? Can I still explain their similarities in virtue of their occupying similar niches? I don’t see why not: just because a property is emergent and relational doesn’t mean that it cannot re-occur. It may be that, in the badger-kiwi case, the kind of environments the two co-evolved with did happen to align in similar kinds of ways. And, moreover, by considering them as similar in this way, we open up new questions about what those relationships might be.

To conclude then, I’m not so worried about niche-talk amongst biologists: I don’t think they’re committing Whitehead’s fallacy (or, at least, not in their actual investigations). Given this cursory examination, their practice doesn’t seem to reveal a kind of commitment to the problematic kind of concreteness we identified earlier. Rather, it strikes me to be a very appropriate, productive kind of concreteness: the kind that recognises that organisms evolve in concert with their environment, that local details often matter crucially for how those relations go, and that we have the capacity to come to understand those relations.