In a pair of old posts I asked, what’s the biggest ecological idea you’ve ever changed your mind about?
Here I’ll pose a slightly different question: What’s the biggest change of mind any ecologist has ever had about any ecological idea?
For instance, imagine if AJ Nicholson had said “On reflection, Andrewartha and Birch were right, density dependence is irrelevant for population dynamics.” Imagine Robert MacArthur saying, “I’ve changed my mind, simple mathematical models can’t teach us anything useful about ecology.” Or imagine me saying, “Whoops, sorry everybody, the intermediate disturbance hypothesis is right after all, my bad.”*
Has any ecologist ever done anything like this? I can’t think of anyone off the top of my head, which may well just illustrate the limitations of the top of my head.
Meghan suggested Dan Simberloff for having changed his mind about island biogeography theory. Are there other examples from ecology?
The only example I can think of is from philosophy. Ludwig Wittgenstein famously developed two opposing philosophies of language, the second of which (ordinary language philosophy) refutes the first (logical atomism).
I’m curious about this because it feeds into our long-running discussions of issues like how fast ecology is progressing, why some ecological ideas seem so persistent, and why the people who know the most about an ecological idea often seem to disagree most about it.
*I find this last one both amusing and worrisome to imagine.
Dynamic Ecology 
If we extend this to evolutionary biologists, as opposed to ecologists sensu stricto, then David Lack famously changed his mind about the role of bill size differences in Galapagos finches – see: https://en.wikipedia.org/wiki/David_Lack#Darwin's_finches
Ooh good one. I didn’t know that.
I have had a few moments like that. One of those where when I realised that the resource concentration hypothesis, that explains insects distributions, was simply wrong, and based on a faulty argument. This realisation arose when I tried to model the processes underlying the distributions and then could not match the output of the model and the current dogma, where the predictions were opposite to each other. To me, this shows what happens when you try to formulate quantitative predictions verbally without doing the math. The surprising observation (or perhaps not) was the large number of prominent ecologists that used and still uses the original resource concentration hypothesis despite the lack of logic in its formulation.
Ooh, interesting. Can you point me to any papers on this?
I’d also like to see the papers! Or preprints if possible.
The original paper describing the verbal hypothesis was Root 1973. Ecol. Monogr., 1973. 43: 95. Our main papers describing the problems with this hypothesis were published in Ecol Lett, 2005. 8: 1057 and in Oecologia, 2013. 173: 1333. The main problem with the verbal hypothesis was that Root used different scales when describing how immigration and emigration rates vary with patch size and other aspects of resource patches. For instance, the basic argument was that immigration increased with patch size (or other aspects) while emigration decreased with patch size, creating a positive relationship between patch size and the species density. A problem with his argument is that it measures immigration rate as the number of immigrants and emigration rate as the probability to emigrate. Howevev, even though the number of immigrants increase with patch size, they are also distributed over a larger area.
We (me Nora Underwood, Brian Inouye an Thomas Verschut) are planning to summarise these findings in an overview paper that describes how to build a more robust theory and also points out possible roads ahead. And, I can certainly send reprints, but you can also find them at my research gate-site.
I remember sitting in the audience at the awards ceremony at the ESA meetings when I was a postdoc and being struck by the citation for Simberloff’s Eminent Ecologist award, since it noted that part of the reason he got the award was his willingness to cast stones at his own earlier work:
“Few ecologists among us have the courage to publicly challenge our own paradigm in this way, particularly once it has become widely accepted.”
from: http://www.esa.org/history/Awards/bulletin/eminent2006.pdf
…which is interesting, because isn’t island biogeography theory widely considered to have stood the test of time better than most in ecology? I feel like I’ve seen some poll that suggested that was the case. Being able to change your mind is laudable, but presumably only if the original idea was actually a bad one (e.g., broken stick)
“I feel like I’ve seen some poll that suggested that was the case.”
https://dynamicecology.wordpress.com/2015/08/03/poll-results-which-big-ideas-in-ecology-were-successful-and-which-unsuccessful/
Thanks for teeing me up. The check is in the mail. 🙂
“Being able to change your mind is laudable, but presumably only if the original idea was actually a bad one (e.g., broken stick)”
Yes, exactly.
Changes of mind about big ideas presumably are very rare. The interesting question is why. Is it because the relevant data and theory rarely change? Because many people resist changing their minds even when the data and theory change? Some of both?
And of course, many of the big ideas about which ecologists might (or might not) change their minds aren’t purely empirical. What you think of, say, island biogeography theory isn’t ordinarily as simple as “I think it’s right” or “I think it’s wrong”. You might also think it’s a fruitful starting point for hypothesis generation, or an elegant idea, or a rich idea with many implications, or a provocative idea, or an oversimplified idea, or etc. So deciding which ideas are “bad” inevitably involves some professional judgment, on which there’s some (but not limitless) scope for reasonable disagreement.
I hadn’t remember such an overwhelming vote (83%) for considering island biogeography theory “successful”, which makes it all the more interesting that one of the few examples people came up with here involved someone going from thumbs-up to thumbs-down on what others consider maybe THE most successful ecological theory! And now some bait for Jeremy. True or false, the IDH is (cut and paste from your comment):
– a fruitful starting point for hypothesis generation
– an elegant idea
– a rich idea with many implications
– a provocative idea
– an oversimplified idea
Don’t make me block you.
Kidding, obviously. 🙂 My short answer is “none of the above”.
My longer answer is that the IDH was a perfectly good intuition inspired by some empirical observations. It turns out that those intuitions are completely wrong (as you discover when you try to express them mathematically). And the empirical pattern those wrong intuitions predict mostly doesn’t show up in the data either (though as an aside the usual statistical tests for that pattern is technically flawed and can be improved upon…). So there’s no longer any theoretical or empirical reason to pursue the IDH as a research program. I do think it should still be taught to students as a cautionary tale. And I think there’s plenty of reason to continue doing empirical and theoretical work on the ecological effects of disturbances. There’s just no reason for any of that work to make even passing reference to the IDH.
Yes, but the island biogeography theory is also fairly loosely formulated and it is not clear what you mean that it has stood the test of time.
MacArthur pretty decisively recanted his broken stick model for species distributions, even becyring the fact that others were still carrying on with it.
John Harte developed a fractal model to explain basic macroecological patterns like SAD and SAR. Then he replaced it with the HEAP model. Which he then replaced with the METE/MaxEnt theory. In each stage he got more elegant and better fits to the data.
This week I’ve been reading a lot of papers on allometry – but in the context of geomorphology. (The geomorphologists were citing biologists/ecologists.) There was flash of work on geomorphic allometry in the ’70s, which was itself responding to, in essence, a raft of power-law relationships, mostly for river networks, put forward in the ’60s. (Meanwhile, SJ Gould’s big Biol. Rev. article, “Allometry and size in ontogeny and phylogeny,” came out in 1966.) All seemed to fizzle by about 1980 – big arguments in the discipline about how difficult it was to connect these myriad scaling laws to physical mechanism(s). People wandered off in other directions. But much of that conceptual framing from the ’70s remains very rich material, and remote-sensing has changed the sheer volume of Earth-surface observations we can make. Still difficult to link quantitative pattern to mechanistic explanation – and these cool roots are all there in this decades-old literature. (The biology/ecology work to which they refer is even older, of course.)
I can’t speak for ecology specifically, but IMO lots of big ideas in science stick around or people don’t change their minds about them because there isn’t enough definitive data to clearly distinguish among competing hypothesis.
Imagine, for example, if the Cretaceous-Paleogene boundary impact crater had been subducted rather than preserved – the ecology vs impact controversy regarding dinosaur extinctions would still be raging and the concept of impact-driven extinctions would remain unconfirmed.
Hi Jim – what’s interesting about that example is that, for outsiders such as ecologists it looks as though the matter is settled. But for insider geologists and volcanologists actually involved in the research, the debate is far from settled.
This recent interview about the history of the debate is very illuminating; I’d not realised how problematical the asteroid hypothesis actually is, that aspects of the fossil record do not support it, and the extent to which acceptance of the asteroid hypothesis was driven by academic politics and rivalries:
https://www.theatlantic.com/magazine/archive/2018/09/dinosaur-extinction-debate/565769/
Hi Jeff,
Cool! That’s great! I love geological controversies because many of them never get resolved. Extinction events are endless battlegrounds. But really even that isn’t the only competing hypothesis. There are still people tucked away at various small schools and state geology bureaus claiming an ecological process for the extinction.
I’m sure if I did a full technical review of the Decaan hypothesis I could find problems with it. It would be nice to see a diagram showing the timing of the critical events.
The main issue that comes quickly to mind with Decaan: The Columbia River Basalt group here in the PNW is about 1/10 the size of the Decaan traps, but still ten thousands times larger than Laki, with no biological mark whatsoever as far as I know, so it’s really not immediately clear if the “scale up Laki” argument is viable. In fact the CRB interflow sediments have beautifully preserved Gingko trunks several feet in diameter.
I have been struggling with similar ideas, albeit in evolution, for some weeks now. Sometimes losing a few nights of sleep. There is nothing new in my comment. I will just be ranting my random thoughts here. Whether ecologist or evolutionary biologist, we all envy physics. I am not a physicist, but I will try to give an example from physics anyway. So, how do you have a big change of heart in physics? Consider the Newtonian concept of gravity and the world of Einstein’s space-time. Newton’s laws were tremendously successful in predicting the orbit of planets and the timing of future eclipses. It was so successful that when there were deviations from the predicted trajectory, it would suggest the presence of other heavenly bodies (that’s how Neptune was discovered). When the precession of Mercury’s perihelion showed deviation from the predicted values, some believed that there should be another planet just like Neptune. But no one could find that hypothesized planet. Although successful in many scenarios, the theory went into crisis when it couldn’t explain the deviation. That’s where Einstein’s theory of relativity came. Not only it could explain the deviations but could also predict the bending of light around the stars. Such predictions were not possible in the Newtonian world. When tests after tests corroborated Einstein’s theory, physicist had a major change of heart. My point is change in heart comes from recognizing that the theory is in crisis in which it cannot explain some aspects of the phenomena. In physics, even small deviations in predicted and observed values can throw a theory into crisis. But even large deviations are normal or expected in ecology and evolution. Unless the ideas are really shitty with very little resemblance of prediction with the observation, we don’t normally consider a theory to be in crisis. Deviations are the norm.
I don’t think I am an ecologist. But still, let us consider a(ny) model of predator-prey interaction and a hypothesized world. Consider an alternate Isle of Royale without any moose or wolf. Let’s introduce specific numbers of both prey and predators. Also, let’s assume that we have a good estimation of parameter values. For the sake of argument let’s keep everything else constant. Given these conditions how accurately can we predict the number of moose and wolves in the coming years? If we have a prediction that is later confirmed by observations, then we would be satisfied with the model. But if we see enough (I don’t know how much is enough) deviations in the observations we will have at least two hypotheses. One, just like before, we would hypothesize that there be another species that is closely interacting with one or both species. We can test this hypothesis either with long term observations. Or, we could solve the inverse problem and try to infer the type of species interaction that would explain the observations and narrow the search criteria. If such endeavors are successful, we would have even more confidence in our model. But say we don’t find another species that is closely interacting with either of those species. Now we will realize the crisis in our model and try to find a one that has better prediction and explanation capability. In that case, there’s a possibility of finding other unexpected mechanisms that give rise to such patterns. Then there’s a possibility that we might have a change of heart regarding mechanisms of such predator-prey interactions. But alas, we don’t live in such a world. Nothing is constant for us to figure out whether the deviation is due to other species, lack of better mechanisms or other various factors. Because of the sheer number of factors associated with any interactions, deviations from prediction is what we expect. Even if a new idea comes along saying some other factor is responsible for some of the pattern in the data, we won’t have a big change of heart since we expected there to be something else. Hence, I think to have a big change of heart, we need to find something that is completely wrong and lacks any sorts of empirical evidence. This is hard even if we have ideas that are just “okay”.
Be it ecology or evolution, we rarely have expectations for exact predictions. If the data roughly shows predicted trends, then we are satisfied (although the hunt for a better model can go on). We are almost sure that we will always find a system where our model won’t work. Instead, we are looking for a system where the model will roughly work. Only if the disagreement between predictions and observations are too bad, we are forced to consider alternatives and get a chance to have a big change of hearts regarding the mechanisms.