In a paper published in Nature in 2001, Marten Scheffer, Steve Carpenter, Jonathan Foley, Carl Folke and Brian Walker reviewed the evidence for large-scale shifts between alternate stable states in different ecosystems, the patterns that emerge from such shifts, and its implications for the management of these ecosystems. Fifteen years after the paper was published, I spoke to Marten Scheffer about how this paper came about, its impact on the field, and what we have learnt since about “catastrophic shifts in ecosystems”.
Citation: Scheffer, M., Carpenter, S., Foley, J. A., Folke, C., & Walker, B. (2001). Catastrophic shifts in ecosystems. Nature, 413(6856), 591-596.
Date of interview: 5th September 2016.
Hari Sridhar: What was your motivation for writing this specific paper? Looking at your publication profile, I realized that you’d already been working in this area of research for a long time. What was the specific motivation for this paper?
Marten Scheffer: I’d been studying lakes, and then I met people working on different ecosystems that seem to have the same kind of observations suggesting alternative stable states. And I found that interesting, and at some point, I thought it might be useful to just put it together in a review paper, so other people could also have a look at it.
HS: Stepping back a bit, could you tell me how you got interested in looking at alternate stable states. Was the interest originally in relation to theory or in field ecology?
MS: Well, I had learned about that theory when I was studying biology, just in general classes of mathematical biology. And then, when I got my first job that was in applied ecology, at a water quality research institute, the observations that people seemed to have, to me, fitted very well to this theory. But definitely, the first part of my career was all quite focused on applied ecology for lake management.
HS: This paper has five authors from different parts of the world. Can you tell us how this group came together and what each person brought to this paper?
MS: Yeah, the authors are basically experts in different systems than the shallow lakes that I had been studying. And with each of them, I had been discussing the general idea of tipping points. And since my own expertise was only in the shallow lakes, I thought it would be a good idea to broaden out the authorship and ask everybody to reflect on it and make sure that the paper is not naive from other points of view, than my shallow look at shallow lakes!
HS: Was there a particular meeting or conference that acted like a trigger for this group coming together?
MS: No. I wrote the first draft of that paper at the Chilean coast where I used to go every year for a month or so; kind of like a kind of mini sabbatical. So, I spent half my day writing and the other half just walking on the beach and cooking and that kind of things. And, for me, that works very nicely as an annual rhythm. And this paper was my own assignment, so to say, for that year. So I wrote the first draft quite in isolation, but then the others helped a lot in sharpening it.
HS: In which year did you write the first draft?
MS: When was the publication of the paper? 2001? Which month?
HS: October
MS: October. Okay, then it must have been during the northern hemisphere winter of 2000-2001; December-January.
HS: Did the group of authors ever come together in one place?
MS: No, we never came together in one place. We all knew each other. And we have come together in various places, both before and after. But there was not one particular meeting, especially designed to write this paper.
HS: There are certain terms that you have used in this paper, which now are used widely in the literature on this subject. I was curious whether these were the first times these terms were used in this research? For example, ‘catastrophic shift’, ‘catastrophic transition’ and ‘early-warning signals’?
MS: I think ‘early-warning signals’ was not in the 2001 paper. ‘Catastrophic shifts’ basically came from dynamical systems theory, where you have catastrophic bifurcations, which are sometimes called, I think, blue sky bifurcations, which are points in the parameter space where, all of a sudden, an attractor disappears. And it disappears because it meets with an unstable saddle point, in this case. But all the bifurcations have the characteristic that a microscopic change in a parameter can cause a large jump in the system. In that world of dynamical systems theory bifurcation analysis, we call those catastrophic shifts.
HS: The reason I asked about early-warning signals is because, in a section of the theoretical framework, you use the term early-warning signals within quotes. So I was wondering if this was the first time that the term was used in this literature.
MS: Oh, I forgot that I had that term in this paper. Later, of course, we have done much more on this.
HS: So would this have been the first time you used the term?
MS: I guess so. But by then I wasn’t yet thinking of the elaborate work on early-warning signals that later we have been doing,
HS: Another important term in the paper is “resilience”, which came from Holling’s paper. In the Acknowledgments you thank Holling for playing a key role in stimulating many discussions. Can you tell us a little about how you knew Holling and whether you had worked with him earlier?
MS: Yeah, sure. I am a lake ecologist. On one of the lake ecology conferences, I met with Steve Carpenter. He’s also a lake ecologist. And he has, just like me, a lot of theoretical interest in ecosystem stability and resilience. And he was part of a network of scientists, organized by Holling that was called the Resilience Network. And we talked on that conference and we had seen each other’s publications, but that was the first time we met. And Steve thought that I would greatly enjoy this network of scientists and they might be interested in my work. So he introduced me to Holling. And the next thing I knew was I was invited to a workshop of that Resilience Network in a tent camp in the savanna of Zimbabwe. That was the philosophy of Holling, that to make people do really deep interdisciplinary work you need peace of mind and time. And he usually gathered small groups of people on islands where you couldn’t go away. Well, this wasn’t an island, but we couldn’t go away because there were tigers and lions and everything around. Tigers, not I think, but definitely lions and all kinds of other things. So, yeah, we were there. And that’s a great kind of setting because you talk all day, you sit at night by the campfire, you keep talking. And there I met Holling, of course, as well as a nice other group of scientists, both ecologists working in different systems and also economists and various social scientists. So, since then, we have met many times and that experience has been really enriching for me.
HS: Do you remember the year of the lake conference and the tent camp?
MS: I don’t know. It was in Denmark, a conference on shallow lakes. I think that probably must have been the early 90s.
HS: And the camp in Zimbabwe?
MS: Yeah. Also in early 90s. I’m quite bad at remembering dates.
HS: Would you remember how long you might have taken to write up a draft of the paper?
MS: Yeah, I guess maybe three days or so.
HS: Wow. Were all the discussions with co-authors over email?
MS: Yeah, all over email. Actually, at first, I didn’t plan to send the paper to Nature at all. I was thinking to send it to, I think, Ecology and Evolution. But then Steve carpenter, one of the co-authors, thought that it might actually be an eye-opener to more people and suggested to send it to Nature. And they liked it. So I think that helped to make that paper so highly-cited.
HS: What did the other authors contribute to this paper? Was it mainly the examples of the different systems that feature in the in the paper?
MS: Definitely the examples that feature in the paper. And also the framing of the overall paper – to have the framing just right for a broader audience, having it correct for different kinds of ecosystems, having it correct from a theoretical point of view. To make an impact, a paper should resonate well with a broad audience and it definitely helped a lot to have this quite broad group helping on that.
HS: Some of the figures from this paper have had a strong influence on subsequent research. Could you share with us a little about these figures, especially Figure 3 in the paper? Did you make that figure?
MS: Yeah, I made that figure but, of course, that is not at all new. I recently found out probably the earliest source of similar figures. And that was in the context of cell biology, where cells, during their specialization, they commit to a certain path. They become, for instance, either a liver cell or a blond cell or a bone cell, and that was nicely illustrated with similar, even more elaborate, figure of balls rolling on hills. So it’s not a new way of depicting that, but I guess the way it came together and the way we wrote it just made sense to people. And why so many people found it useful.
HS: Did you come across this cell biology example recently?
MS: Yeah, that’s like the earliest source that I could find, but, I don’t know, I must have seen similar figures before. It’s definitely not a way of depicting it that I came up with.
HS: Another person you acknowledge in the paper is P. Yodzis.
MS: Yeah, that’s Peter Yodzis. I’m not sure, I don’t remember exactly how he helped. He’s a theoretical ecologist. And maybe he was actually a reviewer that signed the review;made a lot of useful comments. I guess that is probably the way that his name comes in the Acknowledgments.
HS: Yeah, you thank him for helping improve the clarity of the manuscript.
MS: Yeah. I’m sure he was a reviewer and he probably signed his review.
HS: Did this paper have a relatively smooth ride through peer review? I’m guessing Nature was the first place you submitted this to.
MS: Yes, yeah. As I remember it, it went fairly smoothly. So it was like a really easy paper to write and to get published. That’s the way it is, you know. Sometimes you work for years on a paper, and you think it’s a really important paper, and then nobody cares, and it’s difficult to get it published. At another moment, you don’t think much of it, you just write some things up that you already had in your mind, and it goes very smoothly and it turns out to become a highly-cited paper. It’s a bit unpredictable.
HS: How was the paper received when it was published? Did it attract a lot of attention in the popular press and in academia?
MS: I think it did. So I don’t recall globally what’s happened in the press, but definitely in my own country, it was well covered by the press.
HS: What about in academia?
MS: Well, you know, I think in academia, the situation was probably a bit divided, in the sense that, it was an older discussion already. There was kind of a movement saying that it was all a bit overstated and the data didn’t really show it and I think Robert May called it post-hoc window dressing. And then there was kind of a backlash such that it was discredited overall a bit. But I think that several ecologists over time have felt that in their systems it did make sense to look at it this way. But they were perhaps for some time a bit the underdog, so to say, in the ecological community. I’m not sure, but I think they were happy to see the paper coming out and then came up with other work supporting it or showing for other systems that there weren’t alternative stable states, at least, it was it was like revitalizing the discussion. And I guess the fact that it appeared in Nature gave a lot of credit to the idea. I guess the time was just right for it, to have such a paper. People found it useful to cite that, and it made sense for them in the systems they were studying.
HS: At the time when you wrote the paper, did you anticipate at all that it would have such a big impact on the field? And do you have a sense of what it’s been mainly cited for?
MS: I definitely didn’t expect it to be a big deal. As I told you, I wasn’t planning to submit it to Nature until Steve Carpenter told me that he thought it would be a good idea. I had the feeling that I was just putting some things together that were out there in the literature already, you know, to clear up my own mind, as it often is when you write a review paper. Then I thought, well, it looks nice, and I’m going to submit it to a nice journal, a decent journal, but I wasn’t thinking of Nature in the first place. So, I didn’t think so much of the paper in the first place when I wrote it. It was just like reflecting my own view on the field at that point.
HS: Do you have a sense of what it typically gets cited for?
MS: Oh, actually, I haven’t followed up on that. I know for sure that many are appropriate, but there are too many to check. So, I wouldn’t know.
HS: Did this paper have any sort of direct impact on your career?
MS: Yeah, definitely, definitely. It gave me, I think, a lot of credibility. And the same is true for all of the other co-authors,. It definitely helped our credibility. Yeah, some others have mentioned that, that, for them, it was an important paper in their career.
HS: Would you say that this paper was critical point in the course that your research took? Did it shape the research you did subsequently?
MS: Not the paper, but I was interested in that general idea, and, obviously, I’ve done a lot of follow-up work on systems with, so to say, with tipping points. Yeah, I guess you get fascinated by an idea and then it’s difficult to escape it. And the paper was, of course, an important step in that, but it was more writing up our thoughts, than something new for our thinking. But it definitely helped our credibility.
HS: Today, it’s 15 years since this paper was published. As a thought experiment, if you were to rewrite this paper today, would your emphasis change in any way?
MS: No, I’m still quite happy with the paper. Of course, for all the examples we cite, we now know more about the background. And also we have evidence for many more systems, ecological systems, but also other complex systems. So now, often, my research is going towards a broader diversity of complex systems. I guess I wouldn’t write that paper anymore. But I’m still quite happy with what is in the paper and how it is phrased.
HS: I wanted to pay attention to certain specific sentences in the paper in which you talk about lacunae in our knowledge at that point and hint at what needs to be done in the future. I want you to reflect on that, today, 15 years later. First is where you say, ‘Observation of a large shift per se is not sufficient, as systems may also respond in a nonlinear way to gradual change if they have no alternative stable states. Also, the power of statistical methods to infer the underlying system properties from noisy time series is poor’. I wanted to ask you whether, today, there are better statistical methods available to infer system properties from time series data.
MS: Yeah, I think there are. Perhaps the most important step forward is that we are now looking at noisy time series, and from that trying to infer resilience of alternative states, in different ways. One way is to compute the stability landscapes directly from big data. Perhaps you’ve seen the paper that we did with Hirota as first author, in Science, on tropical forests. So that’s an example. Again, the mathematical theory is not new, but now we’re looking more at theory of stochastic systems and realizing that you can use that information to, for instance, plot those stability landscapes, and also infer, from the time series, something about the resilience of the system. So, the Nature paper on early-warning signals and the Science paper following up called Anticipating critical transitions. So, yeah, definitely a lot has been done since those days. And again, as you see, typically those things are not novel from a mathematical point of view. It’s just new to link the abstract mathematical theory in a way that works to data. So, in that, definitely, we have really added a lot over the past decades, if you look back at the 2001 paper.
HS: And just after that you say, ‘the strongest cases for the existence of alternative stable states are based on combinations of approaches, such as observations of repeated shifts, studies of feedback mechanisms that tend to maintain the different states, and models showing that these mechanisms can plausibly explain field data’. Now, again, subsequent to your study, have there been studies that use a combination of approaches, and did your paper stimulate this kind of an approach to show alternative stable states?
MS: Yeah, definitely. By the way, I have written, with Steve carpenter, a companion paper to the 2001 paper. And that elaborates on those points. In the lab, if you have a control system, well, then you can show everything, but in nature, it’s like never ending detective work and it’s very hard to be entirely sure about anything. Ecology has been losing a lot of time discussing whether you can actually prove that competition exists, for instance, or top-down control. And so you have to piece together the evidence and try to see whether it’s plausible. And I think that, actually, since 2001, in general, ecology papers have moved more and more towards that kind of combination of experiments, field data, and models. It has become the standard today. And definitely for the lakes that we study ourselves, that has been the case. Since that time, I think, ecology has very much moved in that direction.
HS: I wanted to ask you about the impact of this paper on management. In the paper you say, “The main implication of the insights presented here is that efforts to reduce the risk of unwanted state shifts should address the gradual changes that affect resilience rather than merely control disturbance. The challenge is to sustain a large stability domain rather than to control fluctuations”. Do you see a change in management that reflects the insights from this study, at least in some systems?
MS: I think I do. By the way, basically the same thing has already been said by Holling in the paper in the 70s on ecological resilience. But I think our paper definitely reinforced that idea, and, especially, this little figure with the rolling balls has been really important. I don’t know for all the fields of ecology, but I can definitely speak for the community of lake managers. And they found that figure very useful, and they entirely get it, and they write that that’s the way we should work on resilience. There is a recent policy forum that we did in Science that is called, I think, ‘Creating a safe operating space for the world’s iconic ecosystems’. And that paper, basically, reiterates the point for tropical rain forests, for coral reefs and for wetlands. So, I think, you know, it’s not like you say that kind of things once and then everybody gets it. You should repeat the message. But I was surprised how important a simple appealing figure is to get the message across. You can write a lot of things, but a simple figure, even if it is a gross oversimplification of reality, helps to frame the mind.
HS: In the 15 years since this paper was published, have you ever read the paper again?
MS: Not the entire paper. I haven’t looked at it, probably, a few times, but in general, I don’t read the papers again. But this one I have picked it up, I think, a couple of times and looked at some of the formulations again.
HS: If you compare this paper to papers you write today, is the way you write more or less the same?
MS: I think it is. Yes.
HS: Would you count this as one of your favorites among all the papers you’ve published?
MS: Yeah, I think I do. I like it. Yeah.
HS: Could you tell us a little more about what you like about it? Is it in terms of the impact it’s had?
MS: No, I think it frames a really complex issue in a way that helps people thinking about it. I think for this kind of paper, that’s what makes me happy. Other kinds of papers where you have an original new finding, that’s also very nice. Here, that’s not the case; it’s a review paper. I think I’m happy with the way it turned out, that it became helpful for people and inspiring for people, apparently. So what else can you wish for a review paper?
HS: What would you say to a student about to read this paper today? What should he or she take away from this paper written 15 years ago?
MS: I think the paper speaks for itself. It doesn’t need much explanation.
HS: In relation to what’s happened since then, would you add any caveats or point them in new directions?
MS: I would probably say that this was an important paper in the field. Since that time, research has moved forward to show similar kinds of properties, not only in ecosystems but also in the climate and human systems. And also, we have, now, this new work suggesting how we may actually estimate resilience from data.
0 Comments