In a paper published in the journal Oikos in 1994, Clive Jones, John Lawton and Moshe Shachak introduced the term “Ecosystem Engineer” into the ecological lexicon and laid out a research agenda to explore, identify and quantify ecosystem engineering carried out by living organisms. Twenty-two years after the paper was published, I spoke to Clive Jones about how he got interested in the topic of ecosystem engineering, the making of the paper and the impact the paper has had on subsequent research.
Citation: Jones, C. G., Lawton, J. H., & Shachak, M. (1994). Organisms as ecosystem engineers Oikos 69, 373–386.
Date of interview: 17th August 2016 (on Skype)
Hari Sridhar: Prior to this paper, most of your research was on the chemical ecology of plant-herbivore interactions. What triggered your interest in ecosystem engineering?
Clive Jones: It really came out of a study on rock-eating snails in the Negev desert in Israel. My colleague Moshe Shachak became an Adjunct Scientist at the Cary Institute in 1983, and we had these conversations about the Negev desert. He said he thought there was something very interesting going on in rocks that contain endolithic lichens, a couple of millimetres down inside the rocks. He said that there were what looked like clear colonies on the rocks – not epilithic lichens on the surface – but sort of underneath the surface with clear marks around them. Lichenologists assumed, as is the case with some other lichens, that this was a lichen colony-lichen colony interaction. But Moshe didn’t think so. He thought that small snails were responsible, although he hadn’t done any research on it. I thought this was really interesting from the viewpoint of chemical ecology. I wondered whether endolithic lichens need to defend themselves against being eaten. After all, they were inside the rock and protected. I thought it might be interesting to see if they had some sort of chemical defence.
The next year, Moshe showed up at the Cary Institute with rocks and snails, a sort of portable miniature ecosystem. We could activate the snails by simulating dew, which is what normally happens in the desert. And sure enough they were eating the rocks to get at the endolithic lichens, their food. We thought that was very cool. I did quite a bit of work looking at whether or not the endolithic lichens were in any way chemically defended but nothing interesting emerged from that. What became apparent to me and to Moshe was that what they were doing to the rock was much more interesting; and, being in an institute of ecosystem studies, what that might then mean for the rest of the desert was in the forefront of our minds. We initiated a whole series of studies on the rock-eating snails. There are three species but we focused on two of them; the third is more restricted in its distribution. It turns out by the way, that there are similar rock-eating snails in the Dolomites in Germany and Italy. What our work showed was that these snails had a tremendous effect on the “weathering” of rocks, soil formation and nitrogen transfer. The work was published in Science and Nature and got a lot of attention. It was “Gee whiz” biology! You can imagine. Tiny snails creating about 1000 kilograms of soil per hectare per year, eroding the rocks away, and transferring large amounts of nitrogen to the soil that would otherwise be trapped forever inside the rock after it was taken up from atmospheric dust by the lichens.
When I was drafting the conclusion to the Science paper I started wondering whether there was a larger context to all of this. What, if anything, did it have to do with earthworms forming soil – as in Darwin’s last great work – or to a variety of other examples in the literature that perhaps seemed similar, like corals building reefs for instance? That thinking really was the germ of the 1994 paper. At the same time, I started working in the Negev from about 1987 onwards studying a whole bunch of different organisms all of which did some sort of very clear physical modification to the environment. Shrubs and ants make soil mounds that trap runoff water. Microbial crusts generate runoff on slopes by secreting muco-polysaccharide that acts like a plastic sheet. Porcupines dig pits and make mounds. Geophytes make soil cracks where water infiltrates. The physical changes caused by all these species then affect run on-runoff relationships. A lot of human construction in the Negev, from the Nabataeans 2,000 years ago to people today, involves trapping runoff water using various physical structures. As we started to study these various systems it was clear that all the organisms were having their effects by physically modifying the environment. And these modifications then had large knock-on consequences for other species in the system and for the functioning of the entire ecosystem. For example, productivity of the Negev desert is markedly increased by the combination of organisms that generate runoff water and/or then trap and concentrate it. Were this not to occur, most of the limited rainfall in this area would just rapidly evaporate from the surface.
So these studies and other work we were doing at the Cary Institute really started to catalyze broader thinking and exploration of the literature to find out what exactly it was that we were thinking about? Had it been studied earlier? Obviously yes, there were lots of examples. Is there any sort of conceptual thread holding them together? No, as far as we could tell.
The other factor that helped move this forward was John Lawton also being an Adjunct Scientist at the Cary Institute. John came from population and community ecology as did I, although I was trying to rapidly become an ecosystem ecologist. At that time and still today, we organized a Cary Conference about every two years on some theme in ecology that we thought required greater attention. John and I were talking about the divide between population and community ecology on one side and ecosystem ecology on the other. We decided to make it the focus of a Cary Conference. The divide was historical and in some ways artificial. Today, these fields are far more merged than they were back then. One of the reasons for the merging was the Cary Conference we organized and the book that came out of it, “Linking Species and Ecosystems”.
The book got a lot of attention. Ecosystem engineering certainly appears in the book, but it mainly covered the broader topic of how to integrate population/community ecology with ecosystem ecology. What were the interesting questions, avenues and opportunities? In planning that conference, it was clear to us that linking some of this more general stuff with ecosystem engineering – although it didn’t have the name at that time – was an obvious and clear way in which we could show interesting connections between populations, communities and ecosystems. Planning the Cary Conference also simultaneously catalyzed us to try and crystallize what the heck we meant, which eventually lead to the term ecosystem engineering in that first paper.
So in summary, it was a combination of circumstances really: Empirical research over a number of years, a lot of reading, the challenge of the conference, and trying to integrate what we had learnt and make it general. In addition, John, who was then Director of the Centre for Population Biology at Silwood Park, had started some work in the Ecotron looking at the effects of earthworms. It turned out that many of the effects of earthworms on that system were due to the physical modifications they made. So there was this clear nexus – a focus on the physical modification of the environment – that was leading the way. We didn’t worry so much about all the other ways organisms could affect ecosystems because at that time it was pretty clear that there were good conceptual frameworks for them. An obvious example would be biogeochemistry. So that’s really the story of my shift in interest from plant-herbivore interactions to what eventually became ecosystem engineering.
HS: It’s interesting that you mention the Ecotron. The first interview I did in this series on revisiting old papers is of Shahid Naeem about his work in the Ecotron.
CJ: I started working in the Ecotron after Shahid had left. I wasn’t involved in the earthworm experiments but in some chemical ecology stuff, looking at how climate change might affect plant defence and so forth.
HS: Which year did the Cary conference you mention happen? Was the work presented in this paper, or presented at the conference first?
CJ: The conference was in 1993, the Oikos ecosystem engineering paper came out in 1994, and the book came out in 1995. The engineering paper mostly got written amidst trying to figure out how to organize the conference and getting that going. At the Cary conference, John gave a distillation of the ecosystem engineering paper. I was heading up the conference and so I had a lot of other things to do. Moshe also gave a paper he and I co-authored based on our empirical work in the Negev that explored how to link population dynamics and ecosystem effects using our work with desert isopods as a basis. There were a couple of other papers by people we invited that were about ecosystem engineering.
HS: Did the authors meet often during the writing of this paper, or was it mostly done on the phone?
CJ: The bulk of the work was done when John and Moshe were both present at the Cary Institute for their annual visits; or when John visited to help plan the Cary Conference; or while I was working in the Negev with Moshe. John and I also regularly communicated with each other when he was in the UK as part of our conference planning. I would say this paper was mostly catalyzed by one-on-one discussions followed by drafts and notes exchanged between us. Of course we also had phone calls, but I think the physical presence of the three of us was necessary and important because we were struggling to crystallize what we meant. It really helped to be conversing, drawing all sorts of figures and conceptual frameworks on the board, and making a list of all the examples we knew about from the literature that may or may not be engineering. At the same time, it wasn’t something where we could just get together for a couple of days and bang it out. I also don’t think it was very amenable to what today is a common way of writing papers – exchanging endless drafts among large groups of people and getting comments and feedback via email. Maybe it was representative of that era, but maybe it was also representative of the challenge. We knew what we wanted to do but we didn’t know exactly what that was. That meant there was a lot of iteration, going to and fro to try to crystallize it.
HS: Did you do most of the writing for the paper?
CJ: Mostly John and I. Moshe and I did most of the work to generate the formal flow diagrams. I also had to assemble and integrate all of that with our writing. John was, and still is, a voluminous reader. He did a fabulous job of digging into all sorts of literature, finding interesting things that people had said that maybe related to what we were trying to say. He has a tremendous capacity for reading, absorbing and digesting literature. Although Moshe and I also read a lot of stuff, John was the key to getting all the literature together. Moshe was the key to helping structure the conceptual framework. The physical writing was mostly John and I, and the editing was mostly me and John with, of course, Moshe contributing. It was very much a shared collective endeavour.
HS: Staying on the topic of the writing process – would you and John actually sit together and work on drafts, or would you write something and share it with John, and then he would come back with comments and so on?
CJ: I seem to recall that we would have a general discussion of what bit of the paper we wanted to cover. Depending on the circumstances, this would be a section that either John or I would try to draft first and then share it amongst the three of us.
HS: Do you remember how long it took to write this paper?
CJ: Not that long. Once all the thinking had gelled, it probably didn’t take us more than a couple of months to assemble what we thought was a good working draft. The time for writing the paper was far less than the time involved in trying to conceptually construct what we wanted. Once we had a clear idea of what we wanted to say, it sort of wrote itself.
HS: In Table 1 you list a lot of examples that were potentially cases of ecosystem engineering. Did you specifically do a targeted literature search for this paper, or were these mostly examples that you and the other authors had encountered over the years?
CJ: It was a combination of the two. Obviously, we started with things we already knew at the time. That included some of our own work and some obvious cases from the literature. But we wanted the coverage to be as broad as possible with respect to the kinds of physical modification, kinds of organisms, and kinds of environments they occurred in. So what we would do is look at our draft table and ask “what are the big obvious gaps?” I’m making this one up because I can’t remember a real example from back then. Suppose there was nothing on microbes, we would say “there’s got to be engineering by microbes”, so let’s do some digging on that. I think Moshe used the phrase “a supermarket of examples”, which I thought was very apt. We strongly felt that we had to go beyond just – “a beaver does this”. We needed to show that physical modification of the environment was pervasive and widespread with regard to environment and taxon. That is what really drove our desire to find more and more examples. When one of us came up with a new example we found in the literature it would then catalyze further searches for stuff that might be related to it. The literature prior to the 1994 paper was actually full of examples of organisms doing this. The key challenge really was to figure out how it might all fit together.
HS: How did you actually do the literature searches? Did you have computer bibliographies then, or did you have to plough through hard copies in libraries, searching back from reference lists in papers etc.?
CJ: I’m struggling to try and remember that. I know that we had access to online searching well before that paper at our library. We didn’t necessarily do it ourselves, but we could file a search, some search terms, and see what came up. We couldn’t do what you can do today – go to Google Scholar, type in “Ecosystem Engineer*” and see what comes up. The main problem in the searching was that there was no conceptual framework or label that we could peg this on. We could do specific searches like for “digging” or “nest building”, things like that. So, we had to sort of go bit-by-bit. The other thing we used and I still use, is a tried and tested method of literature searching. You read a paper, and then find all the potentially relevant literature that paper cites, and then keep doing that, going further and further back until you are not finding anything relevant; maybe pre-Darwin! I think that the time and effort we devoted to finding the examples in the literature was probably a longer and more torturous process than would be the case today, but on the other hand, I think it gave us a much more thorough understanding. We had to read a lot of stuff. We read a lot of stuff that hadn’t been read for a long time, but when we did find stuff that was relevant it really informed our thinking. From the collective assembly of all this information it was possible to start to build a conceptual framework that dealt with the key factors determining whether engineering occurs, what sort of effect it’s going to have, how big the effect is going to be, etc.
HS: Was the phrase “ecosystem engineer”, which you coined in this paper, an obvious choice right from the beginning? Did you consider other options?
CJ: No, it wasn’t an obvious choice. I remember that conversation. It was interesting. Both Moshe and I, being more oriented toward ecosystem ecology, said it has to have ecosystem in it. John agreed. But then we wondered – “Ecosystem what?”. It was John who said, “Well, it’s engineering; they are building stuff.” We thought that was short and catchy. I think after some more conversation, because we also considered other complicated options, it just crystallized as a simple, clear, punchy term. Again, like the rest of the paper, it was a collective effort – John came up with “engineer”, and Moshe and I came up with “ecosystem”.
HS: What about the definition, was there a lot of discussion around that?
CJ: Well, that was endless. So we had this vast list of examples and these flow diagrams that Moshe and I worked on. We then asked – what is common to all of these? Are they similar or are they different? If engineering is not trophic ecology, where is the demarcation? How do we exclude the trophic from the definition? We went round and round on that definition. In retrospect, it isn’t the world’s clearest definition, but I think that is a reflection of the fact that we were, in a sense, trying to define what it was not, especially in relation to the trophic paradigm. People have interpreted the definition in different ways and rewritten versions of the definition and so on. To me, the key part of the definition was the physical modification of the environment by organisms and the consequences thereof. In the 1997 paper, we made some modifications to the definition because, in the first paper, for some strange reason that I don’t recall, we excluded ‘living space’. I don’t know why we did that. I wish I could recall why we excluded it. We probably had a reason that seemed logical at that time, but of course that was wrong.
So, in the 1997 paper we included living space explicitly. Maybe we were trying to put boundaries around engineering so that people didn’t think it encompassed everything. Like with all such things, you have a go the first time, and then when you reflect upon it you realise it isn’t entirely clear. When I re-read the definition today I notice that there are clear, consistent aspects to it. We characterised non-trophic as “other than themselves”, which is an okay way of excluding the engineer as a consumable resource. Also, in the first definition we sort of focused on resources in the traditional ecological sense and we didn’t really get into environmental conditions, which are not resources, even though the examples included conditions. I think that was probably our attempt to keep things simple. However, it’s abundantly clear that it is not just the concentration or dissipation of resources, but also the altering of environmental conditions, including those to which an engineer might be completely indifferent, that can have profound effects on organisms. Fortunately or unfortunately, depending on how you look at it, some aspects of the original definition have stuck. Depending on their own perspective, users sometimes rephrase it in different ways. For instance, habitat creation – we did highlight this – is often emphasized as a defining feature. I think that is pretty good if you take a broad definition of what habitat means. But what ‘habitat creation’ doesn’t tell you is how it occurs. So, if you say habitat creation by physical modification of the environment by organisms, and you add a bit of non-trophic in there, you have the elements of a definition as non-trophic physical habitat modification by organisms. An interesting aside is that one of the other words that has come up in the literature is ‘biogenic’, which just means made by an organism. Poop is a biogenic product too! So, yes it was a challenge to come up with a definition that adequately connected the superficially disparate set of examples, while making what we were talking about sufficiently distinctive from, for example, trophic interactions.
HS: Did this paper have a relatively smooth ride through peer-review? Was Oikos the first place it was submitted to?
CJ: It was an extremely smooth ride. The paper went to Oikos, the reviewers were very positive, the editors were very positive, we had to make some relatively minor editing changes and clarifications, and that was it. In retrospect, it is interesting that it just sailed through, given some of the controversies that came later.
HS: In the Acknowledgements, you say this was Publication No. 171 from the Mitrani Center. Could you tell us a little more about the Mitrani Center?
CJ: The Mitrani Center for Desert Ecology is at the Jacob Blaustein Institute for Desert Research in Sede Boqer, and part of Ben Gurion University of the Negev. That’s where Moshe came from and that is the organization that hosted me in my numerous visits during our research there. Publication No. 171 just means it’s the 171st paper from that Center. The Center was a wonderful place to be, with a really nice group of people in a fabulous location deep in the Negev desert. It is now called the Mitrani Department of Desert Ecology.
HS: You thank three agencies for funding support – the Mary Flagler Cary Charitable Trust, the US-Israel Binational Science Foundation, and the NERC Centre for Population Biology, UK. Were these grants specifically for this piece of work?
CJ: I now notice that NSF is not included in the paper acknowledgements, although they helped fund the Cary Conference. I guess it was the timing of the work, and for some reason we thought the funding was not that closely related to the paper itself, even though it sort of was. I don’t know. The Mary Flagler Cary Charitable Trust provided an endowment that helped support the Cary institute. That money was used to support my salary at the time, as well as travel and expenses for John and Moshe. The NERC Centre for Population Biology was headed by John and also provided him some funding support. The US-Israel Binational Science Foundation funded a lot of the work we did in Negev on the rock-eating snails and the intellectual extensions from that work.
HS: Did this paper attract a lot of attention when published? Was it considered controversial at that time?
CJ: I don’t think there was a tremendous amount of attention when it came out. The Cary Conference certainly garnered attention, and so it is hard to separate the two. It received some attention, including from Jim Brown, but there was no real controversy at that time. The early citations were mostly us trying to push it; John did a very good job of marketing. We got positive feedback from people, but this wasn’t one of those papers that suddenly became very highly-cited. It was a gradual process.
HS: Did the idea of ecosystem engineering gain acceptance slowly over the years?
CJ: That’s an interesting story. I guess around the year 2000 was when it started getting a lot of attention, as articles and other papers on the topic started coming out. That’s when it started to catalyze a greater interest, and at the same time, various criticisms. Some people didn’t like it, or at least the term ecosystem engineering. For example, Mary Power thought it was a great paper, a really interesting idea, and important, but she hated the term engineering. She felt engineering implied intent. That is an interesting issue. A lot of subsequent work on the evolutionary implications of engineering has asked if engineers get a feedback from their engineering. Is there a positive feedback to the engineer, for example? Maybe, but it doesn’t mean that they intend to do it. In a sense, it was a sort of a straw man distraction, but it is interesting that that notion of intent has persisted in the literature. We never meant to imply intent at all, but some people construed it that way.
Other people also didn’t like the term we coined, but for different reasons. They asked why needed a new term, didn’t we already know this, what do we get out of this, etc. In a sense, they were saying – you guys are just reinventing the wheel; we knew beavers built dams! Our response was that, while we fully recognise a vast literature full of examples, what we are trying to do is provide a conceptual foundation that allows us to compare and connect all these examples. A good example of this sort of criticism was when a former graduate student of mine, Justin Wright, submitted a paper to Ecology about beaver effects on species richness. One reviewer nastily said “I learnt more about beavers by going to my children’s library than I got out of this paper”. But the reality was that although it seems superficially obvious that if you create habitat you create space for some species, nobody had quantified it. This was the first study that actually quantified, across scales, how many species depend on the activities of an archetype animal engineer, the beaver. That paper eventually got published, has ca. 500 Google scholar citations today, and catalyzed a tremendous amount of work quantifying engineering effects on species diversity. It was, in my opinion, a very important paper. It wasn’t just about the beaver, it was about how you approach the problem of quantifying the effects of engineers on diversity given the habitats they were making or destroying.
Here is another example. An ecologist submitted a paper to Ecology with “ecosystem engineering” in the title. It was a study that perfectly fit the ecosystem engineering definition. She included dynamic models to understand how seagrass beds altered hydrodynamics, sedimentation, and nutrient trapping. She approached it by thinking of the seagrass bed as a physical structure that interacts with its environment. Very nice piece of work. The editors published the paper but only after she took “ecosystem engineering” out of the title. She was allowed to keep it in the keywords!
There were a number of pieces written pro and con the ecosystem engineering concept. Arguments about the breath of the concept have often been made, but the value of its breadth has been well articulated by others, as well as by me and colleagues. After all, “eating and being eaten” is a universal in ecology, but no-one has had a problem with that concept being too broad. Recognition of its ubiquity has helped advance understanding of predator-prey interactions, not impeded it. The same can be said for ecosystem engineering. One of the other big problems that many focused on is the issue of effect magnitude. Our approach came fundamentally out of a mechanistic and process based perspective. A priori, there is no reason to expect an organism to have a large effect or a small effect. You should not base your classification of an organism as an engineer based on the size of its effect. That approach doesn’t give you any predictive or explanatory power. We definitely wanted to stay away from that problem as we felt it was a conflation of mechanism with effect size. We realised that the Holy Grail for ecologists is to be able to predict important and influential species and interactions. But the question is how do you best approach that challenge? We approached it by saying – if you understand what is going on, what the underlying process is, what the mechanisms are, what the pathways of interaction are, maybe we will do a better job of predicting whether or not it’s going to have a large effect. People also didn’t like us using wimpy engineers as exemplar! My favorite example – I use it in lectures – is moving animals cast shade! Or bird nests almost certainly affect local wind turbulence! But neither of those engineering effects of physical structure have any larger meaning. Why? Because the shade cast is completely ephemeral and small relative to the natural variability in shade in any particular area, and there are few if any organisms that care about wind turbulence outside a bird’s nest. But the idea that you should exclude all modifications that do not somehow translate into larger effects was and is conceptual anathema to me. I don’t think that’s the pathway toward greater understanding. I want to be able to predict engineering effect magnitudes, not just label them as such post-facto. Ideally, I want to predict exactly what biotic variables it will affect and how big of an impact that will have on that engineer and other organisms that live in that environment.
HS: This paper has been cited over 4000 times. Do you know what it mostly gets cited for?
CJ: I haven’t done a recent analysis but I have periodically looked at that. I think there are two basic categories. One would be what I would call citations for the core science. These really are papers about ecosystem engineering. They are studying it, figuring out how it works, and what effects it has; and/or they are integrating that process with other ecological interactions. I don’t know what percentage that would be, but it is a significant portion and it is growing. The second category is papers that label a particular organism that does something as an ecosystem engineer, or just use the term, without delving into it too much, often without demonstrating that it is ecosystem engineering in the strict sense. It is a convenient label to use, right or wrong, and it is, unfortunately, fashionable to use the term, even when it may not be warranted!
HS: What kind of impact did this paper have on your career and the course that your research took subsequently?
CJ: Huge! It has been the major emphasis of my career from late 1980’s onwards. Although I have done a lot of other things – my research on plant-herbivore chemical interactions, plant herbivore microbial environment interactions, decomposition and stress, evolutionary aspects of chemical diversity – they have all steadily waned over time. I continued to maintain some long-term collaborative interactions with my colleagues at Cary, most notably Rick Ostfeld, Charlie Canham and Gary Lovett, that came out of my work on the gypsy moth, which I call “the acorn connections”. These are connections in oak forests between oaks and acorns and deer and ticks and mice and chipmunks and Lyme disease and gypsy moth – a network of interactions most of which are trophic. But that apart, most of my research in the last 30 years has been on ecosystem engineering, and it obviously has had a major influence on my career, on the visibility of my work, funding opportunities, invitations, and all the other sort of stuff you get out of a body of work becoming influential. With the exception of a brief period in the midst of all the controversy surrounding engineering, I have steadily worked on the topic for a great many years, before and after it was called engineering. During the period it was controversial I backed off a bit. I wasn’t sure if I should continue it or not, but in the end I decided to go ahead. I recently retired, but engineering has been a primary focus of my work for over 30 years.
HS: Around when was it when you decided to back off a bit?
CJ: Around 2000. Late 90s, early 2000s. At the time, the exposure and visibility that engineering was getting caused a bit of a backlash. I kind of withdrew because, at the time, I didn’t really want to get too mired in the issues. I’m not the sort of person who likes to have arguments with other scientists that risk descending to the personal; I prefer to focus on the core scientific issues. Apart from that brief period, I focused on it and, even today, keep in touch with what’s going on in the field. I retired in 2016, and I’m still working out what I want to do in the future with regard to the science of ecosystem engineering. But a lot of opportunities did come my way as a result of this, from keynote speaker invitations to working groups to international research chairs and visiting professorships. It also created opportunities to interact with new, diverse groups of scientists. Geomorphologists, for example. Evolutionary biologists. Anthropologists. Environmental engineers. The crystallisation and clarification of engineering has opened a number of pathways. Geomorphologists have long studied organisms that do something to the geomorphology. Our work provided a way to integrate and incorporate an ecological perspective. In other words, understand what an organism is doing, how it’s doing it, why it’s doing it, how its dynamics would affect the geomorphology and vice versa. In a similar fashion, this work allowed connections with other groups of people too. The anthropologists who had become very interested in understanding how earlier civilizations modified their environment to harness energy and resources. Or the evolutionary biologists who had become fascinated with the notion of niche construction. The engineering concept has been very influential and important for the people developing niche construction theory – John Odling Smee, Kevin Laland and colleagues – because it allowed them to make concrete, so to speak, the recognition that organisms can make their environment and not just adapt to them. And of course, there are a lot of people now in what I would call green environmental engineering who have developed, I guess, two kinds of approaches that relate to ecosystem engineering. First, simulating what nature’s engineers do, and second and steadily growing, putting nature to work to do the engineering work for us. Oyster reefs and salt marshes for coastal protection are obvious examples of that. So, a lot of connections to areas outside of ecology have developed. Also within ecology, connections have developed with other areas of ecology, in particular, the coupling of trophic interactions with engineering.
HS: At the point when you did this work and published this paper, did you anticipate at all that it would have such an impact?
CJ: I guess we hoped it would. But I don’t think we anticipated how big it would become. I don’t think we ever sat down and thought “how big it is going to be?” We just did it because it seemed interesting and important. I guess we hoped that people would buy into the idea that there was a conceptual core here that linked all these seemingly disparate examples.
HS: I would like to go over specific parts of the paper, to get a sense of where you stand today on what you said then.
HS: In Figure 1 you talk about five possible types of ecosystem engineers. Would that still be the case today, or is that number different?
CJ: That’s an interesting question. The answer would be it could be more, less, or the same number. Those figures were designed to identify pathways of influence involving the physical environment. For any one engineer there might be multiple pathways, in which case core parts of the figure would co-occur. In that figure we also largely ignored the branching ramifications. I think, in essence, you could boil the types down to the two we articulated – autogenic and allogenic. It’s a useful classification that informs expectations about how the system would work. The physical construct itself can be the key feature. The physical construct can then interact with various energy and material flows in the environment. And that has knock-on consequences for other abiotic environmental variables. So, the number of types could be fewer or it could be more, because they can co-occur in different ways. The physical structure of the environment can be a habitat that interacts with energy and kinetic flows to alter the environmental conditions around it that then affects that organism. We never meant this scheme to be a mutually exclusive set of types or a rigid classification. Rather, we were trying to convey – and this was very much driven by ecosystems thinking – the kinds of flows and interactions you should really be concerned about. What is the role of the organism, the structure, and other abiotic variables in that context?
HS: Talking about the feedback from the affected organism to the engineer, you say that, if there is such feedback you expect it to be indirect, involving several intermediate processes and species. Has there been research on this feedback?
CJ: Yes, indeed. There is clear evidence that feedbacks can occur and they have been studied, but perhaps not as much as they should be given their centrality to the dynamics of engineered systems. As for indirect, I guess what we wrote back then was to point out the obvious – that a burrow indirectly affects the occupant. It is an indirect effect in the sense that it is mediated by the interaction of that physical structure with the abiotic environment. The result is a thermally protected environment and it is the thermal protection that feeds back to the organism. The burrow per se is not a thermally protected environment; it results in a thermally protected environment because of its insulation properties relative to the outside environment and the sensitivity of the organism to temperature.
I don’t know what might be the most extended example of an engineering feedback; I would have to think more about that. But the basic purpose underlying the use of the term indirect is to distinguish it from a direct interaction, where the archetype would be predator eats prey. Engineering has to go through at least one additional intermediary, the non-living environment. And when it does that, it invokes physics and energy. The non-living environment can involve a whole host of processes, each of which could affect different species in different ways. There was a paper that came out relatively recently – I got a Google alert on it – I’ll see if I can find that paper – the problem is I can’t remember what organism – anyway, oh yes here we are – fiddler crabs. The paper is in Austral Ecology – “The response of meiofauna and microphytobenthos to engineering effects of fiddler crabs on a subtropical intertidal sandflat” by Citadin et al. The bottom line here is: “The present study demonstrates that the different engineering effects of fiddler crabs are an important source of habitat heterogeneity and a structuring agent of meiofaunal assemblages on subtropical tidal flats.” If you dig a hole there is a pile of dirt next to it, right? The pile of dirt doesn’t necessarily have the same effect as the hole.
In one sense, I probably wouldn’t markedly deviate from the original notions of indirect, intermediate processes and effects on species. I think this is well-supported by research. Some people have focused only on the net effect of the engineer by removing it or manipulating its density without tracing the pathways that underlie the net effect. This can be misleading because the net effect can result from influences other than engineering. A focus on net effects without mechanistic understanding is common in ecology, as is the use of the rather vague terms direct and indirect. Direct, indirect and net very much depend on the pathways and mechanisms. If A eats B, then to me that’s obviously a direct interaction. If A poops and B uses the poop, I’m not so sure what that is. Is it direct or indirect? Nutrient cycling is a good example of a field that does not worry about what is direct and indirect. Ecosystem ecologists just consider a network of flows and connections with positive and negative influences and feedbacks. That’s what an ecosystem is. So you don’t try to parse the ecosystem into only things that are positive or only things that are negative and you try to understand how the net effect comes into being. I think my perspective here is much more aligned with ecosystem ecology than with classic population or community ecology.
HS: Towards the end of the paper you say six factors scale the importance of engineers. Do these six factors still remain the most important, or are there other factors that emerged as important in subsequent research?
CJ: That’s a tricky one. Let’s go through those factors. If I can remember them!
HS: Shall I read them out?
CJ: Yes, let’s do them one by one.
HS: Lifetime per capita activity of individual organisms
CJ: I’m not so sure about the lifetime bit, but I do think that the per capita activity bit – it appears in a lot of our work – is fundamental. Does it have to be life-time? Let’s look at an example. A porcupine digs a hole in a desert, then it moves somewhere else and digs another hole. It keeps doing that during its lifetime. It has dug a certain number of holes before it dies. And there are “N” porcupines in the desert doing the same. If you want to know the number of holes that are dug, the two variables you need are the number of holes per porcupine per whatever time frame – the per capita engineering activity – and the number of porcupines. Life-time per capita activity captures some of that, but what’s most important is the per-capita bit. The next one?
HS: The next one is population density.
CJ: Sure, because engineering effect magnitude in terms of physical modification by a species at any particular scale is the product of its density multiplied by its per capita engineering activity.
HS: The spatial distribution, both locally and regionally, of the population.
CJ: Absolutely. Some engineers have high spatial fidelity and some just wander all over the place. This has a profound influence on the kinds of effects that you are going to get. There’s been a lot of work done on the spatial aspects of engineering that is central to the field. I would definitely include space.
HS: The length of time the population has been present at a site
CJ: That’s an interesting one and it’s a complicated one. It’s certainly very relevant in many circumstances. If the engineer does not move around, the population density stays constant over time, and the per capita engineering activity remains the same, that’s clearly different from an engineer that comes in, digs a hole and never comes back. That, if you like, is the temporal aspect of the engineering and it is just as important as space. If I was to write the paper again, I might change the wording, but the essence would not differ.
HS: The durability of constructs, artifacts and impacts in the absence of the original engineer.
CJ: This is fundamentally important core of the engineering. It is still understudied, especially for highly durable artifacts. If you abandon a physical modification, it will deteriorate physically over time. How quickly it does so, and therefore by implication, how quickly the other environmental effects will disappear, is going to depend on a bunch of things. Durability is the best way of describing all those factors that could influence it – it could be hardness or whatever. The environment in which it is embedded is also critical. If you make a pile of sand in an intertidal area, it’s gone very quickly. So, the durability notion is really important because it deals with the persistence of effects with regard to both the engineer and the other organisms. It captures the essence of the decay of physical structure, which I still think is central. There has been quite a lot of work on that aspect, including a bunch of modelling studies that have incorporated the durability/legacy aspects.
HS: The number and types of resource flows that are modulated by the constructs and artifacts, and the number of other species dependent upon these flows.
CJ: That’s a nice generic statement, but really important! What it recognises is that if you alter the abiotic environment it has meaning to some species and has no meaning to others. It has positive effects on some species; it has negative effects on others. It affects the interactions of some species; it doesn’t affect the interactions of others. In every case, it will depend on what the species cares about. So that last factor is really our umbrella attempt to wrap in the ecological interactions while recognizing their context-dependency. If I was to modify it, I would perhaps explicitly add: “including the engineer”. It was implicit in the original paper. So, in summary, I think those six original factors are still central. In a 2010 paper, also in Oikos, you’ll see that there is a certain degree of expansion and contraction of those original concepts to create time-dependent models. The paper does not have space in it. The physical engineering is the core, but you have to know how many engineers, working how hard, making what, that lasts how long, that has this kind of effects on the environment, that affect what species? I think that core hasn’t fundamentally changed.
HS: You say, “We know of very few field manipulation experiments designed to quantify the impact of ecosystem engineers by removing or adding species”. Have there been many subsequently?
CJ: Oh yes, tons. There’s been a huge body of experimental work on engineering. It has exploded, and subsequent to the paper we also discovered a number of historical experiments in engineering. There have been experiments that manipulate the structure and look at the effects; experiments simulating the engineering, e.g., a plastic tube instead of a burrow; and experiments that add and remove the engineer. Experiments have been very useful in identifying whether or not it’s really engineering that’s causing the effect and what happens to the system if you take away the engineering? It is worth pointing out that interpreting engineer addition and removal experiments requires some care. Engineers do more than just engineer, and so removing the engineer also removes other effects that it has that are not engineering.
HS: You have also partially answered my next question, which is whether we know more today about the influence of the persistence of these products and their engineering on population, community and ecosystem processes. From what you have said so far, my guess is the answer is yes.
CJ: Absolutely, yes. Lots of people are starting to look at the extended influences of engineering in space and time. Yes, that’s been an area of considerable growth.
HS: You say, “if the notion of organisms as ecosystem engineers results simply in an accumulation of ‘just-so’ stories, it will not have been particularly useful”. Again, I’m guessing that hasn’t been the case in the years since this paper, and that this concept has been fairly successful.
CJ: Yes, I think so. What caused us to write that was that we were worried that now that we had a label for something that wasn’t clear beforehand, that what we would end up with is a lot of “My organism is an engineer”. And yes, there has been quite a lot of literature that does that, but that is okay because it is the fodder for synthesis and integration. But if all we had was such a catalogue we wouldn’t have considered the paper very successful. I think the conceptual framework has catalyzed people to try to compare and contrast engineers, try to understand why engineers have effects and where the parallels are elsewhere. So we haven’t ended up with just a stamp collection. We have ended up with a lot more stamps in an ever-expanding album, with emergent generalizations that can be made and are used. So I would say that the research agenda has been pretty successful. It doesn’t mean it is over, there are a lot of cool things that still need to be done.
HS: Have you ever read this paper after it was published?
CJ: I haven’t read it from start to finish since it came out, but I do periodically go in and read something we said or check up on things. That’s because it was a long time ago and my memory is not that great. Sometimes it surprises me. I’m writing something and think – well, that’s interesting, perhaps I should say that. Then I wonder if I said that earlier, go back and check my papers and find that I have. So, there are a number of situations in which I have ended up citing myself saying what I have said before! So, yes, I do periodically reflect on and refer to it, but I haven’t re-read it from start to finish.
HS: When you compare that paper to papers you write today, do you notice any striking differences?
CJ: Not particularly. For me what is striking is that there is a lot in that paper that has not fundamentally changed. What has happened instead is massive enrichment and development. The addition of detail, from modelling to empiricism, and the connections that have then arisen to other areas of ecology and other disciplines are notable. The core message of that paper was that we should focus our attention on physical modification of the environment by organisms because it is pervasive and we have sort of marginalized it. That’s not true anymore, it is not marginalized now. As an ecosystem ecologist, I would have to say that just focusing on engineering alone is not going to get you there, any more than just focusing on trophic interactions will explain all of ecology. But I’m surprised by how much of this is still relevant. In the questions you sent me you asked me whether or not I would ask a student to read it. I would ask her or him to read it and then read something more recent. I guess if a student has read it – I would then ask him or her – did you know that already? If you did, has it in anyway influenced your thinking. Is it already embedded in your mental map of what ecology is about?
HS: Would you count this as a favorite, among all the papers you have published?
CJ: That is a difficult one, because it is compounded by the fact that it is certainly my most impactful paper. It crystallized a large fraction of the rest of my career. It’s very influential to me, but a favorite? That’s an interesting question. I’m not sure what favorite means? It certainly is in within top of the list of papers I wrote. So yes, I would say it is a favorite. I don’t know if it is THE favorite. It reminds me of an old British radio programme called “Desert Island Discs”, where you are asked – “if you were cast away on a desert island and had to take seven record album records with you, and one book other than the Bible and Shakespeare, what would they be?”; then “if only one record, what would it be?” So if I was locked in a cell for the rest of my life and I had to take seven papers that I had authored or co-authored to read endlessly, it would be in the list. But I’m not sure I would pick this one if only allowed one paper!
HS: Let me frame it a little differently: would this figure among the pieces of research you are most proud of?