In a paper published in Ecology in 1998, Mark Ritchie, David Tilman and Johannes Knops reported the results of their 7-year long field experiment to examine the effects of herbivores on nitrogen cycling in a nitrogen-limited oak savanna. The results of their study suggested that herbivores, by reducing the abundance of plants with ntirogen-rich tissues, indirectly reduce the rate of N cycling in the ecosystem. Nineteen years after the paper was published, I spoke to Mark Ritchie about his motivation to carry out this study, memories of field work and what we have learnt since about the effects of herbivores of nitrogen cycling.
Citation: Ritchie, M. E., Tilman, D., & Knops, J. M. (1998). Herbivore effects on plant and nitrogen dynamics in oak savanna. Ecology, 79(1), 165-177.
Date of interview: 12 January 2017 (via Skype)
HS: Your PhD research was on foraging in ground squirrels. Did you switch to working on insect herbivores in your postdoc?
Mark Ritchie: That’s correct.
HS: What was the motivation to do this particular piece of work? This was a seven year experiment. Were you involved in the experiment right from the beginning?
MR: Yes. It was an experiment that I set up for an entirely other purpose, at the beginning of my postdoc, and then after it had run for a few years, we started seeing results that, I guess, we could have anticipated had we thought about it more, but they were pretty dramatic. I originally set up an experiment in which we were going to try and compare insect herbivory with small mammal herbivory in oak savanna. I had created these little 2*2 meter cages in which we could put different insects or small mammals in. That experiment didn’t work very well because the small mammals didn’t cooperate. In some cages, they totally destroyed all the vegetation and in others they didn’t do anything except to sit in there and die. We ended up reducing the experiment to one which was just essentially an herbivory exclosure. We have fenced out deer and small mammals, and sprayed insecticide inside to greatly reduce the insects. At that point in time, no one at Cedar Creek had really worked in oak savanna very much; almost all the people working there had focused on old fields. Not very many people were really studying herbivory, and so understanding pretty much anything about herbivory at Cedar Creek, at that point, was pretty new. After the experiment had run for a while, we noticed that there was this profusion of legumes inside the exclosures. Cedar Creek is quite well known for being nitrogen limited, and there were some hints from some other experiments that herbivores have some control over legumes, so we then decided to focus on that. There was another researcher there, at the time, Johannes Knops, from the University of Nebraska. He was an ecosystem ecologist, and he was thinking, man, the amount of nitrogen in these herbivore exclosures must be huge compared to the outside. At that point in my career, I’d spent almost no time thinking about nitrogen cycling or nitrogen dynamics. And so, we talked to Johannes about what would be necessary to look at nitrogen. I had my team go out and make all the measurements to look at how much nitrogen was in the soil. So even then, it was kind of an experiment in which the story was, “herbivores control legumes, and legumes control nitrogen” and was a good story. But as I started reviewing the literature and started to write the paper about the effects of herbivores on nitrogen cycling, the only thing I could find was the workout of Serengeti by Sam McNaughton that argued that herbivores were accelerating nitrogen cycling by basically converting the plants they eat into dung and urine, which put nitrogen back in the system in a much more usable form that normally exists in the soil. This process meant that plants under grazing would have more nitrogen available for growth than plants that were not subjected to herbivory. Clearly, the opposite of that was happening in the exclosures: by taking out the herbivores, the abundant legumes produced a very nitrogen-rich litter, which decomposed nitrogen available without herbivores. And so, because I clearly recognized that contrast, I decided the conceptual base for the paper was going to be contrasting herbivores decelerating nitrogen cycling versus accelerating it. I really like pushing the theoretical and conceptual background to my papers as far as I can. I thought, well, let’s just draw some diagrams and make arguments about why it would be one way versus the other. Those are the major events and thinking processes that led up to producing that paper.
HS: When those cages failed and you set up exclosures was the plan still to compare mammals and insects?
MR: After the first year, we just decided to abandon the other treatments where we were either adding insect herbivores or adding small mammals, and just switch over to an herbivore exclosure experiment. So yeah, the shift happened pretty quickly in the progression of things.
HS: You mentioned that when you reading up the literature on this, you came across research showing the accelerating effect, by McNaughton. In the paper, you cite other papers that show the opposite decelerating effect. Did you come across that literature later?
MR: Yeah, those were mostly papers that talked about how herbivores were promoting plant species that had recalcitrant litter. There had also been some work done by John Pastor, looking at moose herbivory, and how moose were shifting species composition in the direction of species that had low nitrogen or more slowly decomposing litter. They were shifting from species like aspens and willows to conifers. Conifers have much more slowly decomposing low nitrogen litter than deciduous trees. So there was some evidence for things like that out there, but it wasn’t until around the time that my paper came out, that these other papers started coming out. It was one of those situations where a number of people started noticing these things simultaneously. And so for a stretch of years in that window, there was a student of Michel Loreau, named Claire de Mazancourt, who was doing similar work and she had papers that came out in 1998 and 1999 that got further into details about the nitrogen cycling effects. It’s one of those phenomena that was there waiting to be found by anybody that thought to look for it. At that time in ecology, there was considerable interest in species effects on ecosystems, that systems weren’t just these inexorable physico-chemical systems in which certain materials and energy were either readily available or not readily available. The idea was that, if you shifted the species composition of an environment, then material and energy flows might be very different. This idea had come out in a series of books and other things in the mid-1990s. As people began to think about potential species effects more, I think the effects of herbivores on nutrient cycling was easy to find by comparing mineral nutrient measurements inside and outside fences that already existed. I think also it became a lot easier during that decade to measure nitrogen in the soil because auto analyzers and other equipment became much more available for various research groups around the world. I think that made it a lot easier for people to get the measurements they needed to test for species effects.
HS: Stepping back a bit, tell us how you chose to do your postdoc in this particular study system. Why did you move away from studying squirrels?
MR: I’ve always been interested more in questions and problems than in any particular organism or system. My interest as a PhD student was in optimal foraging theory, which at the time (mid 1980’s) was of huge interest and embroiled in a large amount of controversy. One of the points of controversy was whether or not you could actually show that it could evolve by natural selection. That’s what I set out to do for my PhD. But even as I was doing the work for that project, I was helping my advisor, Gary Belovsky, do experiments with grasshoppers. I did other experiments on my own with grasshoppers, after ground squirrels went into aestivation in mid-summer, which I never published. And then, when I was looking for a postdoc, I was casting a pretty broad net. I had some connections between University of Michigan and Dave Tilman, because Tilman did his PhD at Michigan and his advisors were still there. They were good friends of mine. And so, when I interviewed for the postdoc, Dave asked me some questions about what kinds of things I was interested in. I was definitely interested in this whole issue about herbivores and resources, which was, at that time, and especially in terrestrial systems, completely uninvestigated. And so, it wasn’t that much of a shift to connect foraging by grasshoppers to the more general topic of herbivore-plant-resource interactions. If I had continued with the evolution of foraging behavior, I would likely have gotten deeper into the way in which behavioral cognition and learning allowed foragers to balance trade-offs and make “optimal” decisions. However, with the opportunity to work with Dave Tilman, I decided instead to focus at a higher level of organization on trophic interactions and community dynamics that dominated the classic ecology literature. And then when I got to Cedar Creek and began interacting with people who were working on things like nutrient cycling and ecosystem processes, I began to learn more and figure out ways of making connections between my work and theirs. And then some of it was just pure opportunism, in that certain experiments produced results that forced you to pay attention. But then when you do pay attention, you have figure out what concepts are being advanced by the results.
HS: At what stage did Johannes Knops get involved in the work?
MR: I don’t even remember why he was in the vehicle with me, but we stopped by the site where the exclosures were and I said, “Hey, I want to show you this experiment. It’s pretty cool.” And that’s when he said, “Wow, the amount of nitrogen in this plot is probably incredible; you should measure that.”I was thinking, “OK, yeah, sure, that sounds great – I know nothing about measuring nitrogen mineralization.” But that’s where the great advantage of working at a place like Cedar Creek comes into play. Measuring mineral N in the soils of the plots was relatively easy to do because there were only ten plots. So it was easy to get into the normal processing of hundreds of soil nitrogen samples that went on every summer at Cedar Creek. So that’s how Johannes got involved with that project, and then and as a result of that collaboration we became pretty good friends over the years and were involved in a few papers back in the early 2000s.
HS: Was he already working in the Cedar Creek project when you showed him your experiment?
MR: Yes. He was brought in as a postdoc.
HS: Why did you decide to do these experiments in the oak Savanna?
MR: Savanna restoration was going on at Cedar Creek, due to the fact that a number of people studying peat bogs and lake sediments suggested that the fire frequency in that part of Minnesota and at the prairie-forest border had been higher than what it was now. Suppression of fire led to massive woody encroachment in most of the woodlands in and around that part of Minnesota. And so, Cedar Creek was actually implementing a fire regime there that was supposed to, roughly, reproduce what had happened historically, either by fires set by Native Americans or fires caused by lightning from thunderstorms or whatever. From my point of view, therefore, oak savanna represented more the natural vegetation of that area before European settlement, and before most of it had been cultivated. It was unclear how long the woodlands that were there had actually been woodlands, or whether that was just purely a result of fire suppression. And so, for me, the savanna reflected the coevolutionary history of the region, and so it would be pretty interesting to look species interactions and plant-herbivore interactions in particular. A lot of people didn’t like working in the savanna, because doing small-plot experiments in savanna, often leads to highly variable results, because there’s a lot of heterogeneity. There are a number of things that drive pretty high heterogeneity there. And so, for whatever reason, it just hadn’t been a focus of extensive research. But for me, it was like, “This is as close as we’re gonna get to what this habitat was, and so maybe we can learn something new and different from the well-known dynamics in the old fields.
HS: Right from the beginning, was the plan to do a long-term experiment?
MR: No, it was not. It was supposed to just be an experiment where I would create different abundances of herbivores and see the effect on the plant communities. I expected to have results in a couple of years. It was not intended to be a long term experiment. Eventually, we did run the experiment for, I think, 10-12 years, something like that, before we finally just decided it had run its course. What was interesting is that after about 1995 or 96, we began to have major problems with pocket gophers invading the plots. They shifted the vegetation away from legumes to more woody species that normally would normally only occur in the forests where there were higher soil nutrients. We had no control over the gophers, so I couldn’t experimentally show that that’s what was actually going on. It was more an observation of change in vegetation that was associated with a visually obvious higher density of gopher mounds inside the fences than outside. And because I didn’t really have strong inference with only 10 plots, and I had many, many other things going on by that time, I never really published that. The outcomes in the 1998 Ecology paper is a just testament to the value of long-term research, but no experiment should be continued indefinitely.
HS: You said you planned to finish it in a couple of years, but it went on for seven years. Why did it take so long?
MR: One of the reasons why I let it run a long time is because there weren’t very many plots. By the time I had eliminated the herbivore addition plots that didn’t work, I only had five control plots, and five exclosure plots. And each was only 2×2 meters. By the standards of the field, this was more like an undergraduate experiment. And so, in order to build the power of inference, I thought that maybe the best thing to do was to just let it run. So we let it run through ‘93, ‘94 and‘95. In fact, during the review process – Jim Clark was the associate editor for Ecology that reviewed it – one of the reviewer’s concerns was controlling for time effects or differences in initial conditions, because they were concerned about the small number of plots. By letting it run that long time, I was able to show a time x treatment interaction that proved more convincing. My strategy for letting it run paid off.
HS: Could you give us a sense of what does work involved? What was your daily routine, were you were staying, who was helping you etc.?
MR: The cages were made out of hardware cloth and window screen and required building a metal frame for them to sit on. We had to make them out of metal because of the prescribed fires that occurred. Each plot had a lid that fit on the top, and so we would have to take the lids off and put them in an area that didn’t burn, and then take them back out and put the lids on. That was something that had to be done in the early spring, and then we would take them off at the end of the summer. Perhaps deer ate some things in the exclosure plots during the winter -but it wasn’t ever very obvious that they did that. Sampling didn’t take very much time because there were only 10 plots. I had a couple people who were usually working for me doing various things at Cedar Creek during the summers. We had to sample twice – in June and August- because different plant species reached their phenological peak at different times. It wasn’t an experiment that took a whole lot of effort once it was set up.
HS: When this was happening, were you doing other things?
MR: We were still doing experiments with the grasshoppers in the field cages all the way into 1992.We had experiments where we were excluding birds (as insect predators) and measuring grasshopper and plant responses. I published that work in Ecology in 2000. By 1995, it was the beginning of the time that we were starting to set up the big biodiversity gardens, and so there was a lot of discussion and focus on that initiative. I’d taken my first faculty position in 1991, so I wasn’t out at Cedar Creek all through the summer from 1992 onward, so I had to set clear priorities for what the team would do each summer. I would just come in for an intense two or three week session of sampling. We were also doing some small mammal monitoring around that time. Though I did have a crew there all summer, many experiments required me to be there to get measurements done right. So yeah, there were many other things going on.
HS: Can we go over the names of the people you acknowledge to get a sense of who these people were and how they helped?
MR: Sure.
HS: Andrew Guss
MR: He was an undergraduate field assistant. He was paid. I was faculty member at Utah State. He was helping me during the summer when we did a lot of vegetation sampling and nitrogen sampling.
HS: Bridget McCann
MR: She was another undergraduate from Utah State who came out to work for me at Cedar Creek. She eventually graduated as the top student in the College of Natural Resources at Utah State.
HS: Will Pitt
MR: Will Pitt was my PhD student who did a thesis at Cedar Creek looking at effects of different predators (birds, lizards, spiders) on grasshopper communities.
HS: Susan Shaw
MR: Every year, the Long Term Ecological Research Project would hire a group of interns. Cedar Creek calls them interns, but they’re basically just field assistants to go out and do various manual labor activities like weeding and measuring things, and sorting clipped plot biomass to species. She was one of those. When we were doing some intensive measurements, we were able to use the LTER field crew for two or three days.
HS: Gary Belovsky
MR: Gary Belovsky was my PhD advisor. I’m sure I had some conversations with him about the whole issue about herbivores and nutrient cycling.
HS: Dave Wedin
MR: Dave Wedin was a PhD student of Dave Tilman’s who was working at Cedar Creek and, at that time, was really interested in the details of nitrogen cycling and plant nitrogen budgets in grassland habitats at Cedar Creek, and so we definitely consulted with him in what we’re doing.
HS: Did Dave Tilman and Johannes Knops get involved in doing the experiments, or was it mainly in discussions about design and at the analysis and writing stage?
MR: Dave Tilman was involved in the design of the experiment. When I was setting up the original experiment, as a postdoc, there was a collaborative discussion about what we would set up. He was totally happy with the decision I made to abandon the earlier plan and leave the exclosures up. And I think he came out to look at them once per summer or something like that, which was not insignificant, even though it might seem so. After that, I just kind of kept showing him what I was finding. And then, I think it was in 1994, or 95, there was quite the profusion of legumes inside the exclosures by then, and he came out and said, “You don’t even need statistics to analyze this; you just need photographs.” When I wrote the paper, he reviewed it and made lots of comments on that, and so he definitely assisted in that way. And since he was paying for my postdoc, it was one of those things where he was likely to be a coauthor on it, no matter what happened with the experiment. Like I said earlier, we started having conversations with Johannes Klops when we began to be interested in the nitrogen cycling aspect of the experiment. And, of course, he made comments on the paper once it was written.
HS: In the paper, you say that you expected the white-tailed deer to be particularly important in selectively reducing plant species. Was the white-tailed deer one of your main species of interest in the original experiment you had planned?
MR: Actually, it wasn’t. In the original experiment, the treatments were what we could put in the cage, not what was going on outside the cage. Except that we did have a control just for the sake of control. But based on everything I had seen driving through Cedar Creek, I hadn’t seen overwhelming evidence for high deer herbivory, even though deer were pretty abundant there. Cedar Creek originally was a preserve because it had a huge white cedar swamp, which was a deer wintering area. I think, historically, it always had reasonably high deer densities in around the area. It was difficult at that time to really sort out how intense the herbivory was. And so, that just wasn’t my first thought about what the most important impact was going to be.
HS: When did you realize that it was an important herbivore in the area?
MR: I started looking at some other experiments. Previously, The Minnesota Department of Natural Resources had set up exclosures in the forests surrounding the cedar swamp, and then in the early 1980’s, Richard Inouye and Nancy Huntly, who were initially postdocs on the Cedar Creek LTER and then later co-PIs based at Idaho State University, built exclosures in wooded areas along the edges of some of the old fields. Some of them were brand new fields, and some had been already been established as fields for 15 or more years. We actually found some interesting results in the brand new fields which showed that both deer and pocket gophers were strongly inhibiting secondary succession. But the older fields had different results. There appears to be a kind of resistance to tree establishment imposed by the old field herbaceous vegetation once perennial grasses become dominant in succession, and little tree recruitment occurs whether there are herbivores or not. This “resistance” to herbivory and to colonization represents hysteresis in the sense that the history of a field influences whether herbivores are important or not. got thoroughly established with perennial grasses, then it was difficult for trees to recruit into them, regardless of whether there were deer or not. But early in succession, the deer had a pretty strong effect on recruitment, and one of the species that was involved in early succession was an invasive species called buckthorn. I analyzed and wrote up a paper, but I could never get it published because there was a treatment missing (removal of deer only). So, we did look around for other types of herbivore effects. And I had a student who did a MS thesis on deer effects on soils and plants and earthworms and stuff like that. None of the effects were especially dramatic but the sample sizes were really small. We have deer fences set up over a lot of Cedar Creek, but in many areas there’s just not a lot going on, at the moment, or at the time when the experiments were set up. Other than the legume story, which we saw repeated multiple times in old fields and in larger 30 x 30 m deer exclosures in the savanna. It’s just one of those things where systems have certain kinds of strong effects, and if you get on top of those, you get results.
HS: After this study, did you continue working in the Cedar Creek area for a while?
MR: Yeah, I worked there until 2005.
HS: When was the last time you visited the place?
MR: 2005.
HS: Do you have the sense of how this 0.2 hectare opening in the oak savanna where you did this study has changed since then? Does it look more or less the same today?
MR: Yes, and I know this because, just out of curiosity, I was looking to see if the same deer fences that had been up when I was working there, were still up. You can go to Google Maps and see quite clearly. You can clearly whether those fences are still there. That savanna looks pretty much just like it did then, which doesn’t surprise me because it was, by management, designed to be burned two out of every three years, and so that tends to keep it open.
HS: What about the larger area within which this patch is embedded? Is that also the same in terms of the community of herbivores present?
MR: Yes. It’s still dominated by white-tailed deer. Small mammal populations are very low. The vegetation is very low in nitrogen, because of the nitrogen-poor soils. Insect herbivory is hit or miss depending heavily on climate variation. If you get warm summers, you can build up insect numbers that, the following year, seem to do interesting things, and then you’ll get to cool years where they don’t do much of anything. They’re just sort of there, but they have very little impact. A lot of our attention also shifted over to working in the big biodiversity plots after about 1999. A lot of those papers that I was involved with involved doing lots of sampling in the big biodiversity plots. The place is pretty much the same now as it was then, mostly because it’s just very nitrogen-poor. There have been some other people who did studies on woody plant succession, like looking at the rate at which the edge of the forest would encroach into a field, and calculated that it was about the same as continental drift, meaning very slow, and a few centimeters a year. The main driving force there is just the total lack of nutrients and the strong competition for resources and light that inhibit woody succession.
HS: In your study, you identify Lathyrus venosus as being the dominant legume there. Is that still the case today?
MR: Yes, I believe so. There are a number of other legume species. In old fields Lathyrus is not what’s dominant, but in the savannas it’s by far the most dominant.
HS: Do you remember how long you took to write this paper and when and where you did most of the writing?
MR: I wrote the first draft of the paper during spring 1996. I remember getting comments from Dave Tilman, and I think, in June of 1996, which was when I actually submitted the paper.
I don’t remember spending a lot of time doing it. I think I wrote the whole thing in about two months or something like that. Once I got this idea of comparing decelerating versus accelerating nitrogen cycling, then I was off to the races. At that time, Ecology had a long time lag between submission and publication. Sometimes it was like 18 months, if not two years, between when you submit and when it would get published.
HS: Would you say this paper had a relatively smooth ride to peer review? Also, why did you decide to submit this to Ecology?
MR: I had an earlier paper in Ecology looking at various factors that limited legumes at Cedar Creek. I’d gotten some feedback from Peter Vitousek that he really liked that paper and so I thought, well, if I got that one in Ecology, then, you know, I’ll submit the next one to Ecology. At that time, Ecology Letters was just barely getting started. There weren’t a whole lot of other high profile options, and I thought it was a pretty cool paper. Dave Tilman really liked it. He told me, this is a wonderful paper or something to that effect. He was really happy with it. Based on that, I had the confidence to go for Ecology. The only real issue I remember being hung up on in review was this issue of small number of plots and possible influence of different initial conditions. But then I just wrote Jim Clark and said, well, looking at the year x treatment interaction is what I want to do about things, and he said, that seems like a reasonable approach. So I did, and then it went through. Most of the delay was just getting reviews and it sat in press for, I don’t even know how long, eight or nine months, something like that.
HS: At the time when the paper came out, did it attract a lot of attention?
MR: I think it did attract a lot of attention. Like I said, this whole issue about whether species and trophic interactions could influence ecosystem processes was a hot topic at the time. Anything with Tilman on as a coauthor was more likely to get read by other people. Not that that was strategically planned; it was more usual reasons for having him as a co-author. But I’m sure that influenced how much attention it got. And then, like I said, there were a number of other people that were starting to look at the influence of herbivores on nutrients. And, because it was the opposite argument to what McNaughton had been saying about nutrient cycling, there were lots and lots of people around the world at that time that were in a bit of a fight about things like grazing and its effect on ecosystems. People who worked in the western US thought grazing was bad, while people in Europe thought grazing was wonderful. So that controversy, in part, shaped a lot of what I did between 95 and 2005. A lot of the papers that I did with Han Olff were driven by this idea that the environmental context in which you are studying a trophic interaction will determine what effects that trophic interaction has. We were able to make arguments like, for example in the western US livestock were having negative effects because of fundamental site characteristics and not having to do with anything about cattle per se. And then in Europe, which had high precipitation and high nutrient soils, cattle were important in maintaining plant species diversity, whereas they seem to reduce it in other places, and so on. So, in that context, we developed this idea that, in low nutrient environments, herbivores might decelerate nitrogen cycling by being selective for the most nutritious plants, of which N-rich legumes qualified, whereas in more high nutrient systems, plants would be able to be tolerant of herbivory and therefore nutrients or herbivores would accelerate nutrient cycling. That simple hypothesis was pretty appealing to people who were trying to understand why experimental results varied so much from place to place. To provide some additional context, the whole idea of doing field experiments in ecology only really caught on in the mid 1980s. So we were only six or seven years, really, into the whole idea of going out and doing manipulative field experiments and comparing results across different sites. There were a whole lot of possible effects that hadn’t been explored at that time. I just happened to hit on a number of different effects that were important to people at the time.
HS: This paper has now been cited over 400 times. Do you have a sense of what it mostly gets cited for?
MR: Typically, it gets cited when someone writes a sentence that says, herbivores can have important influence on nutrient cycling in grasslands. The specifics, the results focused on legumes, and the deceleration versus acceleration idea, get much less attention. Somehow, it got onto the list of four or five papers that people think about when they think about herbivores and nutrient cycling.
HS: Did this paper have any kind of direct impact on your career?
MR: Yeah, sure. Most people don’t know that, as a result of doing a sabbatical in the Netherlands in the fall of 1996, and then the subsequent TREE paper and this paper that both came out in 1998, I was then invited to join the Netherlands Institute of Ecology advisory board. And so, I spent a lot of time shuttling back and forth between the US and the Netherlands. And then there was a job opening as the head of one of their research sites. I applied for it and was basically offered the position but ended up turning it down, mostly because of the accident of the exchange rate of dollars to Euros at the time was in favor of dollars, and then later changed. But as a result of that job offer, I went back and asked to go up for a relatively early promotion to full professor, and the committee turned me down just because it hadn’t been enough years. So then, I was actively job hunting, and I was at a conference in 1999, I think, and Sam McNaughton came up to me and said, “We love that paper, we use it in all our graduate seminars.” And then he said Syracuse would have a job opening in about a year and to keep my eye out for that. That basically set the stage that allowed me to move from Utah State to Syracuse, and to pick up Sam’s Africa program, which is what I’ve been busy doing over the last 15 years or so.
HS: Did you follow up on this piece of work through other research?
MR: I had some long-term herbivore exclosures at a field site in Utah near Logan. It was on a private ranch that was involved in doing high-frequency rotational grazing to try and simulate migratory grazing systems of wild herbivores. And the ranch management people were concerned about the relative impacts of wildlife and livestock grazing, and so I got involved in doing experiments related to that. But I did do some measurements of nitrogen cycling and whatnot, but the semi-arid ecosystem in which the fences were built is very slow to change – It took nearly 15 years for the herbivore treatments to have measureable effects. And then, like I said, at Cedar Creek, we started getting really interested in the whole biodiversity-ecosystem functioning experiments. I was heavily involved in their work on the response of insect communities to manipulated biodiversity. The other development in my career also kicked in about then. In the early 2000s, I had done this work that led to the paper in 1999, in Nature, on fractal geometry as a way to quantify heterogeneity in nature and the consequences it had for species coexistence and biodiversity patterns. The initial paper was in 1999 in Nature, and then eventually I published a book in 2009, in the Princeton Monograph in Population Biology series. So I was spending a lot of effort, at that time, doing various experiments and stuff related to that species coexistence and spatial heterogeneity. I moved away for a while from the whole herbivore -ecosystem functioning topic. But then, when I started working in Africa, I started doing work that relates grazing to soil carbon and to nitrogen fixation. A lot of what I’ve been doing recently is related to that.
HS: Today, would you say that the main conclusions from this paper still hold true, more or less? I’ll read out what, I think, is the takeaway from this paper:“Our results support the hypothesis that herbivores indirectly decelerate N cycling by decreasing the abundance of plant species with nitrogen rich tissues.”
MR: I still think that prediction generally holds true. I’ve more or less kept up with that field, about what people are finding, and there’s been some extension of that to environmental gradients such as water availability or nutrient availability. It was extended in some work that was done by a postdoc named Elisabeth Bakker. She had a paper in Ecology Letters that talked about the herbivore effects on diversity depending on the soil nitrogen gradient. Effects on diversity were related to whether herbivores are being selective or not, and such selectivity relates to the underlying soil nutrient supply. Elisabeth’s paper has also been cited a lot more times than I would have expected. But other papers have kind of morphed the nutrient gradient into a productivity gradient, and found some exceptions to the prediction. But I think that’s because “productivity” isn’t really very well specified in terms of the supply of resources for plants, and thus it’s unclear how that might affect plant palatability to herbivores. So, I think it’s an idea that still is kind of in place, as far as what we might expect when we go to different kinds of systems, and comparing the decelerating and accelerating type scenarios. Maybe that’s why it’s been still cited regularly today, because people still feel like the ideas still hold water.
HS: If you were to redo this experiment today, would you change it in anyway?
MR: Yes. If I were to redo the experiment, I would include a treatment to exclude pocket gophers! Because, if I did that, then the acceleration in the nitrogen cycle would have continued unabated, and I have no idea what we’d be finding there. We actually saw the same signal of this in a set of larger exclosures that we had set up in Cedar Creek savannas. At first fenced plots were showing this rapid response of legumes, but then, over time, the legumes seemed to disappear, and my crew was wondering whether it was due to gophers. So we counted pocket gopher mounds, and they were much higher inside the fences than outside. Originally, when we set up those larger exclosures, it was part of an NSF-funded cross-site study in which we were supposed to include a treatment to exclude large and small herbivores, at seven different sites. When it came time to exclude the smaller herbivores, which at Cedar Creek are pocket gophers, to exclude them you have to dig a trench that’s about a meter deep all the way around a plot and bury hardware cloth in the trench. But because the savanna ecosystem was a restoration habitat and there was only a very small amount of it at Cedar Creek, the Cedar Creek planning committee decided pocket gopher exclosures involved too much disturbance. So we were not able to put that treatment in. As a result, we there was a big missed opportunity. So if I had to do it over again, that’s what I would do.
HS: In the paper you say, “Our results come from relatively small plots [] and we do not know how they apply at larger scales.” Has there been subsequent work, either by your own group or by other groups, looking at similar questions at larger scales?
MR: I can’t really pinpoint it. But I’m pretty sure what would have happened if we had been able to do a pocket gopher exclosure was that we would have had a lot more establishment of woody plants, as the soil nitrogen and the total nitrogen in the system built up. But it would have taken a long time and needed a lot of space before you could see such woody encroachment. But the predictions all were that if you put massive numbers and more resources into the system due the nitrogen fixation of the legumes, this would then allow the colonization of taller species that require lots of nitrogen to grow, that are good at shading out other species, and so on and so forth. This would have been a direct prediction of the work that Dave Tilman did in his 1988 book on plant dynamics – and others as well – this kind of a common model of plant succession that kind of gets inhibited at Cedar Creek because of the extremely low nutrient levels in the soil. That statement was just a qualifying statement I put in the paper, because it was pretty obvious that these were really tiny plots. So, there could be a completely different vegetation dynamic if you had more space and spatial heterogeneity that you are working with. But, exactly what would happen, I’m not sure, because I just haven’t seen a system like that, in which we got that much of a rapid response of legumes to exclusion of herbivores. To some extent, it was kind of a perfect storm in the sense that, it was a cool savanna maintained by fire and therefore was capable of supporting plant species that have a high dependency on water. And then it had a high deer density, so there were pretty strong effects outside the fence. And then it turns out there are papers that show that fire actually helps recycle phosphorus6, which may also favor legumes. Then there are lots of other organisms that attack legumes, including insects and soil pathogens, and I’m not sure to what extent the cold climate would inhibit those as well. So, again, there are a number of reasons why the experiment would have produced an unusually strong effect.
HS: You say you analyze “plant tissue and soil from each plot for total N and C [] with a Carlo Erba autoanalyzer. Do you still use such a device today?
MR: Yeah, for doing total nitrogen of plant leaves and soil. I’m not sure if it’s an updated version of that by now. It’s basically an analysis based on combustion. You put the sample in a little can and superheat it, and then based on the combustion and emission of gases, it detects how much nitrogen was in there.
HS: In the last sentence of the paper you say, “A major avenue of future research will be to determine the combination of resource limitation, soil fertility, and disturbance that leads to accelerating vs. decelerating herbivore effects on nutrient cycling. To what extent do you think this has happened in the last 18 years?
It has become, at least in the mammalian herbivore “universe,” a pretty strong research question. Like I said, there were a number of different labs around the world that were investigating this. Since 2010, the Chinese have become the dominant research groups in the field. They’ve essentially redone almost every experiment that’s been done in the LTER network in the US and found largely similar results in similar kinds of environments. It will, probably be fun to analyze that someday, but won’t be me. I’ve just got all kinds of other stuff going on. Even in my own work, I was dominated by that sentence for the next eight years, until I sort of shifted over looking more at species diversity and fractal geometry. But even now, people are submitting papers in PNAS and Nature talking about how environments drive gradients of plant tissue nutrients, and how many different kinds of species and what sizes of species and what their impacts on grazing and soil carbon are. All of that is kind of an extension that came out of the Cedar Creek work.
HS: Have you ever read the paper after it was published?
MR: Not the whole thing through, I’ve looked at the results few times.
HS: Would you count this as one of your favorite pieces of work?
MR: Yeah, sure, I would.
HS: Why? What do you like about it?
MR: It was a ridiculously simple little experiment that happened to have a really strong result. I feel most ecologists would have just kind of published it as, a wow, big effect on legumes. But for me, I was looking for a deeper conceptual understanding and what it was an example of, and I was able to articulate that. That’s one of the things that I really liked about it.
MR: What would you say to a student who is about to read this paper today? You know, would you guide his or her reading in some way? Would you them to other papers they should read along with this? Would you add any caveats to keep in mind?
MR: I would probably try to give them some context, some of the same things we’ve already talked about, like where the field was at the time, where people’s understanding was, what the available literature was. And then, maybe point to a few more modern papers that have built on it. It’s become one of my objectives in graduate education to try and build context. So many people only read what pops up on Web of Science, which only really goes back to about 1995. There’s this whole massive classic literature that most grad students are completely unaware of. And, consequently, they have no context in which to view anything new, and have no points of departure for even for their own work. So, it’s been a major point of emphasis for me, in seminars and things like that, to try and provide that context.
0 Comments