In a study published in Ecology in 1987, Stephen Carpenter and colleagues carried out whole-lake manipulations to show that primary productivity in lakes is regulated both by trophic interactions and abiotic factors. Their findings were considered “revolutionary” because, at the time, lake ecosystems were thought to be primarily under the control of nutrients. Thirty years after the paper was published, I asked Stephen Carpenter about the motivation and making of the study and what we have learnt since about regulation of primary productivity in lakes.
Citation: Carpenter, S.R., Kitchell, J.F., Hodgson, J.R., Cochran, P.A., Elser, J.J., Elser, M.M., Lodge, D.M., Kretchmer, D., He, X. and von Ende, C., 1987. Regulation of lake primary productivity by food web structure. Ecology, 68(6): 1863-1876.
Date of interview: Questions sent by email on 9th December 2017; responses received by email on 15th December 2017.
Hari Sridhar: I would like to start by asking you about the origins of the experiments described in this paper. From looking at your profile, I came to know that your interest in lakes started as early as when you were a Master’s student at U. Wisconsin-Madison. The interest in food webs in lakes however seems to have come later, in the mid-80s. Could you please trace the evolution of your research interests in this system, leading up to these whole-lake food web experiments? You refer to the “cascade hypothesis” from your 1985 BioScience paper. Was that where the seeds of these ideas were sown?
Stephen Carpenter: By 1980 or so it was evident that phosphorus could explain only about half the variability in biomass and production of lake phytoplankton. In order to explain the other half, something independent of phosphorus had to be involved. A number of scientists at the time had a feeling that the food web might be the missing factor. Jim Kitchell and I saw a way to test this idea experimentally using whole lakes. Jim is a fish biologist and I am a limnologist, so we were a natural team to take on the experiments.
HS: Stepping back a little bit, could you tell us how you got interested in lakes?
SC: As an undergraduate I became fascinated by ecosystem ecology, due to the mentorship of Stuart Fisher, a great ecosystem scientist. Stuart was a stream ecologist and prompted an interest in freshwaters as well. I was accepted to grad school at UW-Madison, one of the premier lake ecology programs in the world. Things evolved from there.
HS: This paper has nine authors. Could you tell us how this group came together and what each person brought to this paper?
SC: A big ecosystem experiment requires a lot of participants. Kitchell and I designed the project and co-led the team. I did most of the data analysis and wrote the paper. Hodgson and Cochran worked with Kitchell on the fish sampling and analysis. Jim and Monica Elser were the two technicians who handled the limnological sampling, sample analysis, and database. Incidentally, Jim Elser was my first undergraduate advisee at Notre Dame. After he and Monica obtained MS degrees, they returned to Notre Dame in their technician roles. Jim later got a PhD at U.C.-Davis and is now a well-known limnologist himself, famous for his work on stoichiometry of ecosystems. David Lodge was a postdoc with Jim Kitchell and helped with some of the field work. Don Kretchmer was a technician with Kitchell and also helped with general sample processing. Xi He was a grad student with Kitchell. Carl von Ende is an expert on Chaoborus and conducted the Chaoborus sampling and processing for the paper.
HS: Do you continue to work in “Paul, Peter, and Tuesday lakes” today? When was the last time you visited them? Could you give us a sense of how these lakes have changed since the time of this study, especially in terms of their food webs?
SC: I visited the lakes this summer, and may work on them again if a pending grant proposal is funded. Paul Lake is very similar to what it was in the 1980s. Peter Lake has been manipulated extensively for multiple whole-lake experiments. Currently it has a low population of piscivorous bass that will probably grow to dominate the food web in coming years.
HS: When you think back to the time when you did these experiments, what are your most striking memories? It must have been really hard work (moving 400-odd bass and 4500 minnows; 1128 bass stomach samples!)
SC: The field work was fun. My most striking memory was our surprise that the experiment worked. So many things had to come together for it to work, and they did. Later, we were surprised to see that zooplankton biomass did not change much. The food web had a big impact on the composition of zooplankton but not the total biomass. It was the shift in composition of zooplankton that impacted the phytoplankton.
HS: Would you remember how the figures for this paper were drawn and who drew them?
SC: I rough-drafted them, either by pencil and ruler on graph paper or by computer in the case of the primary production time series. Then Cheryl Hughes, the Zoology Department artist at UW-Madison, rendered them in Indian ink on parchment. This was the common practice for scientific illustration at that time.
HS: You acknowledge two people at the end of your paper. Could you tell us a little more about how you knew them and their contribution to this paper?
SC: Arthur Hasler conducted the world’s first whole-ecosystem experiment in Paul and Peter lakes in the 1950s. He was still living at the time of our experiment. Jim and I knew him quite well. Arthur was a close colleague of Jim’s. Jim replaced Arthur on my graduate committee when Arthur retired in roughly 1976. Cheryl Hughes, as noted above, was the artist.
HS: How long did the writing of this paper take? When and where did you do most of the writing?
SC: Once we agreed on the figures and tables the writing went pretty fast. I wrote the first draft in a day, as I try to do with all my scientific papers. Then I cleaned up the draft and got feedback from the other authors. After a few iterations it was ready to submit. I guess that most of the writing was done in winter 1986-1987, probably in my office at University of Notre Dame. I wrote it longhand on lined tablet paper in pencil, then edited the longhand first draft and typed it up on an old electric typewriter I had scrounged from the Biology department office. It might have been the last paper I wrote that way. I think I got my first desktop computer right around that time, and would have switched to word processing then.
HS: Did this paper have a relatively smooth ride through peer-review? Was Ecology the first place this was submitted to?
SC: The paper was rather revolutionary and some reviewers hated it. The first version was rejected from Science because a reviewer said it contradicted basic tenets of limnology and therefore had to be wrong. Then we wrote a longer more detailed version for Ecology. Gary Sprules was the editor, and I believe that he thought the paper would be really important someday. However he held us to a high standard and the reviews were tough. Whole-ecosystem experiments were unfamiliar back in those days. Ecologists wanted experiments to be done in tiny containers with lots of replicates, but it is impossible to study trophic cascades realistically in tiny jars. When I look at the paper today, I think the amount of statistical detail we included is rather ridiculous. But we had to include it to placate the referees.
HS: What kind of attention did this paper receive when it was published?
SC: Many top ecosystem scientists recognized that the paper was revolutionary, and were quite encouraging to us. I particularly remember that David Schindler and Gene Likens, two giants of the field, were supportive of the paper, as were a number of other scientists. However, much of limnology at the time was oriented around nutrients, and the idea of predator control was considered strange or even impossible by some people. We participated in debates about trophic cascades for about a decade after publication of the paper. I think our 2001 paper in Ecological Monographs answered a lot of the criticism. In any case, trophic cascades affecting ecosystem processes is commonly accepted nowadays.
HS: What kind of impact did this paper have on your career and the future course of your research?
SC: I spent the next 10-15 years on ecosystem experiments to work out details of trophic cascades in lakes and apply the idea to lake management. By the 2000s I was doing experiments on other issues such as carbon sources for food webs and factors affecting ecosystem stability. However, I continued to think about trophic cascades and occasionally wrote papers on that topic, even up until recent years.
HS: Today, 30 years after it was published, would you say that the main conclusion still holds true, more-or-less: “Food web effects and abiotic factors were equally potent regulators of primary production in these experiments. Some of the unexplained variance in primary productivity of the world’s lakes may be attributed to variability in fish populations and its effects on lower trophic levels.”
SC: Yes, that statement is accurate.
HS: If you were to redo this study today, would you do anything differently?
SC: Probably not. If today we were in the same situation with respect to understanding of lake ecosystems, then the same experiment would be important to do.
HS: You say “These studies demonstrate the importance of large-scale, long-term manipulation in the analysis of cascading trophic interactions”. Today, 30 years later, would you say you are satisfied with the extent to which trophic interaction research has adopted this approach (i.e. large-scale, long-term manipulations)?
SC: Yes, because for the past 20 years the major work on trophic cascades in lakes has been whole-lake management applications. For ecosystem science in general, whole-ecosystem experiments are one of the major tools for learning, along with long-term and comparative studies, and modelling.
HS: You say “Although fisheries biology and limnology have evolved as largely separate discipline, they must coalesce as we establish a holistic view of factors and interactions that regulate ecosystem functioning.” Would you say this has happened?
SC: No, they are still largely separate disciplines. However I think there is a general awareness now that lake water quality problems can be tackled through changing chemical inputs, changing the food web, or both.
HS: Have you ever read this paper after it was published? If yes, in what context?
SC: I just skimmed it for this interview, and I re-read it several times while we were writing our book ‘The Trophic Cascade in Lakes’. I’ve used it in class occasionally and would have at least skimmed it again prior to the classes.
HS: Would you count this paper as a favourite, among all the papers you have written?
SC: It is certainly the highest impact paper I ever wrote. I think it has stood the test of time very well. I never really thought about favourite papers but I certainly like this one.
HS: What would you say to a student who is about to read this paper today? What should he or she take away from this paper written 30 years ago? Would you add any caveats?
SC: Well, obviously it was early and rather compelling evidence for the role of trophic cascades, in part because the experiment was conducted in real lakes. There is a lesson there about appropriate scales of research. I think that the paper had a lot of impact because of the whole lake scale. It was controversial because the evidence was so compelling – it directly challenged a prevailing dogma that nutrients controlled everything about lakes, and some scientists felt threatened by this. Of course that dogma is gone now, and students might be surprised that it ever existed. If the students were interested in trophic cascades I would give them some more recent papers to expand their studies. I would tell the students to learn about modern notions of appropriate statistical evidence for big-scale experiments, and to apply those instead of the painful statistical overload that we were forced to present due to the customs of the time.