In a paper published in The American Naturalist in 2004, Jeannine Cavender-Bares, David Ackerly, David Baum and Fakhri Bazzaz provided an explanation for the maintenance of oak diversity in local communities in North Central Florida. Cavender-Bares and colleagues found that co-occurring oaks are more distantly related that expected by chance, a pattern resulting from convergence in traits underlying habitat specialization. Thirteen yearsa after the paper was published, I spoke to Jeannine Cavender-Bares about the origins of her interest in this topic, her memories of field work and what we have learnt since about the patterns and mechanisms underlying co-existence of oak species.
Citation: Cavender-Bares, J., Ackerly, D. D., Baum, D. A., & Bazzaz, F. A. (2004). Phylogenetic overdispersion in Floridian oak communities. The American Naturalist, 163(6), 823-843.
Date of interview: 21 December 2016 (via Skype; updated over email in 2021)
Hari Sridhar: What was your motivation to do this piece of work in relation to the rest of your PhD? How did you decide to work in this particular site in North Central Florida?
Jeannine Cavender-Bares: This was part of the main body of my PhD research. The original idea was to understand the adaptive significance of leaf lifespan. That was before the leaf economics spectrum work, which actually came out at the same time. When I started, there was a lot of discussion in the literature about why plants have different leaf lifespans, its functional significance and how it was related to strategies across resource gradients, in relation to biotic factors. I thought this really required some kind of evolutionary approach, and I wanted to compare congeneric pairs with different leaf lifespans. I needed species from the same genus that were evergreen or deciduous, or had contrasting leaf lifespans, and occurred together in the same place to factor out climate as a confounding issue. So, I needed to find a place where there would be many co-occurring, congeneric species with contrasting leaf lifespans. When I looked into it, it turned out that Northern Central Florida has the highest number of genera with species that have both long and short leaf lifespans. That was why I went to Florida. Originally, I had a whole series of genera that I was looking at. My first trip down, I met with an amateur botanist Angus Gholson. He showed me around the Apalachicola Bluffs and Ravines Preserve and other places in Apalachicola. Through various connections, I ended up in Gainesville, where Terry Henkel—who had botanized frequently with Angus and was a long-time acquaintance and former student of my Dad’s—was living. I ended up looking around in the North Central Florida area. I already knew about there being so many oaks from Walter Judd from the University of Florida, who had told me about them when I met him after a talk he gave at Harvard. Consequently, four pairs of oaks (eight total) were already on my list. But then when I was down there tromping around looking at all these species in their respective habitats, I was so struck by how many oaks there were. When I came back from that initial inquiry trip, just trying to scope out field sites, it struck me that it really might be more interesting to focus on the oak clade and look at the evolutionary history of how different lifespans evolved within that. And that was something that fit with what was going on at Harvard at the time. Michael Donoghue and David Baum were there. Stuart Davies was taking a clade approach looking at Macaranga. David Ackerly was working on Acer, and he would become one of my committee members, an important mentor and subsequently a good friend. It struck me that I could take a clade-based approach with the oaks in Florida, so, I turned my attention to them and stopped working on the other species.
The paper that came out in Am. Nat. was the culmination of many years of data collection and thinking. I first went down to Florida, I think, in about 1996. The paper came out eight years later. That was a long time to think through what I was seeing in Florida, and the questions changed, and I had to interpret the observations I was making and put them in some kind of context. That evolved over time.
HS: Stepping back a bit, how did you get interested in plant community ecology?
JCB: It’s a bit of a windy path. I grew up very interested in plants. I was kind of a plant science kid, growing up. My dad was a biologist; actually, a mycologist. We were outside a lot. My whole family is really into botany. So, I just had a strong connection to plants growing up. But then, I was also very interested in environmental policy. So, I spent a lot of time going in other directions and I almost did a PhD in environmental policy. In college, I studied biology, but also history, political science and environmental policy. During my Masters, I continued following these dual interests. I went to the Yale School of Forestry and Environmental Studies where I could do both science and work on applied issues. I finally had to decide, for my PhD, which direction to go in. I initially enrolled in a PhD at the Kennedy School at Harvard with Bill Clark. I deferred that for a year to do a Fulbright in Germany to work on a book Learning to Manage Global Environmental Risks (2001), Social Learning Group, edited by Clark et al. During that year, I spent a lot of time interviewing the people in Germany working on these issues – scientists, government leaders, members of industry focused on environmental risks, and members of NGOs involved in these global environmental issues. Every time I interviewed the scientists, I realized that I really wanted to be a scientist. I ended up not doing the PhD with Bill, although we have stayed in touch and intellectually engaged in various ways these many years. My new husband, at the time, had reorganized his life to start a PhD at MIT so that we could go to Boston together. Bill graciously gave me a job for a year at Harvard so I could figure out my new plans. Since I was already at Harvard and my husband was now at MIT, it was easier to stay at Harvard than go anywhere else. I reapplied to and got into the biology department there. There was really only one ecology professor there taking students at the time– Fakhri Bazzaz – and so if I wanted to study plant ecology at Harvard, my choice was Fakhri. I really liked his lab, and there was a lot going on—a very big dynamic group. Bill Clark assured me he was “a Scholar and a Gentleman.” Fakhri was nearing the end of his career, but it was a dynamic place to be. The department, at that time, was a very exciting place for organismal biology and evolution. Fakhri was an integrative plant ecologist. You couldn’t really put him into a box. Some people thought he was a physiological ecologist. He did community ecology, but he called himself a plant ecologist. And if you look at the people who have come out of his lab, they all tend to be very integrative. I wouldn’t call myself, necessarily, a community ecologist; I would call myself a plant ecologist. If you pressed me further, I would probably call myself a plant evolutionary and physiological ecologist. Community ecology was something that emerged from the project. It wasn’t what I set out to do, except that I knew I had to understand something about species distributions. The sampling that I did was using the same kind of approaches that community ecologists use. Christine Muth Petrusz was a lab mate at the time, who had worked with Bob Peet at the University of North Carolina. She encouraged me to use their protocols for my sampling. Those protocols have now become the standard for the national VegBank database.
HS: How did authors on this paper come together and what did each bring to this piece of work?
JCB: David Ackerly had also done his PhD with Fakhri Bazzaz. So, we’re, actually, academic siblings. But he finished his PhD a couple of years before I started. He was doing a postdoc with Michael Donoghue at Harvard at the time, at the Arnold Arboretum. We had a little bit of time to overlap and interact. David’s way of thinking about the world just clicked for me, and I wanted to learn the phylogenetic methods he was using. When he then moved to Stanford, I would fly out to the Bay Area and spend a couple weeks there. He is an extremely smart individual and sometimes it can be a challenge to follow what he’s saying because it comes at you really fast. But I always felt I could instantly connect to the way he was thinking about things. In that sense, intellectually, I feel like I’ve always connected with David, and I still collaborate with him. And then, because he became faculty, I was able to have him join my thesis committee.
David Baum was at Harvard at the time, and I did all the molecular work in his lab. He really encouraged me with the phylogeny. Paul Manos thought I would never get this to work. I called him up, originally, hoping that he would just send me the phylogeny. But he was pessimistic about the possibility, which is humorous in retrospect because we later worked together for many years and still do. But David Baum just said, yeah, go ahead and try it, to see where you get. So, that’s what I did. I was bold, I think, because David Baum encouraged me to be bold.
HS: Did the four authors ever meet together as a group during this piece of work?
JCB: Yes, because they were all on my thesis committee. I don’t know how frequently David Baum and David Ackerly would have seen each other, but at my thesis defense they were both sitting around the table together. I was individually interacting with these people. I was seeking out their help in different ways.
HS: In a footnote in the paper, you talk about how you’ve moved from using the terms “attraction” and “repulsion” to using “clustering” and “overdispersion” instead. Give us a sense of how you decided to use these terms?
JCB: I was reacting to the idea of “attraction” and “repulsion” being too explicit about process, when all we really have are patterns. Attraction is indicating that there’s some kind of process going on, whereby there is some mechanism pulling species together. It insinuates more about process than “overdispersion”. And the same thing with “repulsion”: it implicitly invokes a process, when all we have is the observation of the pattern. There is a mathematical issue with “dispersion” itself. I remember discussing, on a phone call with David, should we go with “overdispersion”? I remember we talked it through and came up with “clustering” and “overdispersion” as the word choices on the final version of the paper. The title may have been different in the original submission. Then later, David talked to Robert Colwell, who indicated that “dispersion” is an appropriate mathematical term but adding “over” and “under” are not quite right. For a while, David was publishing everything using the term “evenness”. I switched to that for a while, but the literature seems to have latched on to “overdispersion” and “underdispersion”, or “clustering”. I’m very comfortable with “clustering,” but “evenness” doesn’t quite have the same common language meaning as “overdispersion” does. It’s less clear what evenness actually means. It may be mathematically correct, but it’s harder to communicate what the pattern is with that word. In the end, “overdispersion” is sufficient and “clustering” is very clear. Anyway, “evenness” was not even on the table as an option at that time. That wasn’t something we thought of.
HS: “Evenness” can be confused with “species evenness”.
JCB: Yeah, that’s right.
HS: In earlier papers, you had used the terms “attraction” and “repulsion”. It was sometime between the 2003 paper and this paper that you felt that the terms attraction and repulsion weren’t suitable.
JCB: Yeah. It became clear to me how people were running with the idea of “repulsion” as meaning something like competitive exclusion. I have never been convinced that it has anything to do with competitive exclusion. People started running with that idea, and I was getting papers to review making assumptions to this effect, and so that’s why I really backed off “repulsion.”
HS: What do you remember about the actual doing of this work? What was your daily routine, did you have people to help with different aspects, etc?
JCB: I spent a year in the field just learning about the ecological system in Florida. I literally spent a year just tromping around all over Northern Central Florida trying to figure out where to set up my sites. I put in plots and contacted the professor at the University of Florida who was in charge of the undergrad biology students. After about a year, he gave me some names of absolutely fabulous undergraduates who had time to help – Jimmy Sadle, Jason Teisinger, and Rachel Seman. Those were three undergrads who helped me out in the field in Florida—all absolutely fantastic and all pursuing science-related careers. And then there was Radika Bhaskar, who had actually worked in the lab of Penny Chisholm, my husband’s PhD advisor at MIT. Penny’s group contacted me and said, you know, here’s this great student, can she work in your lab? Fakhri was pretty easygoing about most stuff and so Radika came to work with me, and helped with the molecular work in David Baum’s lab; it was all very fluid. Radika later wanted to do a PhD, so I connected her with David Ackerly. She went on to do her PhD with David Ackerly at Stanford. She then received a fellowship to conduct research in Mexico with Patty Balvanera and then did another postdoc with Erika Edwards. I believe now she has a faculty position at Philadelphia University. I’ve kept in touch to some degree with all of these awesome people. They were undergrads at the time, and now they have full-fledged careers. My cousin Eric Weis was doing a Master’s degree at the University of Rhode Island at the time and was running into some snafu. So, I asked him if he wanted to help me with my greenhouse experiment—he helped me with harvesting the plants and measuring the plants in the greenhouse. There were a number of undergrads who were fabulous and super helpful; some became good friends and are still wonderful people in my life.
HS: Can we go over the other names in the Acknowledgements to get a sense of how you knew these people and how they helped? The first name is K.Kitajima.
JCB: Kaoru Kitajima. She was a new assistant professor at the University of Florida at the time. The first year I was down there, I was just living in Terry Henkel’s house (with his wife and son) after he started his PhD at Duke. These days, Terry is a well-known mycology professor at Humboldt University in California. But at the time, he was kind of a hippie living in Florida, doing botanical collections for the Smithsonian. I stayed in Terry’s house for a year; it was super cheap and for a year I was just on my own; my brother-in-law gave me his old van and I sort of had this lab on wheels. We called the van “Big Blue”. It later died in the field right after I finished everything. But at the start, it was the van and me at the hippie house. And then people at the University of Florida started to find out about me because I was connecting with the undergrads there. Also, Missy Holbrook was one of my really important mentors and committee members at Harvard and she had a lot of contacts down in Florida. So people knew I was there. Kaoru was a physiological ecologist. She invited me to use her lab rather than just running around with my van—my lab on wheels—and so I set up shop in her lab and collaborated with her on a related paper. There’s another paper that came out earlier; some of that work had been done in her lab. It was important to bring her into a collaborative framework because I was spending time in her lab and she also let me stay in her house when she would be off working in the tropics in Panama. She ended up being my local host there and became a good friend. Now she’s at University of Kyoto. I recently went out and visited her, and she’s doing amazing things out there.
Paul Manos also became a good friend and wonderful collaborator, but at the time, I had just cold called him and asked him if he could give me the phylogenetic information for those species. Because that’s what I wanted; I just wanted somebody to give me the phylogeny. But he was like, no, that’s never going to happen. There’s no way you’re ever going to resolve the phylogenetic relationships for the oaks because of the introgression problem. And he had been working in North Carolina and finding introgression among even the live oaks there and couldn’t pull them apart from the white oaks and was very frustrated and cynical; which is really funny and we laugh about it now. We have since collaborated, together with Andrew Hipp, on the genome sequencing of all the oaks. As I mentioned before, David Baum encouraged me to just go ahead and try it. I mean, you could never publish a phylogeny like this now, but I tried it, and the relationships looked pretty darn good and not that different from what the actual genome-wide phylogeny looks like now. I haven’t gone back and compared, but I know that the major groups came out pretty well. I stayed in contact with Paul Manos the whole time and he approved of the final phylogeny at some point. I must have shared it with him, and I ended up giving him the remaining DNA from that project; some of it later went into our collaborative work.
Then you have Kevin Nixon, a world authority on the oaks. Kevin Nixon was Paul’s major advisor and is an interesting character. He’s at Cornell, where Paul did his PhD. Kevin Nixon and Michael Donoghue saw the world differently, but that is another story. Michael Donoghue was involved in all this because he was at Harvard at the time and a very influential figure there. Early on in my dissertation, Michael tried to get me interested in Viburnum instead of Quercus, using a similar clade-based approach. I was intimidated by Kevin Nixon because I knew about his work—Viburnum would have been an alternative. By now, I was spending a lot of time in Florida, and I became somewhat involved in academic life at the University of Florida. I was a bit of a carpetbagger coming down to Florida quite frequently and spending time at the University of Florida. Kevin Nixon came down to give a talk. He was talking about angiosperm fossils and evolutionary history; nothing to do with oaks; I went to his talk .I can’t remember the details now, but somehow I ended up taking him and David Dilcher and perhaps others on a field trip to San Felasco Hammock State Preserve, one of my field sites. Kevin ended up testing me intensely on all the oak species. The oaks elude many people down there, because there are so many of them and a lot of them look alike. But it turns out a lot of that likeness is actually convergent evolution, which is fascinating. But you really do have to spend a lot of time to get good at identifying them. I had been doing it right from the beginning, so I had gotten pretty good at it. I marked key individuals throughout my study site that I was doing physiological work on, and Kevin came around and checked my identification and gave me the big thumbs-up for the project. I learned a lot just going out in the field with him. He may have looked at the phylogeny at some point later before I submitted; I can’t remember. He did check my phylogenetic work in subsequent papers when I actually was studying live oaks (Quercus section Virentes). I later started working on the Virentes, the very group of oaks that he did his whole thesis on, and I was quite worried that he was not going to be very happy about it. At one point, I was invited to give an alumni talk at Cornell University because I had been an undergrad there, and I was hoping maybe Kevin wouldn’t notice that I was there talking about live oaks. It turns out that he was getting honored for a teaching award in the same public seminar I was asked to give the alumni award talk for. And so it was that we were in the same seminar room at the same time—and it actually ended up being wonderful. He asked a whole suite of fascinating questions. He understood the whole thing much better than anybody else, and we ended up having an amazing conversation afterward. It was super fun. My fears of Kevin melted away.
Then there is Cam Webb. He was at Harvard doing a postdoc, at the same time I was there doing a PhD There would be these discussions, I think led by Michael Donoghue. There was quite a hubbub at that time of exciting, evolutionary comparative methods. Peter Stevens was there at the time. It was just becoming possible to put together molecular phylogenies and start thinking about evolution of organisms and their ecology in new ways. Cam was pushing this idea of looking at the relatedness of species in a community ecology context. And, of course, that work was pivotal, but at the time it wasn’t published yet. There were just these conversations going on. I knew about his thinking, and I started explaining the patterns I was seeing in the oaks, and that lit something in his mind. And actually, I’m nearly positive that that 2002 paper that Cam did, that David was also on, wouldn’t have come out the way it did if they hadn’t known about my work. I don’t think that they would be aware that you could get convergent evolution, that you could get environmental filtering on functionally convergent traits—a pattern that can lead to overdispersion. That wasn’t part of their thinking before they knew about the oak stuff. And so I like to think that I actually had an influence on the way that framework came out. If you read that paper carefully it does actually acknowledge the oak work in there. In any case, it’s hard to say; in anything academic there’s a lot of flow back and forth, ideas are being passed around and people influence each other. Cam was definitely part of that mix of people who were influencing my thinking at that time. This paper took an awful long time to really nail down and finish. Originally, I wanted to try his mean phylogenetic distance calculations, because I saw the pattern and I wanted to test it statistically. So he sent me his code, I ended up coding everything in Visual Basic (not in C) but I used the algorithm that he had. But at that time, his initial algorithm for the null models was problematic. The only way to get it to work initially was to have the rare species put in first. For some reason it was just much more computationally efficient to put the rare species in first. And when I ran that in Visual Basic, I got hugely significant results. I was super excited, like, oh man, I should send this to Nature or something. Then I started doing more reading and I may have had a conversation with Nicholas Gotelli, but I started realizing that the algorithm was biased and that I had to change it with the independent swap algorithm. Okay. At that time R was not widely used by ecologists, and there were not many libraries, and so forth. So, I coded the independent swap algorithm myself. It took a long time; I’d just had a baby and everything. When I re-ran the analyses with the independent swap algorithm, it was no longer significant. Those metrics were calculated on individual plots, and what I was doing was taking all the measures on the individual plots and then aggregating them to get a value for the whole landscape. And it was no longer significant the way I was doing it when I used the unbiased algorithm. You can imagine how this set me back and I was like, okay, now what? There goes my Nature paper. I’ve actually never cared too much about trying to try for those big profile papers. I’m really, uncomfortable publishing something that I’m not super confident in. But here I was seeing this pattern, and it was very clear to me that there was all this convergence going on. And then every time I’d find these gradients and I’d find all these species from all the different lineages in those, I knew there was a pattern in there. I just didn’t have the right statistical method for showing it. And so, I guess I just invented—actually reinvented from scratch—the Mantel test; I didn’t realize that’s what I was doing. But I wanted something that would fit the whole system at once, not one plot at a time. So, I ended up just taking the phylogenetic distances between the species and comparing them to their co-occurrence patterns. But I had to come up with a new way to look at co-occurrence, so I based it on Schoener’s Similarity Index. I also wanted to look at their ecological similarity–their niche similarity. That work is published in Ecological Monographs—but I guess it’s in this paper too. Anyway, I looked at their ecological similarity in terms of their abiotic environment and their co-occurrence patterns, and then the overdispersion pattern that’s published in the Am. Nat. paper came out, and it was significant. (Years later we found the same pattern of overdispersion in the oaks for thousands of US Forest Service plots across the entire US). It took me a long time to redo the statistics after I realized that Cam’s algorithm was not going to work. And by then I was dealing with a little baby and I was doing a postdoc; I felt it was taking forever to get this thing published. Jonathan Losos had just published a paper in Nature on Anolis lizards, and felt it was time to get the paper out. Am. Nat.’s a great journal. I think David Baum was trying to get me to send it to Science, but I didn’t have any experience with short format papers, and it was more important to me at that time to be thorough and get it out. That’s what I did. I remember you had a question in here about the review process. Maybe it’s a good time to talk about that now.
HS: Yes, go ahead.
JCB: Jonathan Losos was the subject editor at Am. Nat. He was great. He had really important comments that helped improve the paper. He was the one that actually suggested that I put all the traits in the same format as the Webb et al.2002 framework, and put them on those two axes. That’s been picked up by other people now. That suggestion came from Jonathan Losos. It was really fun to see how the data fell out in doing it. I think it pretty much sailed through the review process.
HS: There are a couple of other names on the list of acknowledgments. A. Wilczek.
JCB: Amity Wilczek. She’s now Academic Dean and Professor of Biology at Deep Springs College. It’s a very unique college. She was my lab mate at the time—she was also a student of Fakhri Bazzaz—and we actually shared a house. I don’t recall her input exactly, but she would have been somebody who read the manuscript perhaps, or that helped me think about concepts. And then there is Kent Cavender-Bares, my husband. He had experience with scaffolding and helped me build the towers that we used to access the canopies of some of the trees. For most of the species I used a bucket truck—which I’d already been using at the Harvard Forest for the first couple chapters in my thesis. But some of the species are only found in ravines, and they’re never going to be in the places where you can drive a bucket truck with an aerial lift. So the only way to get to the top of those canopies to measure their functional traits was to build towers. My husband came down and—along with Jimmy, Jason and Rachel—we put up these towers. Three towers allowed us to get 15 more trees that we wouldn’t have been able to get otherwise.
HS: What were these towers built of?
JCB: They were made of scaffolding. They were very safe and secure. My husband’s really good at construction, and he knew how to assemble these. One of the towers was donated by Henry Gholz, a much loved ecologist originally at the University of Florida, who left for NSF and later tragically died in a climbing accident. Assembling those was some of the hardest work any of us had ever done—Jimmy, Jason and Rachel would attest to that. It was crazy how much work it took to get those towers up for the few measurements we got. But it was super fun to be up in them and people always wanted to visit them. So, for example, when Kevin Nixon came down, the first thing we did was to take him up the Gholz tower, which you could actually walk up because it had stairs. The other two you had to climb up with rope.
HS: Were Jimmy, Jason and Rachel the undergrads you mentioned earlier?
JCB: Yes. There are three papers that were just as much work and just as important as this one. This one’s just has been picked up because of the fascination with the phylogenetic part of it, but the other two papers are equally important in terms of the intellectual investment, what we learned and the quality of the work. Missy Holbrook was a collaborator on the paper that came out in Plant, Cell and Environment, and Kaoru Kitajima is a collaborator on the Ecological Monographs paper. I harnessed multiple people’s brains for this whole effort and they were collaborators on different parts of it.
HS: You mention an NCEAS workshop on phylogeny and community assembly. Do you remember when that happened?
JCB: I don’t remember exactly, but it might have been around 2003 or 2004. I would say that, by that time, all this work was done. That was a meeting that Cam Webb convened and he’d invited all the people that were doing stuff related to phylogenetics and community assembly. It was a really fun meeting. I was told I should go because I was young, it was early on in my career and NCEAS was a special place. I was told it was kind of a prestigious thing to do, and I did. The conversations there were super fun and it led to the special issue that Cam led in Ecology that came out in 2006. I have a paper in there with a couple of people from my lab looking at the importance of scale. That was something that everybody was grappling with: why do we see overdispersion sometimes and clustering other times, There seems to be a scale dimension to this, and what’s up with that? It was pretty obvious to me that the oaks would be clustered if you looked at them at a larger scale—a larger phylogenetic scale, a larger spatial scale—what would that actually look like? I would say that this meeting was important for that subsequent step (Cavender-Bares et al. 2006).
HS: Do you continue to work today in the field sites you used in that paper?
JCB: Well, I kept going for a while. In fact, for the 2006 paper, we went back and surveyed the same site for all the species. We did put permanent rebar in the ground, but there’s a lot of controlled burning there. If anybody wanted to go back, we’ve got all the coordinates, and it would be possible to go back and resample. But they’re not being regularly sampled.
HS: When was the last time you went to any of these sites?
JCB: I actually have a common garden down there. But I’m not monitoring any fields. I have oak species planted in a common garden just north of the San Felasco Hammock State Preserve.
HS: Since these sites are in a Preserve, I’m guessing that they wouldn’t have changed much from the time you sampled them?
JCB: Well, there was a fair amount of prescribed and also natural burning. And there are also a lot of invasive species. The community composition could well have changed.
HS: What is a common garden?
JCB: It is a common environment in which a range of species, populations or genotypes can be planted to factor out the environment as contributing to the phenotype. The phenotypic variation, or the functional variation–among species or among genotypes or among populations—can then be attributed to genetic differences among organisms rather than to environmental variation because individuals are randomly planted across a fairly homogeneous environment. A common garden allows us to essentially factor out the environment as a contributor to the phenotype.
HS: Of the three sites you sampled, why did you pick San Felasco Hammock State Preserve to measure functional traits?
JCB: I picked the site after going to many, many different places. San Felasco is just a fantastic Preserve. It was one of the biggest in the region and it’s also easy to get to. It has old growth forest in it and a whole range of different habitats. That preserve, alone, has 14 or 15 of the total oak species in it and was the obvious place to get most of the species. It has so much heterogeneity in habitat and all three of the main community types there. That was the best place to get most of the species in close proximity. Because I had to access most of the tree canopies with an aerial lift, having them all in one location made fieldwork simpler.
HS: You’ve already said a little about why you chose to do the soil analysis at the University of Florida. Anything else you want to add to that?
JCB: The University of Florida had excellent facilities for analyzing soil. It’s a public land-grant institution with a large agricultural science component. I did all of the soil analyses in the UF soil testing laboratory. They taught me how to do the analyses. Tom Sinclair, who was in agronomy (and a colleague of Missy Holbrook) helped me with the bureaucracy. That simplified everything. It would have been would have been very difficult to set up all of those analytical procedures at Harvard. At the University of Florida, it was all set up; I just had to learn the protocols and run them on my samples.
HS: Does that lab still exist?
HS: Does the greenhouse in Harvard where you did some of the experiments also still exist?
JCB: The greenhouse has an interesting history. They built the facility for Fakhri when they recruited him to come to Harvard. The greenhouse was designed by Fakhri to do to look at elevated CO2 in a replicated fashion. It had six rooms of which three had elevated CO2 and three had ambient CO2. I am not sure of its current status.
HS: Was all the molecular work also done in Harvard?
JCB: Yes, in David Baum’s lab.
HS: Does that lab still exist?
JCB: David got tenure at Harvard, which is not so easy to do, but he and his wife both subsequently went to the University of Wisconsin-Madison.
HS: Would you know what Baum’s lab space is used for now?
JCB: It is possible that Chuck Davis is in that lab now, but I’m not sure.
HS: Do you remember roughly how long it took you to write this paper and when and where you did most of the writing?
JCB: Well, it was a long process. In my thesis, this paper is one of three related chapters. The first paper looked at hydraulic traits along environmental gradients. It was in collaboration with Missy Holbrook and came out in 2001. I wrote the Ecological Monographs paper and this paper simultaneously from the rest of the work. I made the Am. Nat. paper specifically phylogenetic while the Ecological Monographs paper focuses more on niche breadth and examines the relationships of the functional traits to the ecological gradients more thoroughly. I remember in the review process, I did have to explain how these two papers were different. There was so much work embedded in both of them, and at one point one of the reviewers for the Ecological Monographs paper said I should just add the phylogeny. I responded that there was actually a whole separate paper on that, and I had to send the Am. Nat. paper to the editor—Martin Lechowicz was the editor for the Ecological Monographs paper—and he agreed that it was too much to put into a single paper; it was already a monograph. And there was a baby in the middle of this. I got the paper with Missy submitted, and that must have been in the last year of my PhD. I moved with my husband to Washington DC and had a baby that year, in middle of trying to write all of these papers, while also starting a postdoc at the Smithsonian. A baby changes everything. So there was moving, baby, life changes, finding a faculty position, and various other things going on. I spent a long time doing analyses. Once I had figured out what the story was, the writing wasn’t that hard because I’d been thinking about it for years by that time. It was an intense experience because I felt like I had a lot to say. I was engaging with the larger community and going to meetings and interacting with people. Any paper isn’t just something done in isolation. There are a lot of different influences. It took me a long time—it certainly felt like a very long time—to get these papers out. I didn’t actually get them submitted until I started my faculty position at the University of Minnesota.
HS: Did the other authors get involved in the writing?
JCB: I would say Fakhri was very little involved. He was a cheerleader for me. He was a strong supporter. He was not particularly involved. I would say there was a lot of intellectual discussion with David Ackerly and David Baum, which was really important. David Baum helped me with the methods and the phylogeny, made sure that it was done correctly it, for the time; at the time, it was the best we could do. And with David Ackerly it was just many years of intellectual interaction and he would have edited paper.
HS: At the time when the paper got published, do you remember how it was received? Did it attract a lot of attention?
JCB: Well, actually, when it finally got published, I felt like there wasn’t any reaction. It seemed that in Minnesota nobody understood what I was thinking about or trying to do. I actually changed what I was doing. I’m interested in quite a lot of things. I became really interested in trying to understand micro-evolutionary processes and trying to figure out adaptive variation within species across environmental gradients. And so that’s what I set out to work on next. I then became pretty frustrated because I kept getting these papers to review that would show patterns of clustering and overdispersion and then draw grandiose conclusions based on these patterns. I felt people really hadn’t done the thinking and the work to figure out the mechanisms that produced these patterns and were not doing a good job of linking patterns to processes. It seemed that people were putting phylogenies on everything, perhaps as a reaction to the Webb et al. 2002 paper. I don’t know. But that’s why I later wrote the review in Ecology Letters —I felt that somebody had to clarify why you can’t just link a pattern to a process, that multiple processes could generate the same pattern. The Ecology Letters paper was a response to what was going on in the field at that time. But, originally, there wasn’t much response. I felt like, particularly in my own department, there was not much appreciation for the work. I was really surprised when, a few years later, UC Davis tried to recruit me to their evolution and ecology department. I almost went. When they talked about this work as pioneering, I was blown away. I was really, really touched that people appreciated it.
HS: When was this?
JCB: I think that would have been around 2008. I was being appreciated in circles that just didn’t happen to be the immediate circles I was working in at that time. I got a little depressed, actually, because there were all these questions like, was that really your work if your advisors are on there? I always work collaboratively. I depend on being able to talk to people and have a sounding board for my thoughts. I don’t know what to say about that further but there was a little bit of sadness afterwards. It looks different in retrospect.
HS: Later on, it seems to have attracted attention. It’s been cited over 500 (in 2021, almost 800) times. Were you surprised by that later attention?
JCB: It did attract attention before it was published. That was the interesting thing. Within the circle of colleagues thinking about community ecology and phylogenetics, it was really well received. I actually won an award for it as a graduate student when I presented at the Ecological Society of America. I went on the job market and I got six interviews, all fabulous places, and three job offers. I gave talks that encompassed this work, and I’m fairly certain that’s what got me the job at the University of Minnesota, which is a great place. So, it was appreciated before it was published, and people did recognize the importance of the work and the novelty of it. It was only after it was published that I didn’t perceive much of a reaction, but maybe I wasn’t really paying attention.
HS: Do you have a sense of what it mostly gets cited for?
JCB: I think it was actually the first paper that used functional attributes of species together with the community phylogenetic pattern. Jonathan Losos had his lizard paper, but I think it was one of the first studies, or the first study, to show that you can get convergence in traits. I think people had a hard time really understanding that. For years afterwards, I felt like I had to explain that you really can get environmental filtering on convergent traits and end up with a pattern of overdispersion. Now, I really am convinced that it has much more to do with the pattern of parallel adaptive radiation and less to do with competitive exclusion. There may well be biotic reasons why distantly-related species can coexist better than closely-related species, for example, due to contrasting disease resistance and pathogen/herbivore resistance. Those factors are likely to stabilize the pattern.
HS: Anything more you want to add to how this paper impacted your career?
JCB: I was asked to give a lot of talks. And I did. I talked widely at many institutions and universities around the US and abroad about this work.
HS: After this study, how much longer did you continue to work on oaks in Northern Florida?
JCB: It was not possible to maintain a research program down there and in Minnesota. I subsequently started to work in Central America for a long time, as well as in Mexico with Antonio Gonzalez-Rodríguez. I still have common gardens in Florida, but I’m no longer working in the plots that I set up. Steve Pacala once told me that what you do for your thesis lasts your whole career. You just keep basically reinventing your thesis. That’s not quite what I’ve done—I’d like to think I’ve done a few other things—but it was the bedrock for the rest of my career because I learned so much during that period. It was the foundation for what I know now about plant ecology, and it is a reference point whenever I come into contact a new system or read a new paper because I know that system so well. When else do you have that much time to throw yourself into something so fully? It was before I had children, it was before I had faculty responsibilities. I was 100% in. I worked on that project very, very intensively for years, and just tried to investigate everything I could about it, from every angle. I’ve never had the opportunity to approach anything with such thoroughness since, although, collectively, the teams I am working with do all kinds of awesome things. I simply don’t have the time to devote my entire psyche to a project like that now.
HS: It’s now 12 (in 2021, 17) years since you published this work. Would you say that the main conclusions from this work still hold, true more or less?
JCB: I actually do. I’ve gone back and reread the Am. Nat. paper, particularly the discussion and felt like, yeah, that’s pretty much how I still feel about this. There was a lot of thinking that went into those papers at the time, and it was good thinking. I’m more aware now of the parallel adaptive radiation in the larger evolutionary and biogeographic history of the oaks based on the collective work with Andrew Hipp, Paul Manos and others, but I gained insights about those broader questions from the work in Florida.
HS: If you were to read redo this today, would you do anything differently?
JCB: I have to laugh a bit at this question. There’s no way in hell I could ever do that study again .It was a journey. I’m thrilled that I was able to do it as thoroughly as I was at the time. Would I have done something differently? You know, I’ve actually never thought about it that way. I’ve never thought, gosh, if I’d only done this, if I’d only done that. That’s never a thought I’ve had about the project. What I have thought is, I can’t believe I pulled that off and learned so much from that system. It was a once in a lifetime opportunity. It was an incredible amount of work and a lot of things came together to make it possible. I just can’t imagine that all those things could happen again in the same way. So, no! Maybe I would have published it sooner. I would have liked to have gotten it out more quickly, but I’m always kind of slow.
HS: From what you’ve said am I right in understanding that the current phylogeny of oaks is not very different from the one presented in this paper?
JCB: The major groups are well supported by the genome-wide phylogeny with much more sampling and much greater numbers of individuals sampled across their ranges and with thorough genome-wide DNA sampling (see the work led by Andrew Hipp in New Phytologist). Even some of the species-level relationships have been maintained, but not all of them. The aspects of the phylogeny that led to the patterns of trait convergence and phylogenetic overdispersion in communities—those have been upheld
HS: In the discussion of your paper, you say, “Host specificity of pathogens or herbivores within lineages could result in density-dependent interactions that would limit the co-occurrence of closely related species and promote the co-occurrence of distantly related species.” And then you say, “Empirical observation, combined with modeling approaches and experimental manipulations, could be used to test such hypotheses.”To what extent has this happened?
JCB: People have done this kind of thing in other systems. Actually, I still want to do this, and working with a group of oak people now, we may get to do a little bit of this. These are hard experiments to set up and oaks are not easy to manipulate experimentally in a manner that explains long term patterns; their longevity makes it difficult to witness density-dependent interactions. It is easier to do this with slime molds or bacteria or other fast-growing systems. But I think we have a plan in place to try this. People like Greg Gilbert have done some nice work related to this. Cam Webb and this guy at UC Santa Cruz published a paper in 2006 and 2007 that were really important papers related to this. And so there’s certainly been some nice work that’s done. I have yet to do this, but a group of us you’re trying to do something that’s very related to this.
HS: You also say, “Modeling approaches could be used to test the hypothesis that long-term competition between lineages could lead to evolutionary conservatism of traits that promote coexistence.” To what extent has this happened?
JCB: Nathan Kraft and David Ackerly did a whole simulation study and show that the Webb et al. framework is upheld. They also show that when you actually have competitive interaction and you simulate those, you’ll get the expected pattern if the traits are conserved, or if they’re convergent you get this other pattern. That simulation study was pretty important. I wanted to actually do some modeling with introgression to see if introgression is critical to the patterns you see. It may be that introgression operates like competitive exclusion does. You can get gene swapping introgression that ends up looking, from a pattern perspective, like competitive exclusion. And maybe the introgression works in this system the way competitive exclusion does in some other kinds of systems; and that is critical to the pattern. But I could be completely wrong on that. So it would be fun to try a modeling approach that at least looked to see if it was possible to get those patterns from introgression.
HS: Have you ever read this paper after it was published? If you have, in what context?
JCB: I have read it again. A few times, maybe. I don’t remember the context. Maybe, just to see what I said at that time. I’ve actually checked to see how I handled the issue of competitive exclusion versus other hypotheses to explain the pattern. I’ve tried to understand why people interpret that paper so much in terms of competitive exclusion. I do mention it as a possibility, but I really feel that people have run with the pattern as competitive exclusion but that’s not my interpretation. I offered it as a possibility. But I think I’m more thoughtful about it in that paper. And so I’ve gone back to check how I handled it, and I think I did handle in a fairly thoughtful way. And I’m not sure people have read that as much as maybe I would have liked. Again, that was why the 2009 paper seemed important.
HS: When you read it before this interview, what struck you about it? Did it bring back any memories? Did you think your writing today is different from the way you wrote then?
JCB: I don’t really know what to say. I’ve always been a pretty dense writer. If you look at my recent stuff, it’s similarly dense. I guess I can’t remember now. Do I have memories? When I go back and I read the methods, I just think about how much work that all was, and how I can never do that again. And how I’m glad I got to do it then. I’m really grateful that I was able to have that much freedom in my PhD. I think a lot of people, for various kinds of reasons and constraints, they nest themselves inside an ongoing project, or they build our hypotheses from their advisor, and so forth. I was given complete freedom to do whatever I wanted. In the first two years, I struggled with that a lot. I didn’t know what to do. And it took me three years to come up with the idea, and then it took all those additional years to actually do it, and then interpret and figure it all out. That was a lot of freedom to develop myself as a scientist. And I feel that that comes out of when I reread that paper. I can see that process of developing myself as a scientist.
HS: Would you count this as one of your favorite papers?
JCB: Yes, except that I honestly don’t see it as separate from the other two. The Ecological Monographs and the Plant, Cell & Environment papers were part of the same body of work; the same study. Collectively they represent an integrated understanding of the evolution, physiology and ecology of a lineage that explains how species assembled in communities. They set the stage for later work understanding the biogeographic history and assembly of the oaks across the continent.
HS: What would you say to a student who is about to read this paper today? Would you guide his or her writing in some way? Would you point them to other papers that they should read along with this? Would you add any sort of caveats they should keep in mind when reading?
JCB: I suppose I would read this together with the 2001 PCE paper and the Ecological Monographs paper and the 2006 Ecology paper, because they represent a body of thinking about this system, and then more recent papers on the evolutionary and biogeographic history of the oaks (Hipp et al. 2018; Cavender-Bares et al. 2018; Cavender-Bares 2019 NP; Hipp, Manos and Cavender-Bares 2020 Scientific American). It is fascinating to see that the phenomenon of phylogenetic overdispersion observed in Florida provided the clues to understanding how the oaks colonized the continent and became the dominant lineage in North America through sympatric parallel adaptive radiation.The oaks represent a model clade for integrating ecology and evolution and tell the story of how long term processes give rise to observable patterns. The caveat is that they represent a single case study, albeit an important one. Other lineages have other stories to tell, and very different processes can explain similar-looking patterns.
For a graduate student, perhaps a lesson from all this is the extent to which you can forge ahead with your ideas and take some risks. There were risks doing this study, going down to a place I had never been, setting things up, without being sure what I was going to find, my questions and hypotheses changing along the way, giving rise to an evolution of thinking. Many times during this study I was afraid I would fail. Students should have the courage to forge ahead with their ideas and really do their own thinking, but not in isolation. Thinking in a rich context of reading the literature and talking to people that you admire intellectually, and not being afraid to reach out to people whose work inspires you and interacting with them, is so important. To the extent that students develop ideas in an intellectually stimulating environment and have the space to test some things that don’t work out, that sets them up for a career as a scientist. I look around at my colleagues and they’ve all done that. They’ve all taken risks; they’ve all failed at things. What you don’t see in these papers are the pieces that that didn’t work out.