In a paper published in The American Naturalist in 1972, Joan Roughgarden presented a model to predict the course of evolution in a population containing individuals of different types of ecological specialization and put it to test using field data on a species of Anolis lizard collected by Tom Schoener. Thirty-five years after the paper was published I spoke to Joan Roughgarden about how she got interested in the topic, her memories of carrying out this study, and what we have learnt since about the ‘evolution of niche width’.
Citation: Roughgarden, J. (1972). Evolution of niche width. The American Naturalist, 106(952), 683-718.
Date of interview: 27th December 2017 (via Skype)
Hari Sridhar: What was your motivation to do the work presented in this paper?
Joan Roughgarden: Well, the unifying theme to my PhD thesis was the idea of density-dependent selection. And I was motivated by the aspiration to bring ecology and population genetics together. And I remember reading the original paper by Robert MacArthur on r-selection and k-selection, in which he stipulated that one could use the carrying capacity of a species as a measure of fitness. And of course, from a natural selection point of view, you can’t use a population parameter as a property of individual fitness. And so, I thought that was a problem up front. But then, in the meantime, I’d been working through the mathematics of how to derive the basic equations in population genetics, and it was perfectly obvious that you could write an expression for individual fitness that was still a function of population size. So an individual’s fitness could depend on population size, and then you could parameterize the dependency of fitness on population size, in terms of the carrying capacity, and, in particular, the functional form of the dependence is a linear function of density, if you’re going to work in the framework of a logistic equation. So that allowed me to incorporate the parameter the carrying capacity into individual fitness. That then led to the first chapter in my thesis, on density-dependent selection, which was also the first paper that I published. That then became the theme. Now, if I recall correctly, there were two other chapters. I mean, I haven’t looked at my thesis for decades. But one of them was on the evolution of symbiosis; I was very interested in that. That didn’t lead to so much at the time, because I developed that as a cost-benefit model for whether a parasite should transition into being a mutualist. And I embedded that in a population genetic framework but there wasn’t a lot of interest at the time in symbiosis. This is in the 70s, and all the interest by community ecologists was in competition. So, now, of course, there’s much more interest in symbiosis and cooperation, but at that time, it was a real slog. And then the third paper did look at competition but looked at it from an intraspecific point of view. And the idea is whether you could have niche differentiation within a species as well as between species. And so, thinking back to it – I haven’t had a chance to actually check the paper; I’m not even sure I have the paper anymore – the concept that I remember introducing was that of a competition function, because I remember looking at the notion of a convolution. There was a mathematical expression called the convolution, so I was able to use Laplace transforms, which seemed very exotic at the time. And I used it from my training in physics, such as it was. And so the competition function is interesting and encapsulates the notion that the strength of competition can depend on the distance between two phenotypes in niche space. And that then leads to all kinds of conceptualizations about how a community can be structured in terms of distances in niche space. And that’s also consistent with a lot of the thinking at the time, especially from Martin Cody. Cody was a student of MacArthur. He had an influence on me because he would make these maps, like a two-dimensional grid, and then he would locate on the grid the niches of different bird species. And then he would do this for different parts of the world. And it was extremely interesting, and I wish that they were still being done, because you get a picture of the whole community of niches, the way Martin Cody would present it. Parenthetically, I remember visiting Tasmania a couple of years ago. And there there’s a whole mammal community, and I only wish Martin had been with me to lay out how all the different kinds of marsupials were spread out across the niche space in Tasmania. Anyway, so the competition function was a way through which you could potentially have two dimensions in that space. At one point we even worked with a student who worked on a model that was a two dimensional space. And so the competition depended on something like a Euclidean distance in the niche space.
At that time, Richard Levins and, I think, MacArthur together were interested in the idea of developing the competition function from an underlying specification of how the parties overlapped in their resource use. There was interest at the time – tell me if I’m going on too long – in building up to the Lotka-Volterra models from an underlying description of the mechanisms of competition. Because it was taken as a criticism at the time, that the Lotka-Volterra competition equations just sort of stood by themselves, and weren’t relatable to the nuts and bolts species interactions. And so the idea was to look at the actual resource use overlaps, and then develop a sub-model whose output would be the competition function or the competition coefficients. And the sub-model would be specified in terms of particular mechanisms per species and per location, but it would still wind up with a competition function. And I don’t think that wound up being too promising, but there are a lot of papers in that genre at that time. It turns out that the overlap on the different resources wasn’t always comparable because some resources would be more renewable than others. And so the idea sounded nice, but it got complicated really fast. And again, parenthetically, now nobody seems to care about that issue too much. Because in my most recent paper, which is on the evolution of holobionts, things like corals, with a host and a microfauna, I encountered a lot of literature, especially coming out of the microbiology community, where they’re sort of reinventing or doing community ecology to a large extent, largely unaware of the early work by Vandermeer and others, and Gause even, but they’re fitting Lokta-Volterra equations to microbial dynamics just by using regression coefficients, so that the coefficients in the Lotka-Volterra equations emerge as regression coefficients, if you regress the abundances against the changes in the right way. And that’s what we were doing in the 70s! And that was considered phenomenological in the 70s and not mechanistic. But people are quite content with that now, for looking at microbial communities, in large part because there’s so many microbes, and getting to the mechanism of the microbe-microbe interaction is nearly impossible. So, people are content with a phenomenological representation of that. So I encountered all that in this most recent work. But getting back to the 70s, one interesting mathematical consequence of looking at these competition coefficients, has been generated by overlap, by the product of utilization curves, is that you generate a competition function, or if you’re looking at a discrete set of species in a competition matrix, which is positive definite, which means that it has positive eigenvalues to it. And so you wind up getting a very nice structure, mathematically, out of communities that are assembled with competition function. Now, of course, if there’s predator-prey interactions and all sorts of other interactions, that’s a problem, but for something like a guild, which you might consider as being solely a collection of competing species, it could have nice stable properties, in principle. But as you know, that work in the 70s, which was all focused on whether an equilibrium point for community would be positive or whether it would be stable to perturbation and so on, that became somewhat eclipsed by the work that Bob May started involving nonlinear cycles and chaos and so forth. When I was doing the work in competition theory, what I was doing was very mainstream in the theory at that time. But when the mainstream became more focused on nonlinear dynamics, then I couldn’t really track that, I couldn’t go along with that too well, because the species that I was working on were not susceptible to that. It was mostly the insect people who were really turned on by the nonlinear dynamic stuff. You know, the cicadas and all the big cycles that you have and for oak moths and things. But I was working on lizards, and, you know, did for many years, and they did not show oscillations in abundance. We had excellent census data on my study sites for a decade or more; I forget. It’s published in my little book on Anolis lizards. And so we really knew that they did not fluctuate in population size. There was a seasonal cycle, but if you equated it from the same time of the year to another year, you know, it’s the same. And they were in equilibrium in some practical sense of the word. And then I was also working on barnacles – had this project with intertidal barnacles – and there we definitely had fluctuations in abundance. But there was no doubt that that was caused by fluctuations in the currents and in the upwelling and in the currents returning larvae to the natal habitat. So you could plant your flag or plant your buoy, so to speak, a half a mile to a mile offshore, study where the currents were going, and be able to predict, really closely, what the abundance was going to be that subsequent year. And so the explanation for the fluctuations in abundance in barnacles was directly tied to physical events and not to internal nonlinear dynamics. And so the whole focus on non-linear dynamics, to me, from the experience with my field systems, had no relevance. And so I didn’t go in that direction, and, instead, in the 80s-90s started working more on economic type models and other things. So that’s a long answer to your question.
HS: Was the PhD almost entirely theoretical, apart from the analysis of lizard data in this paper?
JR: Yes, it was. I began my fieldwork after my PhD because I didn’t think it was possible to get a job solely as a theoretician. And so I approached Ernest Williams, who was the curator of amphibians and reptiles at Harvard, and also the principal advisor for Tom Schoener and Bob Trivers. And so we had an active group of students at that time, and I joined it really in large part to be in communication with these other people. Ray Huey of Seattle was also part of that group. So then it was because of being with Williams that I then went into the field with him. Now, he was very much a taxonomist, since he was a collector. So our job, when accompanying him, was, literally, just to catch lizards, and then he would preserve them. And then along the way, I was able to sort of piggyback some studies as part of my responsibility to catch lizards. And so I measured a lot of jaw sizes and things of that sort, in order to get an estimate of niche distances, as reflected, to the extent that they’re reflected in jaw length and body size differences; which is pretty close. I collected a bunch of lizards, dissected them and looked at the relationship between the actual size of the prey eaten and the size of the jaw or of the body length; jaw and the body length are almost interchangeable for correlational purposes. So, it was pretty obvious that small lizards were eating small prey, like ants, and big lizards were eating caterpillars and flies. So that meant that they were on different resources. Now, of course, they were different places in the vegetation too. So you did have a, like, a two-axis niche space.
HS: Your affiliation in this paper is the University of Massachusetts. Did you go here immediately after your PhD?
JR: Well, it was actually during it. This was a difficult time in America, like it is again now. At that time, the Vietnam War was raging. And so, after the end of my first year at Harvard, I started teaching at a small college called Simmons College in order to get an occupational deferment from the draft. That was a small undergraduate college with a lot of politics and really unpleasant place to stay. So then I began my third year in Boston at the University of Massachusetts as an instructor, and I was still working on my PhD. I finished my PhD at the end of that year. So I did my PhD in three years. In the second one I was teaching at Simmons College. In the third one, I was teaching as an instructor at University of Massachusetts at the Boston campus, which no longer exists. This was in downtown Boston. And then when I got the PhD at the end of that year, they made me an assistant professor. So my fourth year in Boston was as an assistant professor at University of Massachusetts. And then I moved to Stanford.
HS: Stepping back a bit, which came first the interest in ecology or the interest in mathematics and theory?
JR: Well,the actual genesis of it is that when I was in high school my mother wanted me to become a doctor. And so I began as an undergraduate thinking I was a pre-med, but then I looked at the other people who were pre-med and I didn’t want them as colleagues.They were competitive and nasty and I didn’t want to live my life with these people. And so, at that time, I was taking a course in biology, an honors course,and there was a microbiologist named Wolf Vishniac – it’s amazing how you remember all these names from ages ago, and you forget the name of somebody you met yesterday – and I wound up doing very well there. He encouraged me to go on in biology, which was the first time anyone had ever done that. And I thought I was going to do electronics and design equipment. My father was an engineer, so I was kind of being pressured – if I wasn’t gonna be a doctor, at least I had to be an engineer and design the things that hospitals use – big pieces of equipment and a lot of electronics. I knew some electronics as a hobby. I used to be a ham radio operator, so I knew how to do electronics. So I thought I was going to do that. So then I wound up just taking the bio major, and then I took a course in ecology from someone named Conrad Istock, and it was kind of interesting because I could see that, when it came to anything theoretical, he was struggling. He would try to teach it, but he would botch it; mess it up. Oh, he was a charming man; don’t misunderstand me. I still remember him trying to explain the fundamental theory of natural selection in an ecology course, which is kinda risky, but he did it. But he didn’t know. So I did, I could figure out what’s going on. I said, well, you know, maybe I can make a living here. Maybe I could teach mathematics to biologists. And so I felt that there was a niche for me there. And I also looked at neurobiology because it had a theoretical component, but it never really resonated with me. I like the outdoors more; I’m not a lab person. And so, so then I simply went on. I just start reading privately, on my own, about ecology and ecological theory and population genetic theory, especially. At that time, I didn’t even really make a distinction between ecological theory and population genetic theory. It was all just theory of evolution and ecology jointly. It was only when I got into graduate school, and then especially after graduate school, it became clear to me that, you know, the geneticists think of themselves as really special people, off on the side, and ecologists think of themselves as different. There are different organizations and different societies and different journals. But when I was an undergraduate, I wasn’t aware. It’s often the same way. And I hadn’t realized how cheeky it was, you might say, to envision a unification between ecology and population genetics. I hadn’t realized that that was going to step on anybody’s toes; seemed to be more like a technical problem than a scientific sociology problem.But yeah, that aspiration continues to exist today. I keep coming across people who were calling for unification with genetics with ecology. But there’s a lot of hostility to that at the time. Lewontin was definitely not in favor of it, but he seemed to have an epiphany about 20 years later and he decided that it would be a good idea.The geneticists had kind of a holier than thou attitude, they viewed the ecology as sloppy outdoors; not controlled. And the ecologists were equally snobby and though that geneticists did nothing but work with little bottles of flies; nothing to do with the real world. And so there was definite hostility between both camps. Because of my pedigree, I guess, having come from Harvard and having evolutionary training, I was considered an ecologist by my colleagues, and basically was, but even to this day, I still do a lot of evolutionary and genetical theory. But also at the time, at the time I got my PhD, I knew I needed to do field work to have a career. You see, Bob May was just coming into the field. He had met MacArthur and he was a physicist. And Simon Levin was working; he was obviously a mathematician. They were both older than I was. And so, I knew that I would be unusual in doing theory and field work. At that time, there was no one else active in theory, who was also doing field work, and who had a biology degree, rather than a math or physics degree. Though I had math and physics training, my actual degree and experience was biological. Now, of course, there are a lot of theoreticians who do field work and theory. This is not unusual anymore.
HS: How did you get interested in lizards, and how did you decide to use a lizard dataset in this paper?
JR: That was all because I joined Ernest Williams’s group, and they were focused on lizards,and so when in Rome do as the Romans do. And the idea of using the size of the lizard as an indicator of the prey came from MacArthur, because he was the one who, and before him Hutchinson, all of whom were looking at morphological indicators of ecological position or activity.
HS: Who was your supervisor during your PhD?
JR: William Bossert. Monty Slatkin and I were both students at the same time of Bossert’s. Bossert was in the computer science department, he wasn’t a biologist, but he did collaborate. There was also – actually, you might know him – Madhav Gadgil. He was a little older. He was, I guess, a postdoc at the time, but he also, I believe, was a student of Bill Bossert; so the three of us were together, and Bossert was my advisor. And so, therefore, it was possible to do a theoretical thesis under his supervision, but he still had privileges in the biology department, and so he could be an advisor to someone in the biology department.
HS: If you don’t mind my asking, how come Bossert was not an author on this paper, which formed part of your PhD. Was it common then for PhD students to publish single-author papers?
JR: Well, it was pretty common, but I don’t think Ernest Williams had anything to do with 1972 paper. But it was a subsequent paper – I think it’s 73 or so – where I had a lot of data on lizards. He should have been a co -author on that because I actually took those data while I was traveling on his grant with him- as I mentioned earlier – catching lizards.S o, he should have been a co-author on that. But also, it wasn’t as common. Ernest Williams is not a co-author on Tom Schoener’s papers, or on Bob Trivers’s papers, or Ross Keister, who was in the lab at the time. So, occasionally there would be one, but it wasn’t so common at all. In fact, if you go through the journals at the time you will find a lot of single-author papers.
HS: Let me see if I’ve understood this correctly – Ernest Williams was also an advisor on your PhD?
HS: Can we go over the list of people you acknowledge, to get a sense of how you knew them and how they helped?
JR: Yes. Let’s see how this goes.
HS: You’ve already spoken about Bossert. The next name is J. Cohen.
JR: Oh, Joel Cohen. Yeah. Well, I think he would have been someone I discussed it with.
HS: Was he also in the same department at that time?
JR: No. He was in biology, I believe. But then he also went into public health somehow. He is, is a demographer, basically. He’s now at Rockefeller institution –been at Rockefeller for sometime – he’s mostly known for putting together a database of food webs.
HS: The next name is Madhav Gadgil, who you have already spoken about. Ernst Mayr.
JR: Oh, yeah, Ernst Mayr. Well, yeah, I mean, I talked about it with him. He was the grand father figure; the great man. And he had a graduate seminar in evolutionary biology in which Slatkin and Trivers and I were all members, and so we would talk; it was a good time. I mean, Ernst Mayr would pontificate, and he would hold forth and you would get the view. You knew what the Modern Synthesis was when you heard it from the horse’s mouth.
HS: And then, Tom Schoener, who shared his data, Otto Solbrig, Ernest Williams and EO Wilson. Were all these people at Harvard?
JR: Yeah, in the biology department and the Museum of Comparative Zoology.
HS: And then you thank another set of people for comments on the final draft of the manuscript, which includes Martin Cody, Bruce Menge…
JR: Yeah, he’s an intertidal ecologist at Oregon. His wife is Jane Lubchenco. He and I were at the University of Massachusetts together, in that second and third year, he and I were teaching the same course together – team taught. He was a student of Bob Paine‘s.
HS: Michael Soule.
JR: Well, he was a student of Paul Ehrlich out at Stanford and he worked on lizards. He was generally pretty critical of theory. But I solicited his feedback and took it into account.
HS: Robert Trivers and Van Valen
JR: Oh, Leigh Van Valen. He was at the University of Chicago. He’s known for the Red Queen hypothesis. I don’t remember interacting with him a lot. But he was a theoretically interested person at the time.
Have to say, you know, looking back, it was a really exciting time in the 70s. We really felt like we were revising or revolutionizing the field. Larry Slobodkin used to say, ecology is nothing but the misidentification of beetles on a field trip. And it really was. In fact, my original advisor at Harvard was George Clark, and then I switched to Bossert. There was no conceptualization at all. What MacArthur really brought was an appetite for conceptualization to ecology. Before MacArthur, it was just collecting samples and recording the number of amphipods in a stream or the number of beetles on tree bark; you know, there was nothing to it, conceptually. It was just the description of the contents of various biotas. And so MacArthur brought not only conceptualization but an explanatory goal. So the goal of ecology became to explain what you saw rather than just document what was there. And so that led to a lot of interesting, sometimes rather abstract, conceptualizations of what is the community. It didn’t come out of the blue. You did have Clements and Gleason, you had these old fights at the turn of the century between them. Gleason claiming that there really wasn’t anything called the community; just a bunch of things happened to be blown into the same place. And you had Clements, I believe, arguing the community was an integrated entity – higher order entity – like a super organism. And these guys used to fight it out. And then you had David Lack, who was sort of from that generation too, and fairly conceptual.But MacArthur really changed things. Not only did he want to explain ecology, but he wanted to do it conceptually and with predictive models. So, he paved the way for theory in ecology. Meanwhile, at the same time, in evolution, there was the GC Williams, sort of thing, focusing on the individual, and there was the early Richard Dawkins and Hamilton focusing on the individual as the unit of selection, and not the species. This is before there were any group selection models to explicate that. At that time, in evolution, you know, there’d be the reinterpretation and the reanalysis of phenomenon, in terms of individualistic evolutionary thinking. I remember one time attending a conference in Georgia and there were some of the ecosystem people there – Dac Crossley and one of the Odums – Howard Odum, I guess. We’re talking about how ecosystems put together, and we’ve talked about the role the role of herbivores like caterpillars and butterflies, and this guy gets up and he says, Now, the way I look at consumers is that they’re like the governor on a locomotive, and without the governor a locomotive runs off the track. The governor is like this little gadget that which forbids the engine from just running away. And so he was saying that the herbivores are like the governor on the locomotive, and they serve the function of pruning the plants so that they can grow better. So they’re like the natural gardeners. And so it’s functional, but it’s in service of a higher goal or the object of the fitness of a higher entity. And so Rob Colwell was there at that time, I remember that, and we looked at each other and said, my god, they hadn’t even heard about natural selection and how natural selection works on the individual and gene pools change and you don’t get the evolution for the good of the species or for the good of the ecosystem, in their case. So there was all this reexamination of things at that time. Now, my regret, though, is that I don’t feel that the people who were active in the 70s have grown. This is a long time ago. And now, particularly in England, I think, and, to a lesser extent in the US, evolutionary biology has become very defensive, and you can’t criticize Darwin,and you can’t criticize evolutionary theory and you have people trying to write definitive statements about what evolutionary theory means; you actually get a lot of that from the British. And, as you probably know, I’ve criticized sexual selection theory a lot, which I think is really seriously mistaken. But, in my view, the problems of sexual selection theory represent a great opportunity. It’s not as though the critique is incorrect. I mean, the critique is there; these are real problems. The need to solve them opens up a whole horizon of activity. Instead, the stance of most evolutionary biologists to any criticism is to attack to attack personally. There’s like a fear that if we show any weakness then it will be exploited by the creationists. And so there’s the circle the wagons mentality that, I think, is holding back evolutionary biology. It was really king of the mountains, so to speak, in the 70s, but now it just seems like a beleaguered discipline. At least here in the US, half the people don’t believe in it, which means we haven’t done a good job of reaching out. And, in contrast, ecology has really prospered since the 1970s, especially because of its connection with conservation. So, it’s almost been a flip in the relative stature of evolution and ecology over the last 50 years.
HS: Do you remember how long it took you to write this paper, and when and where you did most of the writing?
JR: Ah, not really. Obviously, it was written longhand. That was in the days when if you made a correction, you used white-out, and you used carbon paper.
HS: I noticed that you say that the computer work was supported by a grant from the Research Computing Center of the university. Forty-five years ago, how difficult was it to do these simulations?
JR: Oh yes, it was. This was before the era of PCs. Bill Bossert, in particular, was key in getting the Computer Science Center to have a large computer – physically large – so it occurred in a big room. You had to interact with it with a teletype. And also, you could bring a deck of cards. So there was a card reader, and you used to have to type out your program by punching out cards. I can’t remember the details of what languages I was writing in then. Early on, I was writing in Fortran. I know some of the simulations are in Fortran, but that’s a card oriented language , and the text starts at column seven, and if as an instruction goes over one card, you put a little x in the first column of the next card to indicate a continuation of the command. So then, you’d have to have job control language at the top, because you’d have to call a compiler,the compiler would then compile it, I guess, and call an assembler or something, and then it would spit out a printout. Now, somewhere in there, there were teletypes that communicated with it. I don’t remember the details. Somehow, you could get the printout sent to your teletype. You didn’t have to get the printout sent to your line printer; I think, you could direct it to your teletype. And I think instead of using cards, you could use one of those paper tapes with little holes in it, and you can input it through the tape reader next to the teletype. Toward the end of the period that I was a grad student, Basic came along. Basic was really a lot more useful than Fortran. And so it’s easy to code a graphmaking capability onto the teletype. So, by typing asterisks and lines and so forth, you could generate a sort of graph-looking thing on a teletype. But it was really crude. And so the reason you had a grant is, of course, the University bought the computer, but in order to get time on it, you’d have to apply internally, which they call the grant. So somebody from the outside who wanted to use it would have to pay at a certain rate,somebody from the inside of university would have to get a grant to pay at the same rate, even though it’s kind of funny money, because they knew when they had a computer, if they’re going to let students use it, they were going to have to give the students paper money or funny money, in order to have time to use it.
HS: Was The American Naturalist an obvious choice for this paper?
JR: Yeah. That was the most conceptual of the journals. MacArthur published there. A lot of the early papers in the 70s were there.
HS: Did this paper have a relatively smooth ride through peer-review?
JR: I believe that one went through pretty smoothly. The one I had trouble with was the first one – the density-dependent selection one. That was in Ecology. It was being edited out of Chicago, and it was given to Brian Charlesworth, who was then a postdoc at Chicago, to review. And then there was something that, I thought was quite unethical that took place, because Charlesworth turns out to have been independently working on the same thing – density-dependent selection. And so he held the paper up and then wrote a sequel paper, which had the aura – it was very snobby – had the aura of mathematising what I was offering, primarily, in terms of computer simulations. I thought that that was unethical and showed that to Bossert, who kind of shrugged his shoulders and said, Yeah, but just keep going. And so the paper came out.
HS: At the time when the paper came out, do you remember how it was received? Did it attract a lot of attention?
JR: Yeah, it was well-received. I think many of the field people felt that it was like a secondary issue. Especially from the community ecology perspective,the focus was on the species as a whole, you know, like, if there are three doves, what is the size of the different doves? How does competition structure species? And then it seemed like a refinement to look within a species to find out what the niche width was. I think that people’s reaction was positive, but they thought it was a secondary issue to the primary issue of the inter-species interactions to begin with.
HS: I realize that your interests have shifted now, but do you still keep track of the literature on this topic?
JR: No. And I think that’s important that, you know, you do what you can and then you move on.You can’t be possessive about work that’s released to the public; that’s released to the peer group. And I don’t think people should spend the rest of their career defending what they did when they were younger. I think you have to let go and move on to other issues -and god knows there’s so many other issues that need research. I know of some of the people my age haven’t let go. They’re still in there plugging away what they’ve been plugging away at for 50 years. And I think that’s wrong, I think that inhibits younger people and inhibits progress in the field. And also, just personally, I would find that kind of boring and stressful; it’s stressful to be on the defense all the time. Whereas, if you get into a new subject, well then you’re taking the lead.
HS: Did you make a conscious decision to stop working on this and work on other things? Or was it that something that just happened, you know, along the way, you started working on other things, and this, slowly, tapered off?
JR: It was a conscious decision. They do overlap, but I did take the conscious decision to start working on the marine system, which was quite different. The criticism that occurred,and may even still occur, for Anolis, is that it’s a specialized system. It’s a model system and not representative of nature more generally. And so I decided to work on barnacles, in part, because nobody could say it was a model system. They’re all over the world. I mean, if even nothing else was like barnacles, the fact that barnacles are one of the biggest species in the world means that that’s worth knowing about just intrinsically; whereas, the interest in Anolis lizards was always derivative upon its implications for general ecological thinking. But considered in their own rights, no one really cares about Anolis; it’s not that important in the grand scheme of biology. Though, on the islands, they eat a lot of insects. In some of our experiments, we did show that if you excluded lizards you did get a rebound in the number of insects. So the cascading kind of effect is definitely real;so lizards do play a role. But the problem is that the Caribbean itself is not a big part of the world. And so, even if they’re a big deal on all the islands of the Caribbean, they’re not a big deal worldwide.
HS: Towards the end of the discussion, you say, “I conclude with an appeal for more research on the basic genetics of continuously varying characters. although the need for models which breakaway from the now classical single locus two-allele viewpoint has long been recognized. the paucity of basic pre-theoretical information on the genetics of continuously varying substances discourages such efforts.” In the rest of this paragraph you are, basically, making the case for more research on the genetics of continuously varying characters. To what extent has that happened?
JR: I haven’t thought about it in a while. I suppose so. I mean, that was written at a time in which the work would have been more classical. So, now, with modern molecular methods, one, I suppose, would decompose variation in quantitative characters into contributing loci spread across the genome. Now, I’m not sure how well that’s worked out. Mostly, I’ve looked into that kind of genetics in connection with genetics of gender and sexuality and so forth, which I have to write on periodically. And there, it seems pretty doubtful. For example, I did a review article last year, or maybe it was earlier this year, that came out in Francisco Ayala’s book on human nature. So I did a critique on evolutionary theories of homosexuality. And I reviewed the literature on the genetic evidence for that. And it’s all over the map. It’s really very inconclusive, and there’s a significant amount of information that’s directly contrary to any idea that there’s a genetic component to homosexuality. If there is something, it’s very little that have the quantitative character, I suppose. And then you find these different labs come up with different indications of genes. For example, one lab will say it’s chromosome 17 that has a hotspot in a certain location, then another lab will say, no, we didn’t find that, but we did find something on chromosome five. And so if you step back, and in a meta-analysis kind of perspective, if you’d expect to find there’s no effect at all of genes on that quantitative character then, then various studies would claim to find different things. And if you step back from it, that’s exactly what you’d expect if there was no effect – different studies would argue that there was a different effect. And so, so then you say, well, homosexuality in humans looks like it is a polymorphism of some sort, but you can’t really find any genetic basis for that, then what is the story? So I think that you’re kind of left with a dissatisfaction with the success of the efforts to do quantitative genetics today. Now, I don’t know if that can be generalized to other characters, like if you’re breeding corn or tomatoes or cattle. I don’t know if the underlying genetics of those kinds of quantitative characters are any clearer. Looks like breeding programs can just go ahead and proceed with ad-hoc methods without knowing anything about the underlying genetics.So, the short answer your question is that I’m not sure that the situation has improved even with modern methods.
HS: Soon after this paper was published, do you know if the ideas in the paper were tested empirically?
JR: No, but every now and then I come up with a paper. They do seem to test out okay, so far as I know. It’s been a long time since I was in that.
HS: What kind of an impact do you think the paper had on your career?
JR: Well, it was definitely helpful because, you know, I had to get tenure, five six years or so after this. But my papers on co-evolution were more current at the time I was up for tenure. If I recall correctly, that was about when my first book came out –Theory of Population Genetics and Evolutionary Ecology. That was also quite influential. So it was quite a productive time. So, I had a pretty good file when I came up for tenure.I began at Stanford in1972, I think. So I would have had the manuscript for the book, certainly, five years after that.
HS: Did the ecology textbook with Paul Ehrlich come out later?
JR: Yeah, that’s quite a bit later.
HS: Have you ever read this paper after it was published?
JR: No. I remember being at a meeting of the Evolution Society, and there was a whole symposium on Niche Theory. I remember thinking, gee, there is symposium on that, and remarking to myself, I used to be active in that area.
HS: Which year was this?
JR: Oh, five-seven years ago. It wasn’t that long ago.
HS: Did you attend any of the talks?
JR: No, there was a conflict. I was presenting in another one.
HS: Would you count this paper as one of your favorites among all the papers you have published?
JR: Ah,well, yeah, I mean this and my very first one on density-dependent selection .I think the first barnacle paper is an important one, the one on demography of an open marine population. And then there’s a Science article about the importance of recruitment. And then I think one of the lizard-oriented papers, the notion of an invasion structured versus a co-evolution structured community, is important. Then I also think the paper I have in PNAS on how to harvest a population, why to set the optimal stock at three quarters of the carrying capacity and how that’s buying a natural insurance, is important. Then, coming more recently, some of the ones with Priya Iyer, especially the first one about the origin of anisogamy. Because I think that really sets the whole discussion of evolution of sexual reproduction off in a different direction than the party line in England, from Parker. And then, Erol and I did a lot of things and I think one of the best we did was on the evolution of payoff matrices, where you embed the payoff matrix in a genetical model and then show how natural selection selects the payoff matrix, thus the whole notion of the two-tier evolution of social behavior. Then this most recent one, which just came out months ago and has been two years in the making, on the evolution of holobionts. I think I may actually get back to research. As you may know, traveling a lot now, doing a lot of photography, selling photographs locally. But, getting back to holobionts, I might, in the second half of 2019, I may get back to that, again, because I just scratched the surface on it. The whole idea of a holobiont is really interesting, in the sense that the genome of a holobiont is the union of the genomes of all the microbes together with the host. But the microbes come and go, so there are ecological processes that bring the microbes together and make them join the host So you’re having selection on the level of the holobiont, and this has led to remarkable integration. So one of the collaborators on the work is Scott Gilbert, who’s this developmental biologist,and it’s just astonishing how much information they’ve pulled together about how deeply, physiologically, integrated the microorganisms are with the host; just really, really integrated. And it’s more than just a transactional ‘I’ll help you if you help me” kind of thing. It’s much, much more intriguing. And the whole notion of the hologenome, therefore, is you have a dynamic genome. So you have selection on the holobiont and holobionts have dynamic genomes because the constituent genes come from the colonization of the host by different microorganisms. In evolution, we’ve always assumed you know that the genome is fixed. Think of it as a string, of course, in the simplest case, with little spots on it for loci. But in the case of the holobiont the genome expands and contracts,and the whole genome is very fluid. So, almost getting back to the 1970s, it merges ecology and evolution. It’s ecological processes that underlie the colonization of the host by the microbes. And meanwhile, there’s selection and genes and all that sort of stuff going on.So that’s a really intriguing problem and I may get back to that in a year. [Added by interviewee to trascript in 2020: I did return to this subject, cf.http://dx.doi.org/10.3998/ptpbio.16039257.0012.002 and https://www.biorxiv.org/content/10.1101/2020.04.10.036350v2.] But I’m going to Africa in a couple of months for two months, going up the coast for all the way from Cape Town up to Lisbon, and then I’m going to Antarctica next January. So my travel itinerary is pretty booked.
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? Would you add any caveats they should keep in mind when reading this paper?
JR: Well, what they should take away is the conceptualization of the niches, the notion of the competition function, the notion of the carrying capacity as a function of the niche position, and the competition function as a function of difference in niche positions between the two competitors. I mean, that’s the underlying conceptualization in the paper. That’s what they should take away.
The significance of the paper is that it shows why an ecological community does not consist of one species filling all the available niches but instead consists of multiple species, assuming the species are sexually reproducing, as are birds for example. An asexual species old enough to accumulate enough favorable mutations can spread out over all the niches in the environment, but the random mating in a sexual species continually diminishes the frequency of the outlier phenotypes and reinforces the frequency of the common phenotypes. A sexual species phenotype distribution cannot mirror the distribution of available niches, but is instead distorted by the mating system’s influence on the phenotype distribution. This leaves some niches relatively vacant, providing opportunity for more species to invade, leading to a buildup in the habitat of species diversity involving relatively specialized species rather than leaving a habitat filled with a single highly polymorphic species.