In a paper published in Science in 1998, Jonathan Losos, Todd Jackman, Allan Larson, Kevin de Queiroz, and Lourdes Rodrı́guez-Schettino, using morphometric and phylogenetic analyses, showed that the same set of habitat specialists or “ecomorphs” of Anolis lizards had evolved independently on four Greater Antillean islands, suggesting that adaptive radiation in similar environments can produce similar evolutionary outcomes. Nineteen years after the paper was published, I spoke to Jonathan Losos about his motivation to do this study, how this group of authors came together and what we have learnt since about adaptive radiation in Anolis lizards.
Hari Sridhar: What was your motivation to do this particular piece of work?
Jonathan Losos: This paper is about the fact that on the large islands in the Caribbean, what we call the Greater Antilles, the Anolis lizards are very similar. On each island, there is a diversity of different species, morphologically, that occupy different habitats, and behave differently, and the interesting thing is that you have the same set of habitat specialists on each island; with a few exceptions. We call these habitat specialists “ecomorphs”. The existence of these ecomorphs had been first pointed out by Ernest Williams, the curator of herpetology at the Museum of Comparative Zoology at Harvard, and the Grand Old Man of Anolis. He’s the one who put Anolis on the map of evolutionary biology. He and some of his students, most notably, someone named Stan Rand, had pointed out this phenomenon about the ecomorphs, and had said that they had evolved independently. They based that assessment on an understanding of phylogeny based on morphological characters. It was based on the sort of analyses that were common in the pre-cladistic era. By today’s standards, we wouldn’t think much of these analyses, but Williams was actually ahead of his time in thinking in these ways. That was the landscape when I started working on anoles. I think that the ecomorph convergence phenomenon was not particularly well known. Part of it was that people weren’t convinced by Williams’s analyses. It’s not that anyone objected to the idea; they just didn’t pay a lot of attention to it. Now, I’m not entirely sure why that is. In other words, the idea was certainly out there, and as someone who had grown up academically knowing about anoles, I certainly knew, and suspected that this was correct, but it hadn’t gotten a lot of wider attention. I should point out, at this point, that I was an undergraduate working in Williams’s lab at Harvard, working with this last graduate student, Greg Mayer. And so, I was steeped in knowledge of anoles from an early age.
I should say one other thing. Research on Anolis, since the early 1970s, had been really at the forefront of evolutionary ecology, thanks to Williams and the people he trained. Every time a new method of phylogenetic inference was developed, it was immediately adopted by people working on anoles I’m referring particularly to the sorts of molecular methods for inferring phylogeny, starting with looking at karyotypes, looking at protein evolution in ways that people no longer have even heard about – something called micro-complement fixation – electrophoresis and so on. People working on anoles had jumped on every single approach. And so, when DNA sequencing was first developed, it was obvious to look at the anoles and try to figure out their phylogeny, and to test the ecomorph idea. And so, that’s what I set out to do. Initially, that was going to be my postdoctoral project. And I actually got a postdoctoral fellowship from the National Science Foundation to do that. But for various reasons, I took a different postdoc and put that project off. But finally, when I became a faculty member at Washington University in St. Louis, and started my long-term collaboration with the molecular systematistat Washington University, Allan Larson, – also a herpetologist – we decided to try to do this. That’s the long introduction to where this paper came from. It was a desire to test Williams’s ideas about the evolution of these ecomorphs – were they actually convergent on different islands– using what was, at that time, state-of-the-art phylogenetic systematic methods.
HS: Stepping back a bit, how did you get interested in lizards and in anoles in particular? Was this something that you knew you were going to study even in your undergraduate days?
JL: Well, certainly like many herpetologists, and like many biologists in general, I got my start with dinosaurs. I was fascinated by dinosaurs from a young age. I was legendary in nursery school for my basket of plastic dinosaurs; I would bring that to nursery school every day. As I grew older, my interest changed from extinct reptiles to living ones. A very key event in my development was when I was in sixth grade, about age 12. I convinced my mother to get a baby Caiman, which is a type of alligator, as a pet. They were fascinating, if nasty, little animals and I became more and more fascinated with reptiles. I did science projects in eighth and 12th grade, both using the common North American Anolis lizard, Anolis carolinensis, not because I had any particular interest in anoles, but simply because they were very common in pet stores, and so easy to obtain. And so, I had a background on anoles, and when I went to college, I said that I wanted to try studying biology, and herpetology in particular, and see if I really became interested. And, it just so happened that the herpetologist at Harvard when I was an undergraduate was Ernest Williams. And so, I quickly learned a lot about Anolis. That’s how I got to this project.
HS: When was the first time you visited these islands?
JL: That was in the summer of 1982, after my second year in college, my sophomore year. I went along as a field assistant with Williams’s graduate student, Greg Mayer, to Jamaica for three weeks to collect Anolis lizards.
HS: How did this group of authors come together and what did each bring to this piece of work?
JL: As I said, Larson was a molecular systematist and a faculty member at Washington University, where I joined the faculty. He’d been there for about six years at the time. Larson and I knew each other from Berkeley, where we had both been together at the Museum of Vertebrate Zoology; Allan was a postdoc when I was a graduate student there. Kevin de Queiroz is a morphological systematist at the Smithsonian, who works on lizards and had an interest in anoles. He and I were graduate students together at Berkeley as well, so we knew each other for a long time. Although this paper didn’t include morphological systematics, he knew a lot about systematics in general, and our goal was eventually to incorporate morphology into these studies. Todd Jackman was the postdoc who worked with Larson and me at Washington University. We hired him after we got a National Science Foundation grant. Jackman, too, was a Berkeley product. He also had been a graduate student at the Museum of Vertebrate Zoology. He had started a few years after I began, so we overlapped some at Berkeley. So, we knew each other and I was delighted to hire him when we got the chance, because he was an expert on molecular systematics. Lourdes Rodrı́guez-Schettino, very sadly, died three years ago [as of 2020]. She was a Cuban herpetologist, probably, the world’s authority on Cuban lizards. She was our collaborator for the research we did in Cuba, which was essential for this project. It was not the focus of our Cuban work, but it was essential for us getting tissue samples from Cuban species, which were necessary for this project.
HS: In a note in the paper you say you measured 10 to 15 individuals of 46 species I’m trying to understand the sequence of events.Was this data that you had already collected in the course of other work,or did you go out and collect it, specifically, for this study?
JL: You know, I need to revise what I just told you about Lourdes Rodrı́guez-Schettino, because this answers your question, in part. The morphological measurements on the Cuban species were actually critical for this study. Those were collected in collaboration with Lourdes. So, the collaboration with her was, actually, an instrumental part of this project. I say that because, in answer to your question, the data presented in this paper had been collected over a period of 10 or 12 years; it was really an extension of my PhD work. My PhD had focused on the anoles of Puerto Rico and Jamaica, the two smallest of the Greater Antillean Islands, right. At the end of my PhD, I went to the Dominican Republic and collected data. Those data didn’t actually make it into my PhD. I then visit there, again, later on. But finally, what we really needed for the project was the data from Cuba. And those days, and, unfortunately, still, to some extent, today, working in Cuba is not easy, particularly for American scientists, for a number of reasons. The primary reason is the very unfortunate and stupid US trade embargo with Cuba, which is a real obstacle, although scientific research is actually one of the exceptions in the law.But it still makes it difficult from the American side. And then on the Cuban side, at times, they’re not all that welcoming to American scientists either. We’ve been trying for many years to get to Cuba before we finally set up a collaboration with Lourdes. So, the short answer is that the data in the paper had been collected over a period of a decade or more.
HS: What about the molecular data?
JL: That was more recent. We had been accumulating tissue samples for quite some time,and there were some available in collections as well. Again, getting these samples from Cuba was what made the paper possible. Our first trip to Cuba was in 1995. I think that’s when we collected many of the tissues we used from Cuba, That was a long process, but really getting to Cuba was key for that. And then, Todd Jackman was brought on as a postdoc, and he spearheaded the molecular systematic analyses.
HS: Did you collect all the morphological data during your PhD?
JL: I collected all of it. As I was saying, it really was over the entire period.The Jamaican and Puerto Rican data mostly during my PhD, the Dominican Republic data at the very end of my PhD, and the Cuban data after that.
HS: Did the other authors also visit the field sites?
JL: Allan Larson is not a field person. He’s a laboratory scientist, although he does get into the field, occasionally. He did not go on any of our trips. De Queiroz and I had gone to the Dominican Republic in the early 1990s and we’ve gone to a number of other places. Again, as I said, what really made the paper finally possible was the work in Cuba.We had a trip to Cuba in 1995, in which de Queiroz, Jackman and I met up with Lourdes and several of her Cuban colleagues, and we traveled around and collected data then.
HS: Did all five authors ever meet as a group?
JL: No.It was even more difficult for Lourdes to get to the United States than for us to get to Cuba. She did actually visit us later, but the five of us never met together.
HS: These are sites that you continue to work in till today. How have these sites changed from the time you worked there for this paper?
JL: Well, certainly, like everywhere else in the world, things are going on and environments are changing. And much of that change is not for the better, although not entirely. Puerto Rico is actually reforesting. They cut down a lot of their forest in the last 100-150 years, but now Puerto Rico is becoming more forested than it was a few years ago. In general, habitat degradation and disruption is going on. One of the things about Anolis lizards is that the common species that you see when you go to any of the islands do quite well in disturbed habitats. And so, it fools you when you go to study, at least, the common species because they’re everywhere. You have to remind yourself, that they’re everywhere and do so well because this is trashy, disrupted habitat. But certainly, like everywhere else, there are problems, and habitat is disappearing more than being recreated.
HS: How often do you visit these sites these days? When was the last time you visited one of these sites?
JL: Well, let me think about this.I continue to work on anoles, but the focus has changed a bit. I have a long-running experimental study in the Bahamas where we go to every year; I was there in May. A lot of the other research going on with collaborators is looking at the anoles in Central and South America, comparing continental evolution with island evolution. Over the last five to seven years, we’ve traveled widely throughout Central and South America collecting similar data. It’s been a number of years since I’ve been to any of the islands you are talking about, to collect data. Some of the questions that were raised in that paper are being research, but led by other people, especially former students of mine who are now faculty members themselves. Luke Mahler at University of Toronto and Richard Glor at University of Kansas are actively working in the Greater Antilles. I know Luke was in the Dominican Republic recently working on a project that’s very much along the lines of this paper we’re talking about; going deeper and in greater detail. Both Luke and Rich are trying to get research programs going in Cuba because the diversity of that island is extraordinary.There are 64 species of anoles there, and there’s a lot left to learn.
HS: Who came up with the names for the ecomorphs?
JL: They were Williams’s names.
HS: Where was the molecular work done?
JL: It was done in Allan Larson’s lab. Our collaboration was so productive because his lab took charge of the molecular aspects of the work and my lab took the lead on the morphological work. And there were a number of graduate students who worked jointly with us.
HS: Do you remember how long it took you to write the paper,and when and where you did most of the writing?
JL: To be honest, I don’t have many recollections of the writing of the paper. The one thing I do recall are the phylogenic comparative analyses, reconstructing and inferring the evolutionary patterns of ecomorph evolution. This was at the dawn of the phylogenetic comparative method era. We didn’t have any of the incredibly sophisticated and intensive computer programs that we have now. The methods we used were based on parsimony, primarily. If I recall correctly, a number of these reconstructions were done by hand. We did simulations, which were done in a very simple computer program, and then the reconstructions, I think, I did them by hand. Actually, I remember doing some of those as I was traveling in South Africa, doing some research on a very different lizard project. I remember being in my room at night, reconstructing ancestral character states, calculating it by hand. I don’t remember the writing of the paper very much, but I do remember the reconstructions very clearly.
HS: How did you decide to submit it to Science?
JL: Well, we thought this was an exciting finding.At the time, there were no examples of this phenomenon of replicated adaptive radiation. As I said, Williams had proposed this, but it hadn’t gotten a lot of widespread attention. There were no other comparable examples. People had suggested this had happened,in very vague ways,but had not tested hypotheses in a phylogenetic framework. And so, we thought it was newsworthy and decided to shoot for the top journals. We, actually, sent it to Nature first, where it was rejected without review.
HS: Did it have a relatively smooth ride through at Science?
JL: Oh, boy. You know, it is conceivable that I could go find the files. They were on paper; not digital. I really don’t have a clear recollection. I vaguely remember that the two reviews were pretty positive. I don’t recall having a lot of trouble in the review process, but I’m not very sure.
HS: How was it received when it was published? Did it attract a lot of attention?
JL: Let me think about this for a moment. It did receive a lot of attention. It was written up as a news article in Science, which was nice. I was a little bit annoyed by that article, because the journalist, to my mind, went out of her way to find somebody who would say something that wasn’t altogether positive. It was not a particularly cutting criticism; it was off point. And, actually, I was particularly annoyed because I had told the journalist what I thought would have been a real interest to people, and she had ignored it, which was what Stephen Jay Gould said about this paper. She didn’t follow up on that. And to this day, when I tell people about the article, I say it’s relevant to the ideas that Gould was talking about, and I’d like to know what Gould thought of it. Obviously, 20 years later, that still annoys me that she had done that. I believe it got a fair amount of publicity in other places too, but I don’t recall specifically.
HS: What did Gould say about the paper?
JL: I should plug our blog Anole Annals here, because we have a post on that from a few years ago. Gould had proposed this idea that, “If you replay the tape of life, you will get a very different outcome.” That’s an idea that’s been very influential, basically that random contingency can send the evolution of life in different directions, and so, where we’ve ended up today is just the result of happenstance and a very unlikely outcome; things could have been very different. We pointed out in our paper that our results seem to contradict Gould’s idea, because you might consider independent evolution of these lizards on four different islands as four replays of the same phenomenon. You start with an ancestral anole and they diversify and yet we got the same outcome every time; a very deterministic outcome. We said that this seems to disagree with Gould’s thesis. I’m trying to remember where he discussed this; I’m blanking on it. Gould basically said that this wasn’t what he was talking about. He was talking about deep evolutionary time and the origin of animal body plans, related to the famous Burgess Shale fauna and so on. He thought this was too recent and too constrained an example, not that he disagreed with, with our result, but just that it wasn’t a test of the idea as he was proposing.
HS: Have you been surprised by the kind of impact this paper has had, and do you have a sense of what it mostly gets cited for?
JL: I’m not that surprised, because of two things. One, the paper is a nice illustration of using phylogenetic methods to study patterns of evolutionary diversification at the macroevolutionary level. Of course, it has become a very, very common, very important approach today, but back then it wasn’t. I think it has a status of a very nice early paper taking this approach, particularly, using a molecular DNA-based phylogeny to look at patterns of evolution. The interest in adaptive radiations and convergent evolution has just exploded in the last 20 years. Second, the particular phenomenon that we reported, what is now called replicated adaptive radiation, has been reported more and more in recent years. I do think that Anolis was the first well-documented example. So, the paper certainly has cachet in that respect.
HS: Is it usually cited as being the first study to show replicated adaptive radiation?
JL: I’m not sure people say it’s the first, but when they talk about that phenomenon, they do cite the paper a lot. To be honest, I couldn’t tell you exactly how often it’s cited for one thing versus another. I would say it also gets cited for the idea of adaptive radiations, in general. Anolis has become a very well-known group for the study of adaptive radiation, and this might be the most-cited paper on anole adaptive radiation. I wouldn’t be surprised. So, I think, to some extent, it’s become an easy go to reference when you’re referring to anoles and adaptive radiation.
HS: What kind of impact did this paper have on your career and your research trajectory?
JL: It was important in a number of respects.My collaborators and I had been pushing towards this project for 10 years. So, it was certainly very rewarding to see it come to fruition. As I mentioned, I had initially thought of doing my postdoc on this project, so it had been a long standing goal. The fact that it got published in a high profile place like Science meant that it got a lot of attention, and that certainly was good for our careers. And, I think it played an important role in establishing anoles as an important group to understand macroevolutionary patterns. Anoles are now widely considered as one of the best examples, along with Cichlids and Darwin’s finches. This paper probably was instrumental in bringing anoles to the attention of a wider audience. In that respect, it played an important role in our careers in a number of ways. My research program through the years has basically had two prongs: one is macroevolutionary historical studies and the other is experimental field studies of ongoing evolution.We do evolution experiments in nature. I had a paper in Nature the year before on that experimental work. And so, the combination of these two papers, I think, really played a very significant role in advancing my career. Having said that, I feel that we put way too much stock in publications in Science and Nature. Everyone agrees that silly for us to pay so much attention to these two journals, which have many issues in how they select and publish papers. But, nonetheless, the fact is that we do. So, this paper appearing in a high-profile journal and getting attention in the media certainly played a big role in moving my career forward. I do want to digress now for a moment and make sure that I acknowledge the very important significance and the incredible foresight of Ernest Williams, the man who really got Anolis research started. He’s the one who recognized that there are interesting evolutionary patterns in anoles, and that those patterns were of interest, not just for people who’re interested in Caribbean lizards, but that have broad implications for evolution in general. He really deserves the credit for starting this research program. Moreover, he pretty much figured things out correctly. His guess on the phylogenetic relationships of anoles and their implications for convergent evolution were almost entirely correct. It turns out that he was misled in a few cases by morphological convergence,but his general ideas about the phylogeny of anoles turned out to be correct. One of the funny things about that is he based his inferences on morphological characters, osteological skeletal characters, but, as I mentioned earlier, his approach of inferring the phylogeny was very much using intuition.There were no fancy programs back there to build a phylogeny. It was the expert looking at the data and intuiting the pattern. Yet, he mostly got it right. And the funny thing is that people have gone back and done formal analyses with osteological datasets and they do not get his results. They use cladistics and fancy computer programs, but cannot get the result that Williams got just by eyeballing the data and using his intuition. I take that as a credit to Williams. Somebody who really knows the organisms does better than then statistical programs, with all their assumptions and so on. Anyway, Williams deserves a huge amount of credit for seeing the opportunities, the interesting patterns, and for inferring correctly how the ecomorphs had evolved. He did this in a paper in Evolutionary Biology, which came out as a volume once a year. In this paper, he basically inferred phylogenies and then mapped characters on those phylogenies to infer evolutionary history. This was an approach that was years ahead of anyone else. Now, the methods were quite primitive, but the approach of mapping characters on a phylogeny to test hypotheses and draw inferences was well ahead of its time. That’s a paper that really deserves a lot of credit for forging the way forward.
HS: Does the phylogeny look more-or-less the same even today?
JL: It does, and that’s very gratifying. We’re very happy about that. That phylogeny was built on mitochondrial DNA. As people have now shifted to looking both at nuclear genes and much larger data sets, mitochondrial DNA often turns out to give a misleading signal, to some extent. The nuclear data sets for anoles have taken a little longer than we would have liked to come along, but we’re getting there now. But it turns out that the mitochondrial DNA is pretty much correct. Minor details might be different, but the big picture has not changed at all, as we get better data sets.
HS: In one of the notes in the paper, you say that you don’t know what the ecomorph of the ancestral state might be,and because of that, there’s uncertainty in the number of transitions. Do you have a better idea of that now?
JL: No, I think the whole approach just doesn’t work. This is ironic, because I published a paper – might have been part of my thesis – in 1992 where we reconstructed the order in which ecomorphs evolved, using parsimony methods. It was a cool paper. It suggested that the anoles on two islands had evolved following the same pattern. But I now think the whole exercise is hopeless, because when you have a group in which homoplasy is rampant, any kind of confidence in inferred ancestral character states is very limited. To infer ancestral states, you need there to be a strong phylogenetic signal. The assumptions of these methods simply aren’t met when these traits evolve relatively rapidly. I just think there are some questions we can’t answer with phylogenetic methods. This is one of those cases.
HS: You also say that you didn’t include all species in the analysis.When you built phylogenies subsequently did you include more species?
JL: Yes, more species have been included in subsequent papers. We’re trying to get all species of anoles, but getting some out of Cuba, and some of the more remote places in the Dominican Republic is difficult. So, we haven’t gotten there yet. But the goal is to have an all species phylogeny. We’ll get there eventually. But, certainly, the phylogenies that were published subsequently are more inclusive than what was published back then.
HS: In another place in the paper you mention that phylogenetic relationships of two species from Hispaniola – koopmani and darlingtoni – are unknown. Do we know more today?
HS: Yes, we do,and it’s very interesting. They are not closely related to similar ecomorph species on Hispaniola. They represent multiple origins of an ecomorph type on a single island. Although, it’s a little complicated because, Hispaniola, the island that contains Haiti and the Dominican Republic, in the past was multiple islands. And so, it’s possible that these ecomorphs evolved when the islands were separate. We’re not certain about that.
HS: Would you say that the main conclusions of this paper still hold true, more or less?
JL: Yes, absolutely.
HS: In the last paragraph of the paper, you say, “The phylogenetic analysis reveals only two cases in which an ecomorph has evolved more than once on a single island. Interspecific competition, which is intense among anoles (23) and may drive their adaptive radiation (9, 24), is probably responsible; once an ecomorph niche is filled on an island, other species are excluded from utilizing that niche. Thus, the importance of historical contingency depends on the frame of reference: Among islands, it has little discernible effect in that the same ecomorphs evolve on each island, whereas within each island, prior evolutionary events limit the options available to particular species and thus determine the directions in which evolution can proceed.” Could you reflect on these lines in the context of what we have learnt since?
JL: I haven’t read that paragraph in quite some time, and I’m quite pleased to hear you read it, because I think it’s all true today. Everything we’ve learned since then has not changed that perspective in any significant way.We know of a couple more multiple origins of an ecomorph on an island, mostly on Hispaniola, which is a geologically complicated Island. But, for the most part, those conclusions hold today. I think that the contrast we drew between lack of contingency at the highest level, but evidence for it at the lower level, is still an interesting contrast that turns out to be correct.
HS: If you were to do a redo this study, or do a similar study,what would you do differently?
JL: The biggest difference would relate to the standards in the field and what is possible.We’ll certainly get much better nuclear markers, maybe incorporate morphological data, which had been our goal then as well. More species is always nice. I think, if we did it today, we might try to get more of them. But, in general, I would take the same approach we did then.
HS: Have you ever read this paper after it was published?
JL: I have. I can’t remember exactly why. We taught it in some courses. Of course, we’ve done further papers along these lines, and we needed to go back and reread what we said. So, yes, I have read it periodically.
HS: Would you consider this as one of your favorite pieces of work?
HS: What do you like about it?
JL: It highlighted an important phenomenon – replicated adaptive radiation – that hadn’t been widely studied. I think it got people thinking, in general, about whether you see this phenomenon in other groups. It certainly is, to my mind, the most interesting single aspect of anole evolution. And, basically, my entire career has been built around studying things related to anole ecomorph evolution and replicated adaptive radiation. And so, certainly, much of what I’ve done in the last 20 years has followed from this paper. It was a watershed paper in my career. Certainly, if you had to ask me my favorite couple of papers, it would certainly be in that list.
HS: What would you say to a student who is about to read this paper today? Would you guide his or her reading in any way? Would you point them to other papers that will be useful to read along with this? And would you add any caveats to their reading of this paper?
JL: That’s an interesting question.Well, I would suggest that perhaps they should read Ernest Williams 1972 to get a bit of a historical perspective about where this paper came from. Let me think about other papers that would make a nice parallel with this.I think some of papers on other interesting groups would be worth reading. Rosie Gillespie’s paper on spider evolution in Hawaii published in Science a few years later was a nice contrast to our paper, with results that were similar in some ways and different in others. That’s a very nice system and a very nice paper.The cichlid literature is fascinating, but no single paper stands out on this particular topic of ecomorphological evolution, in part, because there are so many species in those lakes, and it’s a little bit unwieldy to take them on. But they should certainly look at some of the work on cichlids that’s come out of a number of laboratories, mostly in Europe, in the last few years. Cichlids are a fascinating contrast, and I would point people towards them. And then, on the general topic of adaptive radiation and convergence, Dolph Schluter’s book, The Ecology of Adaptive Radiation, even though it’s now close to 20 years old, is just a fabulously interesting book, still full of ideas worth following up on. Finally, I would mention my book, Improbable Destinies: Fate, Chance, and the Future of Evolution, which was published in 2017. Written for a general audience, the book both describes my personal story of studying anole ecomorphs in the context of Stephen Jay Gould’s ideas about the role of historical contingency and determinism in evolution.