In 1964, Paul Ehrlich and Peter Raven published a paper in Evolution that reported the findings of their literature survey of larval food plants of butterflies. Based on the patterns they found, they hypothesized that caterpillar-host plant relationships result from “coevolution”, and that the secondary chemicals produced by plants, which till then were thought to be excretory products, were probably evolved by plants to defend themselves against herbivory. Fifty-two years after the paper was published, I spoke to Paul Ehrlich about the origin of this paper, his collaboration with Peter Raven, and the impact of this paper on coevolution research.
Citation: Ehrlich, P. R., & Raven, P. H. (1964). Butterflies and plants: a study in coevolution. Evolution, 18(4), 586-608.
Date of interview: 29th September 2016 (via Skype)
Hari Sridhar: What was your motivation to do this study? By this time, you had already done a lot of work on butterflies, but was it around this time that you started getting interested in questions around evolution? What triggered this interest?
Paul Ehrlich: My entire research programme was aimed at understanding how populations evolved. I was working on a series of natural populations right on Stanford campus, and one of the questions I was interested in answering was whether populations lost genetic variability as they shrunk or whether there would be heavy selection for maintaining genetic variability. So I was always interested, I primarily thought of myself as an evolutionist in those days, particularly because ecology was a semi-dead subject. There was almost no theory, no great principles and though I don’t think there is any real division between ecology and evolution I was calling myself an evolutionist then. The beginnings of coevolution actually came from a simple question that I asked my colleague Peter Raven. We used to have coffee together every day and talk about evolution. Peter was, and still is, the greatest plant evolutionist. I don’t know anything about plants. I was lucky to have colleagues that could easily identify them for me and tell me whatever I needed to know. Not just Peter, but also Richard Holm and John Thomas on our faculty here, so my closest colleagues were botanists and that was good for me because I was totally ignorant about botany. One day at the coffee table I said to Peter: you know it’s crazy, these butterflies I’m working on feed on Plantaginaceae or Scrophuliaraceae, and it just doesn’t make any sense. And he said: of course it makes sense. Didn’t you know that the Plantaginaceae – the plantagos – are just wind-pollinated Scrophuliaraceae? I said: no, I’d get rid of any graduate student that carried that kind of trivia around in his or her head! I didn’t have any idea. So, this started a conversation between Peter and me on what butterflies’ caterpillars ate. And I was really up on that because I’d worked on butterflies my entire life. I was then also in the middle of writing a field guide to the butterflies of North America. And it gradually dawned on us as we looked in the literature that there was more information on the food habits of caterpillars – butterfly caterpillars- than any other large group of plant-eating insects. And that’s because collectors like to have perfect specimens. They would go out and find the caterpillars, raise them, wait till the butterfly emerged from the pupa – the chrysalis, the resting stage where the caterpillar’s turned into an adult. The caterpillar is sort of an eating growing machine and the adult is a reproducing dispersing machine and collectors get the adult to emerge from the chrysalis, wait till its wings are fully expanded, and it’s perfectly beautiful and unmarked, unlike how it would be if you caught it in a net or it had flown around for a while. They would kill it, mount it, put it in their collection, and then if nobody before had known what the caterpillars ate, write a little note for an entomological journal describing what the food plant was. So we were able to mine that enormous pile of literature, and it quickly became clear to us that there were many patterns that were related to the so-called secondary chemicals that the plants produced which were then thought to be excretory products. And it has always been stunning to me that Peter or I or anybody else – actually there are couple of exceptions – but basically we just assumed that the plant physiologists were right and these were just excretory products. It never dawned on Peter or me to ask why any organism would put a huge amount of energy into making an excretory product and then stick it in its ear, and not excrete it. It became clear to us that the plants were making poisons to keep insects from eating them and butterflies were evolving ways to overcome those defences and there was sort of an evolutionary arms race in which the plants were trying to develop something that would keep the butterflies off or kill their caterpillars and the butterflies countered that by evolving things that would allow them to detoxify the poisons, eat parts of the plant not loaded with poisons, and so on. That was the basis of our 1964 paper. We coined the term coevolution because we thought there was not enough attention being paid to this in those days. If you looked at the evolutionary literature it was much about evolving the ability to survive cold temperatures or live in more salty water or things like that. The fact that organisms were putting evolutionary pressures on each other wasn’t acknowledged enough. We’re enormously proud of the paper because we did it without looking at a single organism living or dead – we only looked in the literature – proving the old saying that “study nature, not books” is wrong! If you want to get a paper that is highly-cited you should study books and not nature!
Darwin had ideas about this. If you read “The Origin of Species” carefully you’ll see that he had coevolutionary ideas. Wallace had the basic ideas of natural selection that Darwin had, but Darwin published with a lot of data, his book was loaded with examples. That’s why we speak of Darwinism rather than ‘Wallaceism’. And I think that our paper took the same approach – we had a massive amount of data on what butterflies ate, which was interesting in itself and also boosted the idea of plant-herbivore coevolution. As you know it has now spread in all kinds of directions. In sociology, I think Steve Schneider wrote a book on coevolution of climate and life. Robin Dunbar wrote about the coevolution of the neo-cortex and group size in mammals. So we just had very good luck. We came up with an interesting idea, had a lot of data to support it, coined a word that seemed obvious and it took off. I think the last time I looked on Google Scholar there were, I can’t remember but a huge number of citations. I don’t pay much attention to that sort of thing but you can check that if you want.
HS: Was the term “coevolution” used for the first time in this paper?
PE: I don’t know. It may have been mentioned elsewhere by somebody before because it’s an obvious idea, but as far as I know – and, of course, I’m an extremely egotistical person – is that Peter and I really introduced it into the biological literature. Even today, if you look, if you check papers on coevolution many still cite our paper. There are papers saying let’s look back on it or what’s happened since Ehrlich & Raven, were Ehrlich & Raven right and so on. Exactly what any self-promoting scientist wants to hear!
HS: Was this the obvious term to use for it? Did you consider other terms?
PE: You are asking me to think back 50 years! Peter and I were already friends, and we are still friends. We spent a weekend together about a month ago, with his wife and my wife too. It’s very hard to sort out what idea came where.
HS: Stepping back a bit, could you tell us how you got interested in butterflies?
PE: That’s easy. I went to summer camps that had lots of beautiful butterflies around. I started out in a camp of YMCA, even though I wasn’t a Christian! And there were lots of beautiful butterflies. I can still remember seeing Battus philenor, which is a poisonous butterfly that serves as a model in a mimicry complex. I remember watching them nectaring and so on. I went to a summer camp after that in Vermont where the nature councillor started people on collecting butterflies and I got totally hooked on it. So hooked that if I had a large place filled with cabinets I could start a collection again tomorrow! I’m a real collector. When you look at a drawer full of butterflies and you see the changing patterns it just shouts evolution at you. It gets you like bird watching out in places in the world you would never go otherwise. I’ve always had a travel lust and I’ve been almost everywhere and it’s been driven by chasing butterflies and watching birds.
HS: When you had that initial discussion with Peter Raven, what was the species of butterfly you were working on that you mentioned to him?
PE: When I came to Stanford in 1959, I wanted to set up a situation to study those questions I mentioned earlier, about what happens to genetic variability and how things evolve in response to changing situations and so on. In the fall of 1959, I went to a meeting of the West Coast Lepidoperists’ society in Santa Barbara. When I was an undergraduate, the guy who eventually became my major professor – Charles Michener, who died this year at the age of 96 – had told me about the Lepidoperists’ Society and I had joined it when I was 15. Mich at the time was working on butterflies. He turned out to be the world’s greatest expert on bees and bee evolution and behaviour. I’d known him since I was 15 and he was 29. We only published one paper together even though we were in contact for 60 some years. Do you want to hear about that paper? We showed that, in Costa Rica, if you left 10% of your coffee farm in forest, rather than cutting it down and planting it in coffee, you made more money, because the forest was the habitat of the pollinators, and pollinated coffee is worth more money than un-pollinated coffee. The beans are better and you get more money for them. The lead author on that paper was a guy named Taylor Ricketts who was a graduate student at the time. He is now the head of the Gund Institute of Ecological Economics at the University of Vermont. The second author was my colleague here, Gretchen Daily, and she was Taylor’s major professor. The third author was me, and I was Gretchen Daily’s major professor. And the fourth author was Michener, who did the bees, and he was my major professor. So we had four academic generations, on the same paper, in the right order!
HS: Give us a sense of how this study progressed from the time you had the initial idea. How long did you take to do the analysis and put the paper together?
PE: Not accurately, but I would guess about 3 months. I think one of Peter and my common characteristics is that we tend to work fast, get it done and get it out there. I would imagine that is roughly the time, but I’m not sure. It could have been longer. I think the paper came out at the end of 64 or early 65. And the initial question I asked Peter was probably around April of 64. And consider that it takes time to write the paper, after we submit it goes out for review, then there is a drag time and you have to read proof and so on. So I imagine we worked on it hard for several months and then did the other stuff as time went on.
HS: Had you already compile the data or did you need to do it fresh for this study?
PE: No, we compiled it fresh. We knew where it existed, because I knew the butterfly literature very well and Peter knew the plant literature very well. He knew the relationships of the plants and I knew the relationships of the butterflies. So we coevolved! It was a matter of putting it together and checking things out, because I didn’t have in my head what plants that several thousand species of butterflies ate. I just knew where that information was and I had a good library that covered the butterflies I’d been building because I was using butterflies as my experimental system in evolution and ecology.
HS: So it was a mutualism between an entomologist and a botanist!
HS: In the paper you say “An excellent review of a long series of experiments pertinent to this subject has recently been presented by Merz (1959). Much of the following is based on his account.”
PE: I have no recollection of this. Do you want me to look at the paper?
HS: Sure. It looks like it is in German.
PE: Dear god! Do you speak German?
HS: No, I don’t
PE: Me neither, not much, I had to learn it for my PhD. Let me see if I can find it in a minute. I don’t want to waste your time. You ask me another question while I’m looking and I’ll mumble at you.
HS: Was submitting this paper to Evolution an obvious choice at that time? Did you consider other journals?
PE: The most obvious choice. If I had any idea how important it was to become, I’d probably have tried to submit to Science and they would have turned it down! It’s too data rich. I think we made exactly the right choice. I can’t imagine where else we would have thought of sending it. I just discovered it became a citation classic in 1984.
HS: Did it go through peer-review easily?
PE: I think so; I don’t remember anything special about it. I’ve had problems on some papers but not on this one.
HS: And at the time of the paper was published did it attract a lot of attention?
PE: It got a fair amount of attention because we basically said that the botanists were wrong. If you look at the whole thing on the secondary compounds being excretory products, other people had mentioned before that that was not the case, but we had the evidence. If you go and dig up a botany text from, say, 1955 or something like that, you’ll find maybe four pages on herbivory. If you wrote a botany text now – I don’t know if they do anymore – you’d find half the book is on herbivory. So it was from that point of view, controversial, and there was a lot of discussion in meetings and so on.
I’m getting closer. Wait a minute. Okay, I got the paper now, we’re talking about Merz.
HS: You mention the Merz paper specifically as being pertinent to the subject. I wanted to know whether it played an important role and was some kind of motivation in the making of your paper.
You say “an excellent review of a long series of experiments pertinent to the subject has recently been presented by Merz (1959)”.
PE: All I can see now is the title of the paper. I have zero memory of it from 50 years ago. But the title is “Plants and Caterpillars”. “Gross-schmetterlingsraupen” is macro Lepidoptera caterpillars, but I have no idea what the word Futterwahl means. And even if I had the paper I probably wouldn’t have much more idea! I got through the German exam in Kansas for my PhD through treachery and deceit, not by learning German!
HS: In that section of the paper, you talk about Merz’s experiments, and also work by other insect physiologists – Dethier and Thorsteinson.
PE: Let’s see, where I go to find the thing in the paper.
HS: Page number 587.
PE: 587, okay hang on. It’s a long paper, jeez. Okay. Okay the experiments by Merz. He basically removed one of the plant defences and found that then the caterpillars would eat it. So that was a fairly fundamental thing. The plant was covered with glandular hairs, it would have been likely coat the larva with nasty stuff and so they tended to be repelled by it. Yeah, what else is there let me see.
Oh yeah, people were interested in how plants defended themselves. Vince Dethier was a very bright guy I vaguely knew but I can’t remember. He’s better known for some neurobiology work as I recall. But again you are asking me to go back to a literature I haven’t seen for 50 years. I don’t know. But we were not the first people to think about how plants defended themselves right. There are lots of ideas of evolution out there, when Darwin put together the whole thing with the evidence, but one of the things that we had was so much evidence that they could not ignore whether or not these things were excretory products. I’m mumbling because I’m looking for Fluck (1963) now. There we go. Yeah, he had an article, fairly recent one, on secondary plant products but I don’t know why I can’t tell what it said from the title.
HS: You list a number of people who helped you by evaluating the validity of public records and commenting on unpublished data. Were these people you knew personally?
PE: What page is that on?
HS: This is on page 589.
PE: 589. You have really taken me back; I’m going to become nostalgic. Yeah, I got it. I’ll just go down the list and comment
d’Almedia: I knew by reputation only
Bellinger: I had met
de Biezanko: Had down a lot of stuff in Brazil
Lincoln Brower: Was a friend
Clarke: I knew faintly
Harry Clench: Harry and I had described the only distinct new species of butterfly found in North America before this work was done, so Harry was very close and very helpful
Comstock: I don’t remember
Dickson: I don’t remember
Downey: Was a friend
Etcheverry: I knew faintly
Hayward: I only knew by reputation
Same with Howarth
Same with Iwase
Langer: I knew by reputation only
Don MacNeill: I knew well, he was local
Murray: Doesn’t ring a bell
Sevastopulo: Was a big deal in Africa and I knew from the literature
Shirozu: Ditto in Japan
Shull: I only knew by reputation
van Someren: I knew by reputation
John Downey: Was a friend
Gordon Orians: He is a very distinguished ecologist one, of the leading, you must know him. Gordon is still going and he’s brilliant
Geissman and Thorne were botanists.
So you got some idea?
HS: Yes, that’s very useful, thank you so much.
PE: Remember I had spent a lot of time exchanging butterflies as an amateur, I was writing the butterfly book and so on, and so I had contact with a lot of these people anyway.
HS: I want to go over two sentences that I see as the main conclusions from this paper and I’d like you to reflect on them today. First, you say that secondary plant substances played the leading role in determining patterns of utilization. Would you say that is still generally true?
PE: I think that’s generally true, but do you want to get some outside views on that? I don’t know, I think it is generally true, but people like John… I’m having a senior moment John. …I will find it… but there are people out there, a lot of good people that are working hard on coevolution today and it’s gone a long way. There are many people who can give you a more current view on that. But I don’t feel my reputation is at stake if it turns out that’s not true. I think plant secondary compounds are the major way that plants are defending themselves against herbivores in general but that doesn’t mean that there aren’t many plants that defend themselves against herbivores by having spines.
HS: Is John Thompson the name you thinking of?
PE: Thompson yeah. John is very bright and he’s been on top of it. There’s also a lady Judith Becerra or something very close to that. She’s got a lot of good work and she’s I think in Mexico
HS: Towards the end of the paper you say “the importance of reciprocal selective responses between ecologically closely linked organisms has been vastly underrated in considerations of the origins of organic diversity. Indeed the plant-herbivore interface maybe the major zone of interaction responsible for generating terrestrial organic diversity”. What are your t your thoughts on that today?
PE: People have written a fair amount on that subsequently and I think my guess is largely true. But don’t have any personal research on that that would bear directly on that. That is not an area that I’ve been to again. I faded on the coevolution game probably 30 years ago.
HS: Yes, I was looking it up and if I’m not mistaken I think 1987 is the last time you wrote a paper with the word coevolution in the title.
PE: That’s possible. I could have put it on the title of my favourite paper for this year. That is one that we wrote…I’m mumbling I’m also trying to find Judith’s surname. Yeah, it’s J. X. Becerra. B. E. C. E. R. R. A.
Now back to your question: we could have put coevolution in the title of the paper that I just published with a colleague. It looks like many butterflies actually have pictures of aposematic (poisonous) caterpillars on their wings. Talk about the coevolution of plants and butterflies! You get the caterpillars having defences against plant poisons and then using the plant poisons to defend themselves, and then butterflies having pictures of the caterpillars on their wings to divert predators from the butterflies making them think the butterflies are poisonous.
HS: Is this something that’s discovered recently?
PE: Just been discovered by an artist colleague of mine, Darryl Wheye, and it’s still controversial. So I have actually got my fingers back into the game.
HS: By the time of this paper was published you were already fairly well established in the field, but did this paper have any kind of direct impact on your career?
PE: I think so. I think this work, and the work that I did on the Checkerspot butterfly are the most cited. I was studying two species of Checkerspot butterfly: the Bay Checkerspot (Euphydryas editha) and the Chalsadone Checkerspot (Euphydryas chalsadona), and these were what I was asking Peter about, which lead to this paper, and I continued to work on them until E. editha went extinct on our campus here a decade or so ago .The coevolution work is certainly the best cited scientific work I’ve ever been involved in, the second would be the checkerspot work probably, and then a lot of other stuff. I’ve moved mostly into human biology by the eighties and nineties but I still kept my hand in reef fishes and things like that. Who knows, but I think it certainly didn’t hurt my career. On the other hand, in a sense, the butterfly work hurt my salary! I was doing the work here in a reserve on Stanford campus called Jasper Ridge. I had worked very hard to establish the ridge as a reserve. At that time my salary was terrible at Stanford and I really resented it. So I went to the provost, who I knew quite well, and said: why are you paying me such a lousy salary, it’s just disgraceful. I have more publications than anybody else in the biology department, maybe more than anybody else in the university and you pay me this crappy salary. He said: Paul, we know you won’t leave because you’re stuck with your work on Jasper ridge! I had lots offers to go to other universities but my long-term research was on Stanford campus. That tells you something about academic politics!
HS: And till today you are at Stanford
PE: Well, I retired actually. I realised because I was so long in the retirement plan retiring actually raised my income.
HS: After writing this paper, did you start doing empirical work on coevolution soon afterwards?
PE: Immediately. The most important single thing that I did afterwards, which basically convinced me that Peter and I were absolutely correct, was in Colorado where I was doing field work on butterfly populations. There is a little Lycaenid butterfly, a so called “blue,” named Glaucopsyche lygdamus which feed on lupines. I noticed this butterfly laying its eggs on Lupine inflorescences, on the flowering stalks of Lupines. It occurred to me that there is a very simple experiment I could do, and I recruited a fine botanist, Dennis Breedlove to help.
Even though there were biting flies that pretty much destroyed us during the process of it, which was to label up a bunch of plants in pairs and then, using a hand lens, go over the inflorescences of every other plant and remove the eggs. You can find the paper easily enough or I can send it to you, but the main finding was that when we removed the eggs the plants had enormously greater seed set. I can’t remember exactly what it was, but I think something like 50 percent of the reproduction was lost on plants from which we didn’t remove the eggs. This made it very clear that the tiny butterfly was an enormously important predator and an evolutionary force on the lupines, because when something destroys half your offspring or two-thirds of your offspring every year, that’s a really big selective force. I did other stuff related to blue butterflies and their food plants in Colorado for awhile. That waste main direct experimental stuff I did on coevolution. Later I’ve also looked at coevolution in reef fishes and things like that. It’s everywhere, that’s the interesting thing about it. Once the idea got out there, and again I’m not sure we invented the word but the idea got into the literature through us, it went everywhere, in sociology and all kinds of biology. It’s a term that’s still widely used today in all kinds of contexts
HS: You said in the 1980s you started working more on human biology. Was this a conscious switch?
PE: A number things led me to it. One was the fact that the population situation was getting less and less attention, particularly in the early 90s when my colleagues and I thought it was desperate. I think that if you just look at the whole world situation today the main action is required in areas we think of as social sciences. Questions like how people could possibly vote for Donald Trump is a big question in my mind. Why haven’t people been able to make the fairly simple changes that most ecologists and evolutionists would say are necessary to preserve our civilization? I spend a lot of time studying that and I wrote a book called “Human Natures” which covers that. Early on, I did the first work with a psychologist on the effects of crowding on human behaviour. I love being in the field, for instance showing that the little butterfly really killed a lot of the reproductive capacity of Lupines. I just love doing that stuff and so I will still try and keep my hand in that area. I can’t work on reef fishes anymore because, at 86, my back is so bad that I can’t get my flippers on and off hanging off of a boat, but otherwise I’d still be doing reef fish research. But as far as the lab goes, I’m more interested in looking at the entire world situation. Have you ever been to the MAHB website?
PE: Go to mahb.stanford.edu. If you have any interest in saving the world it takes about two minutes to sign up and you can post your own thoughts. That’s the kind of thing I’m involved with. I’m involved with a big group trying to figure out how to get people thinking about not only preventing a collapse but how to recover if we have one, which seems very likely.
HS: Though you have moved away from doing research on butterflies and coevolution, do you still keep track of the literature?
PE: Not closely. I keep track of the big things but I’m too busy trying to keep track of literature on what’s happening to the world, between toxics and climate change, population growth, over-consumption, the chances of nuclear war, the chances of vast plagues, land-use change, the world food problems etching trying to keep track of all that literature I have to ignore the things that actually interest me more. That’s why I had such a good time with the recent paper on caterpillar patterns on butterfly wings.
HS: in the 50 years since this paper was published have you ever read it again?
PE: I occasionally look at it again. I’ve never read it from end to end again, but people sometimes, like you, send me a question and I look it up.
HS: If you compare this paper to the papers you write today ,what would you say are the most striking differences in the way you write?
PE: The most striking difference is that the way I write today is trying to influence us to have a world where the issues raised by Ehrlich & Raven in 1964 can still be studied. Where we can learn more, where we will have a civilization that will support people like you to look at heterospecific flocking in birds and there will actually be birds to flock. So the main difference is that I’m not looking at interesting questions in evolution and standard evolution and ecology nowhere near as much as I would like to. Although, I’m in the middle of a book right now and I’ve co-authored a paper in review in Proceedings of the National Academy of Sciences on population extinctions, and I have done quite a bit on the extinction crisis. So I still try to keep my fingers in that area but I have to be very choosy.
HS: Would you count this as one of your favourites from all the papers you published?
PE: Probably my favourite. We live in an extraordinarily complex universe and a complex planet with billions of populations of millions of species. That we were able to find a principle that explained an awful lot of it in very diverse areas and generated what by now is a gigantic literature was a real pleasure. I would say my favourite papers would be this one and then possibly what I just sent you, about caterpillar patterns on butterfly wings, because it allowed me to get back into an area that I love.
HS: What would you say to a student who’s about to read this paper today? What should he or she take away from this paper published 50 years ago? Would you add any caveats?
PE: When my graduate students come to me with an idea for a dissertation, I always say: okay, that’s an interesting idea, but tell me, if this idea works out really well, what would the title of the paper you would submit to Science be? That is really important. A student starting out, at least in the kind of profession you and I are in, has to establish a reputation in some area by working on something that is of broad enough interest that they can try to publish in the journal like Science or PNAS. So I would tell a student reading this paper, not to focus so much on the specifics, but to pay attention to how we managed to find a general principle in a lot of noisy data.