Revisiting an old paper, but a good one

I have to brag a bit in the title because if you say a paper is an “oldie,” you have to also say “it’s a goodie”. But I think this one is—it’s the first of two papers I wrote with my then-grad-student Allen Orr on the time course of speciation in Drosophila.  And it’s one of the few good ideas I’ve ever had. I don’t know how often it’s been cited—I don’t look up stuff like that—but it has been influential in inspiring others to do related work. I’m writing about this paper because I recently revisited it in an interview (see below).

Here’s a very brief summary of what we did. I realized one day, when I was at the University of Maryland, that there existed a tremendous amount of data about the sexual isolation and hybrid sterility/inviability of various Drosophila (fruit fly) species tested in the lab. There also existed, separately, a large amount of data on the “genetic distance” between these species as judged from gel electrophoresis. This genetic difference is a rough measure of the times since the species diverged. The more similar the electrophoretic profiles, the younger the species. (The actual real-time calibration of the distance is hard, as Drosophila has no fossil record, but we did our best.)

You could, I realized, take various pairs of species, see how much reproductive isolation they had between them—how much mating discrimination and whether the hybrids were viable and fertile—and correlate that with the genetic distance between members of each pair. If you plotted genetic distance against the degree of genetic isolation, you could get a “time course” of speciation, seeing which forms of isolation evolved earliest, what rate they evolved at, whether it would make a difference if the species lived in the same or different areas, and so on.

Of course there are lots of issues here, one being that measures of reproductive divergence between various pairs of species aren’t evolutionarily independent, so we had to do phylogenetic corrections. Further, sexual isolation and sterility/inviability are only two of the reproductive barriers that separate species, and we had to neglect types of genetic isolation that could operate in nature but couldn’t be measured in the lab (e.g., different preferences for food or microhabitats).

The results, though, were surprisingly clean and enlightening. For example, we found that sexual isolation—but not hybrid inviability—evolves ten times faster between species now found in the same area than those now found in different areas. This result, which has held up in repeats of our work, suggests that natural selection “reinforces”, or strengthens, mate discrimination between species when they live in the same place. That’s probably because there is a genetic penalty to be paid, in the form of hybrid problems, if you actually mate with the “wrong” species; and you only have that kind of selection operating in species that live in the same area, and have a chance to produce hybrids.

Here’s a graph from the second of our paper of papers showing two plots of the degree of sexual isolation between pairs of species (y axis) against their electrophoretic genetic distance (a measure of the divergence time between members of each pair). “Allopatric” taxa are pairs of species that are geographically isolated at present, while “sympatric” taxa are pairs of species that live in the same general area. (These data are phylogenetically corrected.) You can see that the degree of sexual isolation appears much earlier (at lower genetic distances) when the taxa live in the same area. This is a very striking result that is highly statistically significant. It suggests that natural selection operates on species living in the same place to “reinforce” their sexual isolation. You don’t see this difference for hybrid sterility or inviability, which are not expected to be reinforced by selection.

I digress, but it’s nice to think about this good old work.  Allen came on board the project at the beginning, and we spent several years collecting the data (which was scattered all over the literature), calculating statistics when only raw data were given, and analyzing the data. Thus the paper didn’t come out (in Evolution) until 1989, three years after we’d moved to Chicago.

Then electrophoretic data and reproductive-isolation data continued to accumulate, so in 1997 we published an update of the 1989 paper. The additional data confirmed the patterns we’d seen before. And now, since nobody does electrophoresis any more, and estimates of genetic divergence come from DNA sequences, we can’t do this analysis further. (DNA-sequence data does not exist for most of the species we used.)  Similar work has been done in fish and tomatoes, and at least two researchers have redone our analyses in flies using different techniques (the conclusions remain good).

The references to our two papers are given at the bottom, along with the links to them (free access).

This long introduction just wrote itself, when what I really want to do is call your attention to an interview I did about that first paper with Hari Sridhar at his site Reflections on Papers Past. Hari, a a post-doctoral researcher at the National Centre for Biological Sciences, Bengaluru (formerly Bangalore), India, has been interviewing scientists about well known papers in ecology and evolution since 2016.  He was kind enough to interview me about the first Coyne and Orr paper, and you can see the interview by clicking on the link below. I haven’t read the final version, which is a transcript of an audio conversation, so be aware that it’s spoken language. I did read a draft and corrected a few phrases that were unintelligible over the phone.

If you’re interested in papers in ecology and evolution, you might have a wander round Hari’s site; there are lots of interesting papers and interviews, many with people I know.

Click below to see the interview.

I want to add that although Allen was my grad student during much of the time we wrote these papers, it was a total collaboration. As with all my students, I don’t micromanage their work or ever tell them what research to do. Allen was interested in the project from the beginning, and contributed tons of work and many ideas to the two papers. And our collaboration continued in what I consider my most important scientific accomplishment, the book Speciation (Coyne and Orr, 2004; note that the book is now expensive but was about $50 when it first came out).

Here are Allen and I at the Evolution meetings in Portland in 2010. Allen was president of the Society for the Study of Evolution, and I was an incoming President, so he briefed me about the job. We had a great time in Portland, as that was before the city went nuts.

And as a measure of the fame of our work, you can’t get bigger than this. My collaboration with Allen was featured in the 2001 movie “Evolution” (a dreadful film!), as a scrawled reference on the blackboard behind two of the stars, David Duchovny and Orlando Jones. See below. It says “Read Coyne and Orr.  ‘Drosophila’ pp. xx8-450”. Note that the page numbers don’t correspond to either paper that we wrote, though it may refer to the book. But even in the book those pages don’t correspond to anything that would be a reading assignment.

Another ex-student of mine, Mohamed Noor, called me up and said he’d seen the movie and noticed a reference to our paper on the blackboard. I didn’t believe him, so I had to go see the movie myself. Sure enough, we were in there! Someone later sent me a screenshot (below).

I would call that real fame! Pity they got the page numbers wrong. I’ve always wondered who wrote that on the board and how they knew about our work.

_______________

Coyne, J. A. and H.A. Orr.  1989.  Patterns of speciation in Drosophila.  Evolution 43: 362-381.

Coyne, J. A., and H. A. Orr. 1997. “Patterns of speciation in Drosophila” revisited. Evolution 51:295-303.