My all-time favorite scientific paper


Current events (E.O. Wilson saying that scientists don’t need to be good at math) give me a great reason to introduce what might be my favorite scientific paper.

I have three reasons for choosing this paper to share with you. One minor reason is that, from one ant man for another more illustrious ant man, I’d like to be one of the few scientists to publicly say something nice about E.O. Wilson this week without any kind of caveat.

Second, the content of this paper, and the fact of its existence, frames Wilson’s message about science and math that dovetails with my recent writing on how to design a research program.

Last, since this paper was published it has been a source of inspiration to me as a scientist.

Without further ado, here’s the paper:

Wilson, E.O. 2005. Oribatid mite predation by small ants of the genus Pheidole. Insectes Sociaux 52: 263-265. There is a paywall – email me if you’d like a copy.

Here is the abstract of this three-pager in its entirety:

Using “cafeteria experiments” with forest soil and litter, I obtained evidence that at least some small Neotropical species of Pheidole prey on a wide array of slow-moving invertebrates, favoring those of approximately their own size. The most frequent prey were oribatid mites, a disproportion evidently due in part to the abundance of these organisms. The ants have no difficulty breaking through the calcified exoskeleton of the mites.

What is the deal with this, and why is it inspirational? Please humor me by reading on if I haven’t lost you already.

This paper was published in the year 2005. In 2003, after several decades of effort, Wilson had published a monumental revision of the most species-rich genus of ant, Pheidole. Any taxonomist can appreciate the sheer enormity of this effort that had Wilson’s attention over the years. Clearly, it’s a work of love. Most Pheidole are tiny in size. They’re charming little ants, if nondescript, and not really different from one another in obvious ways that could account for their richness.

Like most years, 2005 was a good year for Wilson. He wrote three PNAS papers, two with his long-time friend and colleague Bert Hölldobler. He also wrote a controversial paper in Social Research arguing that altruism doesn’t principally arise from kin selection, a precursor to Wilson’s now full-fledged group selection posture. He had a book chapter come out, oh, and also he published a big book introducing the concept of gene-culture coevolution. And then there was this little paper, one of my favorite papers ever, in Insectes Sociaux.

If you want to understand and measure the diversity of ants, the first place to start is to sample the leaf litter. A whole book has been written about how to do this, actually. That’s where the action is, in terms of functional and taxonomic diversity. Pretty much wherever you go on the entire planet, the most common thing that you’ll find in the litter is Pheidole. They’re cosmopolitan, if not sophisticated. If the importance of a taxon is measured by its diversity, abundance and distribution, then Pheidole are the most important ants. (I guess you could argue for carpenter ants, too. But why? They’re so boring.)

Wilson has argued time and time again that ants are really important, they rule the world, they have the same biomass as people, and all that stuff. So, since Pheidole are the ants that rule among the ants, then we’ve got to really have figured out these ants, right? After all, they’re easy to find, they show up at baits, they’re easy to work with.

So what can we, as the community of ant biologists, tell you about the natural history, life history and habits of these Pheidole that live in leaf litter? Here’s a quick list of features:

  • _
  • _
  • _

That’s only a slight underexaggeration.  Okay, so, I can at least tell you what they eat.

No, I can’t.

Actually, I can.  Why? Because E.O. goddamn Wilson, at 79 years of age, after reaching the pinnacle of his career twenty different times and receiving every honor you could invent, decided to do the little experiment to figure this out. He wrote it up as a sole authored paper in a specialized journal.

It turns out they love oribatid mites. Now you know.

(This is not insignificant, actually, for the field of chemical ecology. Two years after the Wilson paper, Ralph Saporito sorted out that mite alkaloids end up in ants, which end up in poison frogs as their chemical defenses. The frogs also eat the mites directly, too.)

Wilson had spent decades slowly churning on the revision of Pheidole. After spending all that time at the scope and in the museum sorting out the genus, he can’t be blamed for thinking, “what do we know about these ladies after all?” Instead of just wondering, he did the experiment. You gotta love that spirit.

It’s rare for a midcareer PI of a typical lab to do a little experiment of one’s own like this and take the time to write it up. And then there’s EO Wilson doing his own experiments, among a string of high-profile papers, books, gala appearances and being a reliable stand-up mentor to junior colleagues. This communicates an unabashed love for these ants, for discovery, for natural history, and for answering unanswered questions wherever they lead you. Wilson is the consummate tinkerer.

This paper is by no means an outlier. Studies like these pepper his CV, sandwiched with his major theories and findings. To me, these are the actual meat of the sandwich. (Or tofu or something. I don’t eat meat.) To those of us who study ants, that’s what makes Wilson a rockstar. He’d be super-awesome without any of the books and big theories formulated by collaborations with mathematicians. His productivity, keen sense of natural history, an eye for observation and an interest in discovering questions as well as answers has been a trademark of his ant-centered work. The man loves ants, and it shows.

When this paper had come out, I had been working on the ecology of litter-nesting ants in tropical rainforests for about ten years. There were many ideas that I was pursuing, and I’m proud of what I’ve done and excited about what lies ahead. This has been rewarding because so little is known about the biology of these animals, despite their abundance and diversity.

After ten years, if you had asked me, so what do they eat? I wouldn’t have been able to tell you. How many zoologists do you know who can’t tell you the diet of their study organism?

Isn’t that odd that I didn’t know what these ants eat? That nobody knew, at all? Hell yes, it’s odd. Wilson saw it was odd. And he did something about it. The publication of this paper was but a speck, if a speck at all, on the face of his career. For those of us who study litter ants, this was very important. Any one of us could have done it. But you know what? We didn’t, while Wilson did.

That’s what badass science looks like, in my book. And it doesn’t require partial differential equations.

Footnote: You might be wondering, by the way, how can you not know what they eat if you work with them all the time? The answer is, essentially, that these are really small ants. A massive colony fits in a microcentrifuge tube, and a smallish one can fit in a 2 cm piece of straw. You won’t see what’s between their mandibles in the wild, and can’t make out the refuse in nests, either.

19 thoughts on “My all-time favorite scientific paper

  1. Terry – You have officially inspired me to love my career even more than I did 20 minutes ago. Thanks for putting it back into perspective. Cheers!

  2. A favorite of mine, too, although I’m more partial to the full natural history masterpieces of Tschinkel, Roces, Hoelldobler, and Deyrup. Anyhow, this post is actually quite relevant to all of the chest thumping going on over at Dynamic Ecology about Wilson’s editorial. I suspect that Wilson’s deep reverence for natural history may be a the heart of his piece and he is hoping to inspire folks who might be more inclined in that direction to stay in science. I’m quantitatively oriented and so I appreciate Jeremy’s arguments and all the others, too, but they miss a crucial point that so many others miss, too: what do quantitative skills provide if you don’t even have the faintest idea about what you are looking at? A good example may be something like a process like decomposition. We know what goes in, we can measure output (decomp. rate) but we can’t say anything about the process. So we see that decomp is highly variable and we can model it to our hearts content and describe it mathematically. But really, we still would have no idea about what is really going on. This is the fallacy of modern ecology: that we have any idea what we are talking about most of the time. We are too simplistic and too short on descriptions of species and natural processes to “skip a grade” to the mathematical descriptions of things we don’t know much about. Quantitatively oriented folks seem less likely to focus on descriptive works that are so sorely needed. This may be Wilson’s hope here – encourage the people with a keen eye for description to stay in science.

    • Amen! This one is special to me in particular because it’s with my specific community, and because of the context in which Wilson did it all. I’ve also regularly said that another one of my major inspirations has been Tschinkel’s Sociometry paper, also in Ins Soc. That’s why I’ve never said no to Ins Soc, as it’s been so good to me. (Also, they know not to hit me up too often.)

  3. Awesome post. But I’m offended that you think Campnotus are boring!!!!!!!!

  4. Some years ago I met a person who had been a grad student of Robert MacArthur. I was surprised, because the person is a natural historian. He said that MacArthur had been bringing in students with mathematical skills and trying to teach them natural history. That didn’t work, so he tried bringing in students with natural history skills and trying to teach them math. That didn’t work either. If I was good at math I could easily have been pulled into the fray on mathematical models of kin vs. group selection. Instead, I produced a natural history model. In my mind, the next challenge is to see if the natural history model can be captured in a new mathematical model and tested theoretically. The mathematical model could generate “what if?” questions that then could be tested empirically. Progress would be made by an alternation of studies done by scientists that fall into the categories of MacArthur’s two kinds of students. I can’t do the mathematical modeling, so I hope someone will. I would look forward to new natural history studies to test the “what if?” questions that emerge.

    • This is a great story, thanks for visiting Jim. That’s true, all that you’ve figured out (and are still trying to figure out) has emerged from the fact that you’ve spent the time thinking about the natural history and the phenomena. This embodies a lot of what Wilson wrote in the WSJ, and given the arc of you work, is a better example. As Andrew Thaler wrote there’s a difference between “good at math” and “literate at math.”

  5. A nice read, Terry. But I must come to the defense of those boring carpenter ants.
    Seems to me “carpenter ants” wouldn’t be at all boring, if we knew more about them. I know, most people (including a few non-myrmecologist entomologists I’ve met!) think they just eat wood and make our houses collapse, but there’s a lot more to Camponotus than the carpenters. They’re not xylophages, but what in fact do they eat? Why are their eggs yellow to bright orange, and why are subgenus (Colobopsis) eggs so big and hotdog-shaped? How are their worker subcastes determined, and do they differ functionally? And, how did there come to be so darn many species of Camponotus, nearly as many as Pheidole?

    PS – Is Polyrhachis actually a part of Camponotus?

    • I was just trying to think of something more boring than Pheidole. But you’re right, for the same reasons that Pheidole are mysterious, Camponotus are even more so. They’re so similar in phenotype yet so diverse. And since their nests are hard to access, whole colonies aren’t collected much in the field. Yeah, what do they eat? Where do they get their N? The caste thing in Camponotus is messier than Pheidole, which is (kind of) worked out. I think the richness answer will probably end up coming from Corrie Moreau, who has made a great start
      As for Polyrhachis, well, I’m better at climbing trees at building them. And I’m not good at climbing trees.

  6. Reading this got me thinking back on my own old post on the papers that most influenced me (not quite the same as my favorite paper, admittedly). Let’s just say Terry and I came to ecology from rather different directions. ;-)

    A related genre is authors giving insights into their own papers. Your papers often feel very different to you than they do to others. For instance, the paper of mine that I’m proudest of is far from my most cited:

    And my fellow blogger Meg has great stories about an early paper of hers that almost caused her to quit science, and a later paper of hers that was rejected by Ecology before going on to win the Mercer Award:

  7. I find this so inspirational. I didn’t know about the paper, thank you for posting it! What you said really is clear – Wilson does *not* follow the archetype of someone who gets a public image and stops doing science. It was also the feeling I got out of reading his Naturalist, and I think it’s a super super important part of what great scientists do – really love their field, ceaselessly, from top to bottom. Anyone who’s read Feynman’s books gets the same impression – ceaseless, unqualified love for science. Seems like a nice way to live, to me.

  8. This is going in my ‘having a horrible day academically :. needing a pick-me-up’ folder. So wonderful to see the love of a system shining through so clearly in the actions of such a well-respected scientist. Thanks for writing this.

Leave a Reply