The (r)evolution of ecology in the 1950s and ’60s (UPDATED)

Thanks to the influence of Robert MacArthur and other key figures, and external factors like the rise of environmentalism, ecology in the late 1950s and 1960s transitioned from an obscure and mostly descriptive discipline, to a modern science concerned with general principles and hypothesis testing and attracting significant public attention and support. It was recognized at the time that this transition was both needed, and already underway (see the ESA’s 1965 report on the state and future of ecology, which is a fascinating read).

But how did that transition come about, exactly? Was it really a matter of a few revolutionary geniuses like MacArthur coming along and demonstrating a totally new way to do ecology, which replaced the old-fashioned stuff? It sure looks that way if you look at the list of Mercer Award winning papers. Some of the Mercer Award winners from that time are foundational classics of modern ecology, while others are old fashioned papers that are now forgotten. By about 1971, the “competitive replacement” looks complete, and from then on the Mercer award always goes to recognizably modern work.

But if you look more closely within the Mercer Award winning papers, a rather different picture emerges…

(UPDATE: to my embarrassment, everything that follows is just an inferior recapitulation of stuff Mike Kaspari said much better in the ESA Bulletin several years ago. I’d either read Mike’s piece and forgotten it, or missed it, but either way it’s embarrassing. Thanks very much to Mike for sharing his piece in the comments. You should all click through and read it, it’s full of insight.)

Back in the 1940s through the early 60s, the Mercer award often went to site-specific, descriptive studies of plant community composition and succession. For instance, Dyksterhuis 1948, “The vegetation of the Western Cross Timbers”. The apparent exceptions start appearing suddenly in the mid-50s and early 60s, and jump out at a modern reader (at least, they do to me). Odum and Odum 1955! (the Odums! ecosystems! trophic structure! productivity!) MacArthur 1958! (MacArthur! competition! niche overlap! hypothesis testing!) Connell 1961! (Connell! competition! a field experiment!)

Seeing famous classics intermingled with now-forgotten papers makes you wonder about the internal deliberations of the award committee. How could they give the Mercer award to a descriptive study of succession like Olson 1958 one year, and to a pioneering modern study like MacArthur 1958 the very next year? And having given the award to a pioneering modern study, how could they ever go back to giving it to old-fashioned work?

Well, pretty easily, actually. Because when you go back and look at them, those pioneering modern studies don’t look totally modern. Rather, they’re a mix of modern and old-fashioned bits.

For instance, MacArthur 1958 starts by talking about general principles of population regulation and coexistence and how to test them, the first figure is the phase plane diagram of a Lotka-Volterra competition model, and there are a bunch of statistical null hypothesis tests. All very modern. But most of the paper is excruciatingly detailed observational natural history data on warbler behavior. There’s a whole table and associated discussion devoted to the frequency with which two species of warblers hop across branches as opposed to along branches! Most of this stuff is only tangentially connected to how warblers coexist, and it looks just like the descriptive case studies that modern work replaced. If you were being a bit uncharitable, you could argue that MacArthur 1958 is just Olson 1958 with null hypothesis tests and discussion of Lotka and Volterra bolted on.*

Which of course is a bit like saying that the Origin of Species is just natural history with speculations about evolution bolted on. Scientific revolutionaries are almost invariably transitional figures, and you misunderstand them if you see them as totally different from their peers. To pick a far more revolutionary example than MacArthur, consider Isaac Newton, inventor of modern physics and indeed modern science. But he was also really into alchemy. In his own mind, and in the mind of his peers, there was no discontinuity between his physics and his alchemy. It’s only from our contemporary perspective that Newton looks like a modern scientific genius who happened to have a bizarre, unscientific side interest in turning lead into gold. That doesn’t mean our contemporary perspective is wrong–that there’s really no difference between science and alchemy. It just means that scientific revolutions aren’t actually revolutionary. They’re evolutionary. In the early going, they always inherit many of the features of what preceded them. Indeed, insofar as they don’t, it’s difficult for them to succeed at all.

*Same for Odum and Odum 1955. Yes, there’s a lot in it that’s genuinely new–there’s a reason it’s remembered today! But there’s also a lot in it that’s very much of its time. For instance, there’s lots of natural historical description of the atoll with no obvious connection to general principles and hypotheses like pyramids of biomass or whether the reef community is in a stable steady state. Same for Connell 1961, if anything even more so. The transplant experiment along the environmental gradient for which we remember that paper is only a small part of it. Much of the paper is given over to detailed natural historical observations of weekly variation in barnacle settlement and mortality, comparison with past observational studies of where barnacles settle, the movement behavior of whelks, etc. And that stuff isn’t just background–it’s the core of the paper. Connell doesn’t pitch and summarize the paper as a study of general principles of species distributions along environmental gradients, or as an illustration of the power of a novel experimental approach. He presents it as an unusually detailed and accurate record of recruitment and survival under natural conditions. Indeed, having reread Connell 1961, I wonder a little whether, if you reran the tape of life and played it again, Connell 1961 would still end up as a foundational classic. That’s not to criticize Connell 1961 at all. It’s just to say that whether or not a paper ends up becoming a foundational classic is in part a matter of factors and forces outside the paper’s control. See this old post for discussion of how credit for the success of an idea should be apportioned: to the idea itself, or to subsequent work applying and using the idea.