So my flight out was delayed sufficiently that I’d have missed my connecting flight, leaving me with no choice but to fly out tomorrow. So I returned to the meeting from the airport, just in time to catch a few talks.
Peter Adler showed that, if you want to infer plant demography from oft-measured plant traits like wood density and specific leaf area, the glass is half-full, or half-empty depending on your point of view. Interspecific variation in some traits (but not SLA) is related to interspecific variation in demographic rates. But there’s sufficient variation around the line that if your goal was to, say, predict the demographic rates of some particular species for which you only had trait data, the prediction would be so imprecise as to be useless for most purposes. Next big question: what about density-dependence of demographic rates, and can that be related to traits? That might be an avenue to start linking trait data to modern ideas about species coexistence (a topic I plan to continue harping on).
Robin Snyder used theoretically-motivated analyses of a long-term dataset to show that grassland plants living in temporally-variable environments actually are selected to exhibit increased temporal variability in fecundity–that is, selected to become temporal niche specialists. This is opposite to the conventional wisdom in life history evolution theory, and reflects the nonlinear dependence of demographic rates on environmental conditions and species’ densities.
Last but far from least, Peter Chesson (who I was glad to see drew a crowd even late Friday morning) pointed out the major limitation of modern coexistence theory: it assumes stationary environments, environments which may fluctuate, but with constant mean, variance, and other statistical moments. Real world environments are not stationary, and not just because of anthropogenic global warming. I’ve had empirically-minded ecologists point this fact out to me, usually as an excuse for dismissing or ignoring existing coexistence theory. For instance, a commenter on this old post argues that because real environments are nonstationary, all dynamics are transients, and so ecologists should feel free to just study dynamics on whatever timescale they feel like studying without worrying about whether those are “long term” or “asymptotic dynamics”. Of course, the problem with dismissing theory on the grounds that it makes unrealistic assumption X is the risk that someone will develop theory relaxing that unrealistic assumption! Which Peter Chesson has just done. It’s important to appreciate that this is not an easy thing to do, conceptually. It’s not even clear what we might mean by “coexistence” in a nonstationary world. Indeed, as Peter pointed out, in a nonstationary world a species doesn’t even have “an” environment, fluctuating or otherwise–it’s past, present, and future environments can all be profoundly different in terms of the frequency with which any given environmental state or event occurs. Peter showed that one can divide the environment into a stationary component and a non-stationary component. Then he identified the conditions under which nonstationarity all cancels out and so doesn’t change the predictions of stationary theory (basically, when species are at very high total density, and they all respond the nonstationary component of environmental change/fluctuations in the same way). Then he used that special case as a baseline, to look at how lower total densities and different responses to the non-stationary component of the environment can either reduce coexistence (i.e. cause exclusion that would not occur in a stationary environment), or strengthen it. I did find myself wishing his introduction had been shorter, leaving more time to dig in to the key results, but that’s just being greedy. It was nice to end the meeting with a talk by one of ecology’s best minds, tackling a really difficult and important conceptual problem head on. Must do a post sometime about the importance of tackling hard problems instead of just looking for “low hanging fruit”…
It was a great meeting for me and my students, lots of really useful interactions in addition to all the good talks. Looking forward to Minneapolis next year.
Your account of Chesson’s response to conceptual challenges, i.e., non-stationary environments challenge elements of coexistence theory – nicely dovetails with how Lisa Randall responded to a comment on Science Friday today – she said science always leaves useful uncertainty at the edges of our theories, but they remain valid within their scope. Sounds like the post you did on scientific language a while back.