Recently, we invited you to ask us anything. Today’s question comes from Ric Charnov, who asks (paraphrasing): Is modern coexistence theory backed by data?
Jeremy’s answer:
Depends what you mean by “backed.” MCT doesn’t make predictions, it’s not that kind of theory. So it’s not “backed” by data in the sense of “we have a lot of data testing its predictions,” as we do for (say) sex ratio theory. MCT is a partition, a descriptive tool. Like ANOVA. It tells you what the various parts of a whole are, and how big those parts are. But it doesn’t make any predictions about how big the various parts will be in any particular case.
But if by “backed” you just mean “we have many empirical estimates of the quantities MCT defines,” then yes, it’s backed by enough data that you can now do meta-analyses and other data syntheses (e.g., Buche et al. 2022, pointed out by commenter Falko Buschke). Although several recent papers raise concerns about how precise and robust those estimates are (e.g., Terry and Armitage 2024). It turns out to be very difficult to precisely and robustly estimate the quantities that MCT defines.
Which is a bit of problem for MCT, insofar as we want to use as an empirical tool, not just a conceptual tool. To be clear, I think MCT has a lot of value as a conceptual tool, that doesn’t depend at all on our ability to apply it to empirical data. But if our goal is to compile lots of empirical estimates of quantities MCT defines (e.g., “equalizing mechanisms” and “stabilizing mechanisms”), so that we can discover, and then explain, patterns of variation in those estimates, we’re going to be in trouble if any patterns are swamped by various sources of estimation error. Even if you just want to use MCT to do a case study of coexistence in some specific community at some specific location, you’re not going to learn much of anything if you can’t estimate the quantities MCT defines with sufficient precision and robustness.
I feel like there’s growing awareness of this issue among empirical researchers working on MCT. Hopefully that will lead to progress on addressing the issue, for instance through improved study designs or statistical methods.