Posted by: Jeremy Fox | July 26, 2011

An empirical divide in ecology?

Great post over at The EEB and Flow, responding to Lindenmayer and Likens’ recent piece claiming that ecology is losing its empirical, place-based “culture” in favor of modeling, meta-analysis, and data mining. A quick glance at recent issues of our leading journals doesn’t support Lindenmayer and Likens’ argument (most articles are still based on data the authors collected themselves). And increased competition for funding and jobs probably exacerbates intellectual tensions that go all the way back to the founding of ecology as a distinct discipline.

On this blog, I raise a lot of questions about how we do ecology, and I’ve also defended my own way of doing things. But nothing I’ve written is meant to imply that people who approach science differently than I do are somehow doing it wrong. Indeed, many of the papers that most impress me are papers which are very different from anything I would or could ever have done. Ecology is too hard to be “solved” by any one approach, and it’s too bad that some ecologists seem to feel otherwise. Over at The EEB and Flow, Ethan White notes that, when senior ecologists write about the state of the field, they often seem to end up lamenting that not everyone is doing science the way they personally would prefer it to be done. Like Ethan, I tend to find this sort of commentary unfortunate and unhelpful.


  1. Exactly. There’s a really important difference between writing on “Here’s why what I do is really cool and important” (which is great) and writing on “What I do is cool and important and the rest of you are hurting science by wasting valuable time and resources on what you do” (which is not so great).

    • There’s also “Here’s how you could do what you do even better” and “Here’s how we can combine what you do and what I do to learn something neither of us could’ve learned on our own”, both of which are hopefully useful. I should do a blog post about the latter at some point, just need to come up with some good examples of cases where the synthesis of different lines of evidence led to a really satisfying conclusion that couldn’t have been arrived at via any one line of evidence. Any suggested candidates? Local-regional richness relationships might be one. “Evidence for evolution by natural selection” is of course a big, obvious one (maybe too old, and too familiar, to be the best choice for a post…)

      • Yeah, absolutely. I am sadly having a hard time thinking of other good candidates (local-regional is a nice example) in part because I don’t think we tend to do it very well all that often in ecology.

      • More broadly, Phil Warren has done a number of nice “macroecology in microcosms” experiments, at least one of which was published in Oikos. So a post about that work might be a possibility. Indeed, when I interviewed for my postdoc at the CPB I told them that my plan was to do macroecology in microcosms, which I suspect was music to John Lawton’s ears. But by the time I actually took up the position, my interests had shifted, plus I’d realized that Phil was already doing it better than I could.

      • I suppose it depends on exactly how broad a range of evidence you want, how big and general you want the question to be, and exactly what inferences you’re trying to make. Lots of people have, on their own, published studies integrating observational data, experimental data, and theoretical models. That’s actually not at all uncommon–you start with some observed pattern (say, a cline in allele frequency), you do a bit of modeling to develop a hypothesis (e.g., a cline in selection, plus migration), and you do a experiment and conduct additional observations to test that hypothesis (e.g., reciprocal transplants, calculation of Fst from neutral markers). And maybe you follow up with further modeling (e.g., over what range of parameter values do we expect to observe this pattern?), and comparisons with results of similar studies from other systems. Etc.

        Which raises the question, why is this kind of tight integration of different lines of evidence totally routine in some contexts, and unusual or non-existent in other contexts? For instance, local-regional richness studies got there in the end–but it took a while, and along the way I at least had to review lots of papers that seemed to almost willfully ignore any line of evidence besides the one favored by the authors.

        Maybe the difference with local-regional richness studies is that the pattern (linear relationships between local and regional richness) wasn’t so much interesting in and of itself–people mostly weren’t setting out to ask “Why does this pattern exist?” Rather, the local-regional richness pattern was interesting to people because of what it was thought to imply about some other feature of nature (basically, the invasibility of local communities and the strength of competition). And so folks whose preferred line of evidence about invasibility was local-regional richness relationships tended to ignore or downplay other lines of evidence (such as small-scale field experiments directly testing invasibility).

  2. […] up on Dustin’s recent post on bridging ideological divides, and related posts here, here, and here. Wanted to toss out a couple of further thoughts on how to do […]

  3. […] you may have seen earlier either on Jabberwocky, EEB and Flow, or over at Oikos‘ new blog, the most recent piece about how some branch of ecology is ruining ecology has caused some […]

  4. […] on whether ecology is losing its way — other interesting blog posts on this are here and here. I’ll paraphrase what Lindenmayer and Likens said: they’re concerned that ecology is […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s


%d bloggers like this: