Posted by: Jeremy Fox | October 24, 2011

Poll: do you still believe in the IDH zombie? (UPDATED)

My post on how the intermediate disturbance hypothesis is a zombie idea has garnered more pageviews and generated more discussion than anything else I’ve written for this blog. But of course, blog comments, and the emails I’ve received, are a highly non-random sample of all reader reaction. I’m unsure if the post actually changed many minds. So I’m taking a little poll in an attempt to find out.

If you read my original post, please answer the question below. This is just for fun–but it’s not much fun if only a few people respond. 😉

UPDATE (OCT. 26): I slayed a few zombies! Hooray! There have been 27 votes so far. Of those 8 were either unfamiliar with the IDH before reading my post, or already saw it as a zombie. Of the remaining 19, 4 used to believe the IDH but now see it as a zombie, and 12 used to believe the IDH but now aren’t so sure. I consider 16/19 minds changed, or at least shaken up, to be a pretty good batting average. If we make the reasonable assumption that the 8 folks who were unfamiliar with the IDH before, or already saw it as a zombie, now see it as a zombie, we can infer that 24/27 voters now see the IDH as a zombie, or think it might be one.

I know it’s a small, unscientific sample, but it’s still put me in a good mood, even to the point of delusions of grandeur: On this evidence, if every ecologist reads my post, committed zombies will be reduced to a very small minority!

The poll remains open, so keep those votes coming!

UPDATE #2 (NOV. 1): Ok, doesn’t look like we’ve got any more votes coming in, so I’m declaring the quasi-final results. Out of 42 votes cast, 9 already thought the IDH was a zombie and 5 weren’t previously familiar with it. Of the other 28, I convinced the vast majority to completely change their minds (5), or at least created significant doubt (18). And of the remaining 5, who still believe in the IDH zombie, one indicated in the comments that he only voted that way because he read the poll in a different way than I intended, and would probably better be counted as having changed his mind. I’m pretty pleased with that, although I hope those I’ve pushed onto the fence will continue to ask me questions and so give me the chance to continue pushing them onto the zombie-free side. And I hope the small minority who still believe zombie ideas about the IDH start feeling some peer pressure (seriously).

Advertisements

Responses

  1. I voted for the 3rd option, if only because even as an ecology student, your explanation of why the IDH is a zombie flew mostly over my head (maybe it’s time to re-read it)…

    It’s been explained as a relevant theory in now three of my undergrad courses…

    • Sorry the explanation went over your head–I was writing to change the minds of folks who were already familiar with these zombie ideas, rather than writing for students learning about the effects of disturbance for the first time. That said, I hope the posts have been helpful, even if you’re struggling with them a bit. You should know that there are PhD ecologists who are also struggling to wrap their heads around what I wrote, not because it’s “hard” but just because it’s so contrary to everything they (like you) have ever been taught.

      You should tell your profs to read my posts and tell you what they think of them (seriously)! Or just go to your profs and ask questions raised by my posts. If they’re good profs, they won’t be offended–in fact, they’ll be thrilled that you’re taking the initiative to read stuff besides what you’ve been assigned, and that you’re thinking hard and critically about the material they’re teaching rather than just trying to memorize it so you can regurgitate it on an exam.

      One thing to be aware of: the “intermediate disturbance hypothesis” is actually an ill-defined complex of interrelated ideas, some of which are zombies and some of which are not. I think (hope!) I’ve been very specific about which ideas I’m attacking, and why. But it’s possible that your textbook, or your profs, might not be that specific when they teach the IDH, and that could be for good reasons (e.g., lack of time to go into greater detail).

      And at the end of the day, I encourage you to make up your own mind. You shouldn’t just take my word for it that these are zombie ideas, any more than you should take your profs’ word for it that they’re not. You should work hard to understand the competing claims and come to a decision (it sounds like you’re doing that and that’s great).

  2. […] the previous post I asked readers if a post of mine had changed their minds about whether the intermediate […]

  3. I voted “I thought it was a correct, or at least useful, idea before; I still think so” but thought explaining why might be helpful to you.

    I am only generally familiar with the topic, never having read any of the papers you cite, including MacArthur. This means that my vote could very well change if I in fact knew more about the topic. To parse the chosen response out a bit: it’s the “or at least useful” part that I agree with, not the “it was correct” part. Your arguments seem well considered to me, and especially compelling is the one regarding the need to standardize/normalize any comparisons by keeping the long term, averaged mortality rates of control and “treatment” as equal as possible.

    I think the “at least useful” phrase basically encompasses your points related to the fact that IDH is in fact sometimes a valid explanation when you consider nonlinear effects. And also the fact that the time scales required for competitive exclusion’s effects, and hence species diversity fluxes, can be quite long compared to human time scales (especially in environmentally harsh conditions), and I wonder if that doesn’t create a lot of havoc with empirical analyses (i.e. takes way too long to make the appropriate comparisons with low disturbance rate systems, especially e.g. dry and/or cold dominated ones).

    If that makes any sense or helps at all.

    • Hi Jim,

      That makes sense. But it highlights the sensitivity of polls to question phrasing. I inserted the “at least useful” phrase because at some level all models are false, but I didn’t want people basing their answers on that general point. I wasn’t trying to broaden the question to include nonlinear, nonadditive models under the heading “IDH”; I wanted the question to just focus on classical, zombie ideas about the IDH (and I’m very glad that you do see those classical ideas as the zombies they are!) But I can see how my intent was not entirely clear.

  4. My last comment just raised the question in my own mind: experimentally, how would one if fact keep long term, average, mortality rates approx. the same in a system where mortality events are themselves, a disturbance. Thinking of forests in particular, where each tree death, and especially treefall, is pretty much inseparable from disturbance, as far as I can see. In fact that’s pretty much the basis for the whole field of gap dynamics–the deaths *are* the disturbances.

    • Good question. Not sure if this is the best answer in the specific case of forests, but in general, I think the answer is “You use a mathematical model”. That is, if the experiment you’d like to do is, for whatever reason, impossible, you use a well-validated model of the system to conduct the experiment on your computer. For instance, this is what Jon Levine, Peter Adler, and colleagues have done to estimate the strength of the storage effect in grasslands (e.g., Adler et al. 2006 PNAS). It’s impossible to actually create conditions in the field that prevent the storage effect from operating, and so it’s impossible to actually conduct an ideal experiment estimating the strength of the storage effect. But what you can do is parameterize a mathematical model of the system, and then use that model to conduct physically-impossible experiments in order to answer the question you want to ask.

      • Thanks Jeremy. The problem of course is that all models have to be tested (i.e. validated) eventually, and for that you need experimentation. Models are limited in what they can definitively tell you about reality.

      • “The problem of course is that all models have to be tested (i.e. validated)”

        Yes. The point (and this is what’s clever about the Adler et al. approach) is that the validation can involve different experiments and observations than are required to test the prediction of interest.

        I may do a whole post on this at some point. At lot of what’s gone wrong with empirical research on the IDH is that people focus far too much on “test the prediction” and far too little on “validate the model” (where “validate the model” includes “check whether the model obeys the rules of logic” as well as “check whether the model is a sufficiently-accurate approximation of reality”). Testing predictions can be a great way to do science–but there are circumstances in which, on its own, it’s a terrible way to do science. Think of my deliberately silly example, from the comments on a previous post, of manipulating disturbance in order to test the prediction that “diversity peaks at intermediate levels of disturbance because that’s the way the Flying Spaghetti Monster wants it.”

        IIRC, Tim Wootton has a chapter on some of this in the Bernardo & Resetarits Ecological Experiments book, where he talks about testing assumptions as well as testing predictions, as a way to check whether your model’s predictions are right for the wrong reasons (i.e. your model just “got lucky”).

  5. Excellent points–thanks for your responses. Most definitely a worthy topic for a post IMO. I will try to get and look at the Adler et al., and related, papers.

  6. […] so, you’d better speak up. Because the available evidence says that I’m a pretty good zombie slayer. And I have no plans to stop anytime […]

  7. […] seminar audiences despite my best efforts to anticipate and address them before they arise. And the response to my posts on the zombie ideas behind the IDH can be read as an endorsement of my explanatory […]

  8. […] post is directed at the many readers who are still on the fence about whether the intermediate disturbance hypothesis (IDH) is a zombie […]


Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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 )

Google+ photo

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

Connecting to %s

Categories

%d bloggers like this: