My science is better than your science: clashing paradigms in ecology

By Joern Fischer

I recently had a very inspiring discussion with two ecologists who are trained in botany and plant ecology. One of them in particular challenged the survey methods and experimental design we use in our study of farming landscapes in southern Transylvania. To my mind, this discussion was of excellent value because it led to a depth of understanding that I think we rarely have in science. Or, to be less idealistic about it, this kind of discussion has great potential to lead to deep understanding.

Our method, my colleague argued, was not useful to find rare plants. And how meaningful, really, is a plant survey, if you don’t find rare plants with it? I argued the opposite: only if you study the landscape as a whole, including all the bits with few rare plants (which are more common than the areas with rare species), do you get a true appreciation of the landscape. But — my colleague argued — from a conservation perspective, who cares about common species? And I argued that it’s the common species that account for ecosystem functioning (rather than diversity per se), and major changes in land use would in fact cause declines even in common species — thus threatening the functioning of the system as a whole. Our interim conclusion was that we are dealing with different epistemologies (what counts as valuable knowledge?) and paradigms (how we think about things). I found this discussion inspiring because such discussions are rare. Typically, scientists trusting in different paradigms either avoid each other, or fight each other. Perhaps if the discussion had gone on longer, we might have started to fight, but we didn’t!  (:

To me, a useful question for the future is not “which method is right”, or “which conceptual framework is best” — but rather, which method is useful under which circumstances, with what kind of goal, and in what kinds of systems?

If only we acknowledged this, explicitly, more often, I think we would have fewer senseless debates in academia. Instead of avoiding those with different paradigms, we should find them, and talk to them: it’s only that way that we will expand our own horizons.


5 thoughts on “My science is better than your science: clashing paradigms in ecology

  1. Hi Joern,
    this is certainly a worthwhile discussion. I am aware that your reason for posting this was to discuss the broader issue of personal/academic paradigms and ensuing senseless debates . So I apologise at the outset as I can’t help but comment on the specific example you raised around sampling for infrequent or frequent species (and so perhaps start yet another senseless academic debate!).

    While I think you and I perhaps subscribe to a common paradigm in some aspects of landscape and conservation biology, I do personally think there is value in ensuring our sampling strategies detect infrequent as well as frequent and abundant species. Likewise I think it is important to sample all habitats/land uses including the most frequent and dominant and those that are rare. Our ability then to detect effects of land use change may be greater than if only frequent and abundant species were the focus.

    In any landscape most (>90% perhaps) plant species are uncommon and infrequent (only a tiny fraction of them might be formally listed as rare – I have intentionally avoided the use of “rare” ). Infrequent species may respond differently or be more sensitive to land use change than the frequent and abundant species. Also, while individually they might not contribute much to ecosystem function, they may do so as a whole.

    The work we did on plant responses to soil P is a good example (Australian Journal of Botany 59, 670-681 and Journal of Applied Ecology 45, 1274-1283). Infrequent species were more likely to have negative responses to increasing soil P – in particular the responses of individual species of geophytes (orchids, lilies) were most negative. This group is species rich at landscape and site scales (when present), but individual species are typically infrequent and not abundant when present. However, demonstrating that this group was by far the most sensitive was extremely informative. I think it would also be brave to suggest they don’t (didn’t?) contribute much to ecosystem function – for example many of them are associated with specific pollinators and mycorrhiza. We can only guess at what effect the collapse of geophytes has wrought on the biodiversity and functioning of temperate Australian grassy ecosystems.

    anyway food for thought…


    • Thanks Josh — raises one question though: how much effort should you invest per site (given a trade-off with number of sites versus effort per site), if you “want it all” — including all the infrequent species? This is a real issue in our landscape — do you have specific suggestions how one ought to deal with this?

  2. Hi Joern, simple answer, no.
    Firstly though, my comments related to landscape frequency more so than site abundance. If everything had an equal chance of being abundant when actually present then the easy answer is to just sample as many sites as possible. But of course many infrequent species also have a greater than average chance of low abundance when present.
    Secondly, I don’t think we can ever expect to detect all species no matter how thorough our sampling effort, even if we have the best botanists at our disposal. So trying to sample all infrequent species would be futile – only in very specific cases (and generally not research orientated) might someone “want it all”.
    Where you decide to invest on the site specific effort versus number of sites curve will of course depend on your research question (which I don’t know) and also what you are going to do with the data. It is unfortunately extremely common in plant sciences to use large plots and intensive sampling and then chuck away all of the infrequent species when it comes time to do analyses (because of the “noise” they might add)! A quick inspection of any issue of Journal of Vegetation Science will provide you with an example…So trying to sample all (well, most) species just because it is what everyone does is silly if the methods of analysis won’t even account for all species.
    Perhaps the best (only?) suggestion I have is to developing some prior understanding (or belief) about whether infrequent species differ from frequent ones in either traits, responses (to whatever it is you are looking at) or local abundance. Such hypotheses could help guide the design of a sampling strategy.
    From a practical point of view, if you do want to maximize sampling of infrequent species I can only just suggest doing some pilot sampling with varying effort (plot size and time). However, once you have that data you still need to decide how many species is enough – so we are back at the beginning!
    So Joern I don’t have a clear answer for you. At the very least though I think it is important to be aware that there is a trade-off and in some circumstances this will matter.


  3. Thanks Josh — I think we broadly agree. We conducted a pilot study last year, comparing using several small plots per site with fewer larger plots; and found strong correlations. We did not specifically assess the issue of whether rare species respond “differently” — yes, that would be useful. Again, thanks for your feedback!

  4. Hi Joern

    I tried to make peace with a young postdoc (who’d never left the education system) who refused to accept that Bayesian stats should not be considered the only method for examining everything! Personally I think it has its place…sometimes.. but when it comes to measuring purely quantitative data , to me its of limited use. The really disappointing aspect to the discussion was her refusal to bend. She’d put everything into her position to the point of religious fervour. At first I put it down to a generation Y thing (no offence to gen Y) coupled with a lack of life experience and real world research but came to the conclusion it was something more. What it was I don’t know but I sense she’d been indoctrinated into the Bayesian world probably by a lecturer or supervisor. It was sad.

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 )

Google+ photo

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

Connecting to %s