What should ecologists learn LESS of?

There are lots of things that it would be nice for ecologists to know more of. Natural history. Math. Programming. Statistical techniques. The mathematical foundations of statistics. Philosophy of science. Genetics. Evolution. Other things.

If you’re like me, you probably think ecologists should know more about at least one of those things, and don’t think ecologists should know less of any of them. After all, you often hear people say “Ecologists should know more about X”. But you never hear anyone say “Ecologists should know less about X”. Which is a problem. If you want ecologists to be trained in more of some things than they currently are, without being trained in less of anything else they are currently trained in, then you want the impossible. Well, unless you also think that undergraduate and graduate programs in ecology should last significantly longer than they do!

Don’t misunderstand, it’s fine for people to say what they think ecologists should know more of. That’s an essential part of revising curricula. But the other half–the less fun, but equally necessary, half–is deciding what to drop in order to free up time for the stuff you want to do more of. Anyone who’s taught a class has had the experience of agonizing over not being able to cover lots of fascinating and tremendously important material, because there’s just not enough time. But I think we sometimes forget that time constraints also operate at the level of entire curricula. So it’s fine to say that ecologists should know more of X. But if that’s all you say, well, that’s the curriculum design equivalent of wishing for a pony.*

Of course, when people say “Ecologists should know more of X”, they aren’t necessarily commenting on the design of ecology curricula. In my admittedly anecdotcal experience, sometimes it seems like they’re really saying, “I know a lot about X, and so it really bugs me when people who know less about X make mistakes that could’ve been prevented had they known more about X.” Of course, nobody ever continues, “On the other hand, I know nothing of Y, and so am totally unaware of all the mistakes people make due to their lack of knowledge of Y, and so can’t really judge the relative importance of knowing X vs. knowing Y.” And sometimes what they’re really saying is “I know more about X than the average ecologist, which is good because the optimal amount to know about X is whatever amount I personally happen to know.” And sometimes they’re really saying something else. But for purposes of this post, I want to take statements like “Ecologists should know more about X” at face value, and think about the hard choices of curriculum design that follow from such statements.

After all, the world is changing, technology is changing, etc., so maybe ecology curricula do need to change to keep up (they’ve certainly changed in the past). Maybe we really do all need to know more about X, in which case we need to make some hard choices and figure out how to free up the time for everybody to learn more about X.

So let’s talk about those hard choices. As a conversation starter and mind-focuser, below is a little poll. It asks you to name the one thing you think it’s most important for ecologists to learn more of, and the one thing you think ecologists should learn less of, in order to free up time for them to learn more of whatever it is you think they should learn more of. Both questions are required, so you can’t complete the poll by just wishing for a pony and saying what you think ecologists should learn more of. If you don’t think ecologists need to learn more of anything, there’s an option for that (in which case you’re allowed to say they don’t need to learn less of anything either). And if you think different ecologists need to learn more of different things, or less of different things, you have that option. That’s the option you’d pick if you think ecology should involve lots of collaboration among differently-trained specialists. But reasonable as that last option might well be, I’m hoping you don’t all chicken out and take it. 🙂

Note that you can think of the poll as encompassing undergraduate and graduate training collectively (which is how I think of it), or as focusing on one or the other (e.g., because you think undergraduate curricula are fine but graduate curricula need revamping).

p.s. Before anyone complains about the way the poll is structured: yes, I obviously could’ve structured it differently. But no structure would’ve pleased everyone. I went with this poll because it seemed like a fun conversation starter, which is all it’s meant to be. It’s not a scientific sample from any well-defined population. Also, this poll was easy to write; you get the polls you pay for on this blog. If you don’t like the poll, no worries, just ignore it. You can still comment on what changes you’d like to see to ecology curricula–but no wishing for ponies! 🙂

*Of course, you can also argue that ecologists should learn the same things, but better or differently than they currently do. See for instance Fred Barraquand’s comment on a recent post. That’s an important point, but it’s orthogonal to this post.

28 thoughts on “What should ecologists learn LESS of?

  1. More natural history, less physics. I can’t believe how many ecologists wouldn’t be able to ID their study species in the field. And physics…except for a few subdisciplines (plant physiology? fire behavior?), I don’t think enough ecologists need a background in physics to warrant a strong presence in curricula.

    • Ecology undergraduate curricula ordinarily only require physics because at many universities all science majors are required to take some introductory physics. Is that “physics 101” requirement what you’d suggest cutting to free up time for natural history?

  2. The questions are also biased by the institution in which you work or teach. In my School our ecology-stream students spend a large amount of time on compulsory courses in molecular genetics and biochemistry which are of little relevance or interest to them (I don’t dispute that the basics are essential, only that the depth is excessive). Likewise another colleague elsewhere would bemoan the requirement of all science students having to do Physics 101. We can all point to something ‘our’ students should be doing less of, and it’s usually someone else’s module.

  3. Via Twitter: Ethan White says we need flexible curricula in order to train experts (in whatever they choose to be expert in), and people who can bridge between them to foster collaborations: https://twitter.com/ethanwhite/status/519830985173508096

    I’ll note that this actually is the model for many ecology graduate programs, which often either don’t require any coursework at all, or require coursework but leave it entirely up to the student what courses to take.

  4. Felt my answers were a bit (partially deliberately) odd so thought I’d explain my rationale.

    I ignored the pragmatic, sensible and true answer of “it depends”, because you don’t learn much from the poll with that answer.

    As mentioned above, course requirements differ. In the UK we have no physics 101 type courses. So we don’t need to learn less of that.

    I have gone for learning a *little bit* more of something for a *little bit* less of something. While I think mathematical background to stats is very important, you need quite a lot of content for it to become usefully applicable.

    So I have put extra programming because a little bit of programming (5 hours extra per year?) would make a big difference.

    I have said reduce evolution. I am very interested in evolution, loved learning it at university and basically wanted to do it as a career. However the way it was taught to me was “Here is a bit of a theory/hypothesis about this interesting problem” x 10. One little bit of knowledge per evolutionary problem. I could have skipped most of those and be no lesser a scientist, and if I then wanted to work in that area I would learn that specific content.

    • Good to have some British perspective here, many North Americans are unaware that British undergrads specialize earlier and to a greater extent than North Americans. Helps illustrate just how wide a range of curricula can work.

    • I (also UK) answered exactly as you did Tim, but I didn’t have my reasons so well thought out until I read your comment. You’re right, there is so little programming taught that a tiny bit would be a big improvement. On the other hand, my real position is that basic programming/coding/scripting should have already been taught in (high) school – so what to learn less of in school?

    • There’s a north (Scotland) / south (England & Wales) division in the UK (not sure where Northern Ireland fits though).

      I didn’t specialise in Zoology until 3rd year in a 4 year Scottish BSc course, taking Biology, Chemistry & Psychology in 1st year, then more Zoology oriented modules (with some flexibility) + Psychology in 2nd year. The exposure to Psychology classes in particular gave me an excellent grounding in experimental design & statistics, that’s stayed with me throughout my career. There’s great flexibility in 1st year choices for a 4 year degree – I could have selected 2 “science” courses and 1 non-science course if I’d desired.

      I’m now discovering that trying to squeeze all that extra experience into a 3 year very specialised Biosciences programme in a Welsh university isn’t very easy! The 4 year course, is, in my opinion, better. Maybe it’s a bit closer to the liberal arts experience in N America?

  5. I like the fact that students have the ability to pick the course that interests them the most, so some institutions have a bias more towards one aspect e.g. natural history id skills, and other towards another aspect e.g. more theoretical basis. You wouldn’t want all courses to be the same! Though there are obviously a few key components, which is how I answered the survey. i.e. you can’t be an ecologist without ‘x’ and ‘y’ (e.g. fundamentals of ecological theory which does include a lot of math). I find the natural history debate interesting, being an ecologist without this background (call me a rapid learner!) – I think I’m a pretty reasonable ecologist without this knowledge, but instead I have a strong stats background. I know some excellent ecologists who lack statistical skills but are fantastic at natural history & excellent ecologists. So there is a balance, and a diversity of ecologists that makes working in this area great. The courses on the degree program I teach on I think have a nice balance between statistics and natural history – so students learn both, but no course is perfect and there are always areas that miss out to allow for this balance.
    p.s. I am UK based so the degrees are a little different and tend to specialize more early on e.g. on our marine biology/ecology degrees there is no requirement to do physics or chemistry other than that associated with physical and chemical oceanography

  6. I am not sure we need less of something but better of something. Many departments remain in the dark ages in how they delivery their math, natural history and statistics training to biologists. Improving on what is there may be more productive than cutting. Generic 1st year modules within biology, possibly chemistry has to occur but I would agree that for ecologists physics is probably time wasted. Its always going to be the case that Physicists switch to ecology not the other way…. I think :0)

  7. Fun post, I like the way you call attention to the problem that simply saying ecologists need to learn ‘more of X’ isn’t moving us forward, and I completely agree with your point. I must disagree that the answer requires us to consider what ecologists need to learn “less of”. The material on your list is not orthogonal — if it were, there’d be no need to learn it in the first place.

    If the study of evolution didn’t involve any mathematics we wouldn’t be so worried about teaching it. Unfortunately, I think too often students are either being exposed to the wrong mathematical subjects or being exposed to courses in other departments that aren’t transferring efficiently to their study of ecology. Similarly, we sometimes teach ecological subjects as if they are more orthogonal to these skills than they really are. Addressing these issues would be more productive than swapping one subject for another.

    Physicists aren’t worried about how they are going to fit in thermodynamics and quantum mechanics when physicists also need to know calculus, differential equations, probability and programming. Nor is it simply assumed that such skills will be pre-requisites provided by other departments. Of course I’m not saying anything new here, and two of my early mentors articulated this much better in an article in 2004: http://doi.org/10.1126/science.1095480

    • Good comment Carl. Your point here is why many biology departments teach their own “biostats” courses rather than sending their students to take stats courses from depts. of mathematics and statistics. And more broadly, yes, one way to teach quantitative techniques and programming to ecologists is to work them into ecology courses. We certainly do some of that at Calgary, and we’re far from alone I’m sure. But there can be practical obstacles to this approach. For instance, our students here at Calgary complain that they don’t see the relevance of first-year calculus to their biology degrees–until late in their degree programs when some of them hit upper level courses in mathematical modeling and population ecology. One solution would be to stop requiring them to take calculus from the math dept. and instead create a new course that teaches them the math they need in the context of biology–something like Ellner & Guckenheimer’s book, maybe. But we don’t have the staff in our dept. to start teaching a new intro course to hundreds of students. Don’t get me wrong, you’ve made a great point–but getting to the world you envision would require many places to do a top-to-bottom curriculum redesign and change the content of a lot of their courses. Which maybe just means people should bite the bullet and do it.

      • Yup, I agree that this is not an easy fix and requires a complete curriculum redesign. That’s the same conclusion Bialek and Botstein come to in the 2004 paper, and it’s precisely what they did: http://www.princeton.edu/integratedscience/ Alas I was already in the standard physics pipeline by then and it wasn’t until UC Davis Pop Bio core that I got classical ecology pounded into me. I think the Pop Bio core does a nice job of covering the classical and trending literature while letting the mathematics, statistics, and programming be as much a part of the curriculum as they are of the research itself.

  8. These things aren’t all equal in the amount of training we get now, either. You probably can’t learn to program well in the time you would gain by scrapping the little bit of philosophy of science that we do get.

    So one of the hard things is not just a topic-for-topic swap (hard enough!) but figuring out equivalents of class and lab time that these deserve in a curriculum.

  9. At the graduate level, we should trust students to make the choices appropriate to their preferences and career goals. It would be useful for graduate programs and advisors to have explicit discussions of these trade-offs, though. Of course, there are limits to the courses that a program can offer, but laying out these trade-offs enables students to decide where to shore up their skills elsewhere if needed.

    I disagree somewhat with Carl. There is probably something to be gained by more integrative design, but it also comes at the expense of flexibility for students putting together a curriculum for their own needs. Again, this is more the case at the graduate level than undergraduate.

    • I agree with Noam. Creating integrative courses should not be confused with creating a single integrative course or sequence that everyone has to take. Flexibility in curriculum and the ability to specialize is very valuable — some students will take more courses in evolution, some more courses in community ecology, but both sequences can integrate the teaching of math, statistics, and programming that are integral to their respective research areas.

      We don’t need to invent integrative courses, we just need to teach each subject of in a way that also seeks to teach the skills those subjects require, rather than assuming students can just acquire those skills elsewhere ahead of time.

      • I’m not sure it’s so simple to match a set of quantitative skills to a field, especially when one is considering training for paths outside academic research. For instance, natural history and statistics are fairly orthogonal, and the weight one puts on each would be different if one were aiming for a career as an academic researcher, a conservation area manager, or staff at a consulting firm.

        Both students and professors probably want to look at how their training maps in a space of ecological subfields, taxonomic/geographic expertise, and skill sets (from quantitative to policy/management).

      • Hi Noam, again I completely agree with you. I’m not suggesting that we try to integrate non-orthogonal fields. I am merely observing that the reason we have this collection of things in the first place is that they are _not_ all orthogonal. Nor am I suggesting that this is just about integrating *quantitative* methods. Integration with respect to non-quantitative things matter too: Recent discussions of sexual harassment in field research might be most effective if addressed in field-based courses rather than relegated to the silo of ethics. Natural history ought to be integrated into population dynamics. That is not to say that all students would take such a course to learn natural history — another student may learn all their natural history through field-based courses. Writing, reading, presenting, problem solving, & teamwork are likewise skills that are integral to many but not all subjects and ought to be incorporated into such courses. Teaching should reflect the needs of the subject, not outsource them.

    Having started to get fascinated with computer programming since I was a kid, and continuing programming for many many years after that I realized early on that it is not only important to learn a lot of things, but it is equally important to be able to forget some things quickly to make room for other new stuff. I found that to become great at computer programming it is best if you ignore a lot of the clutter and focus only on the pieces that would propel you forward. So, I completely agree with you about the need to “learn less” about things as a necessity of becoming competent in something.

    This said, I would advocate a different model all together. I do understand the presence of a credit hours system and the following of a “fill the bucket” model in which the bucket (our brain) has a fixed capacity and you cannot take more than such capacity. Despite this model being fine and useful, yet it misses a lot of what happens in real life:
    1- It misses the interactions between various subjects and the synergy that can result from such interactions.
    2- It also misses the fact that our brains are expandable in terms of how much knowledge you can fit into them, if you use the right methods for that.
    3- It also misses the ability of the brain to perform better and better with practice, just as athletes keep breaking world records one time after the other.

    I believe that by linking different subjects with one another early on and using a holistic approach there would be literally no limit to the capacity of the human brain. I understand that this contradicts with the current (undergraduate) educational system where students are presented, especially at first, with distinct subjects that they study in relative isolation from one another. Yet perhaps this might inspire a new form of looking at syllabi all together. I do realize, however, that such interdisciplinary approach for undergraduate studies is decades far away, if ever it would be implemented, and that it is not without its drawbacks too.

    I’d love to hear current criticism against this thought though, as I am aware such view of mine is probably pretty romantic or simplistic.

  11. You forgot two options!

    More: all of the above
    Less: sleep

    (or the opposite, of course).

    A rudimentary knowledge of all should be required if only to understand others’ works (a true liberal arts graduate I am, I guess).

    I say more natural history for a number of reasons:

    1) Natural history forms the core of ecology. In order to evaluate others’ work (as a reviewer or citing it yourself), it is integral that you know basic natural history of many organisms. You can’t question the applicability of another system to yours (and thus, most theory) without a familiarity with all systems in question. You can, of course, look this sort of information up, but as EVERY ecology paper has some organisms in it (and natural history allows you to examine the true applicability of a theoretical model, too), this is more basic than any other subject to understanding the paper

    2) It is a subject that is far easier to learn in person from experienced naturalists. “Abstract” (e.g. not necessarily hands-on) disciplines (stats, physics, math, evolution, economics, etc.) can be learned in books. Natural history is necessarily hands-on. I’d argue that besides pedagogy, it is the most hands-on discipline in your list. Explaining in writing how to tell an herbaceous perennial from an annual in a book is hard (look for leaves of last year? check out the roots?); it is rather simple to show the differences in the field. And noticing things like signs of an animal are even harder to explain.

    3) It is not going to massively change on you. In contrast, the genetic techniques of 20 years ago are rather different than today, as are the statistical ones, and even the programming (though the logic behind various languages is pretty similar it seems in my experience).

    For reason #3, I’d say we teach less molecular biology/genetics/statistics. When you need a technique or very specific knowledge, you can learn it on your own (you should have the basic knowledge to build upon) – chances are that it will change in the near future anyway.

    So maybe my answer for what to do less would have to be sleep. 🙂

    • Ok, I’ll bite: natural history is not “the” core of ecology (I don’t think there *is* a “core” in that sense…) And not EVERY paper has natural history in it–in fact, there are whole journals of ecology papers that mostly omit natural history. And no, you don’t always need to know natural history to determine the applicability of a theoretical model. See for instance this:


      or this:


      or this:


      • I’ll – respectfully – disagree, sort-of.

        Fisher’s passage is right, certainly. But its tangential to my point. To be an ecologist, it seems to me, you need to be able to critically examine the literature – 99% of which deals with actual organisms.

        As for the second link, I agree with that, too. But in your quote it says: “Quantitative ecologists are only loosely anchored by the natural history of particular systems. Even the word “systems” is a giveaway; we see organisms as realizations of ideas, not as furry, feathery, or green individuals…. In either case, we are primarily interested in how we can use organisms to understand general principles rather than in the particular organisms themselves. This flexibility lets us pursue interesting questions wherever they lead.”

        In order to understand how to use those organisms, a certain degree of knowledge is needed. In order to use other people’s data, you need to have a general idea of the organisms in order to not abuse said data!

        And for the third link. I read that awhile back and I do think my position falls somewhere between yours and the real alarmists. But I don’t agree completely with this implicit (explicit in your piece) argument that not knowing natural history allows new ideas. It is a different viewpoint, but close observation and knowledge of the natural history allows just as many new ideas.

        And as a point-person among the graduate students here (Davis) for identification and natural history questions, it seems that most students are eager to know more than they do. I went to a rather elite college with a great reputation, which has produced quite a few ecology grad students and professors, and we simply didn’t have more than a few basic natural history courses – the opportunity was not there and really isn’t except at a few schools (Cornell, Harvard, etc.).

      • Sorry, still going to respectfully disagree. Different strokes for different folks. I think the optimal strategy for ecology as a whole is what game theorists call a mixed strategy.

        I’m a crap natural historian. But yet, while freely admitting that I don’t read a random sample of “ecology” papers (no one does), I’m pretty confident that I can read and critically examine much more than 1% of the ecology literature. With respect, I suspect you’re overgeneralizing from your own experience. I think you’re overestimating both the fraction of the literature that can only be understood by people with substantial natural history expertise on the organisms concerned, and the amount of natural history expertise needed to understand and evaluate a typical ecology paper.

        I’m sure the students at Davis are eager to learn more natural history, which is great. And since they all come to you for help, I’m sure from your point of view it seems like they’re underserved on the natural history front. Indeed, they might well be. But if you went and talked to, say, the “stats gurus” or the “programming gurus” at Davis (and every biological dept. has people whom everybody tends to flock to for advice on those topics), I bet they’d tell you that the Davis students are underserved when it comes to stats and programming too. And again, they might well be.

        With respect, you slightly misread that old natural history piece. I didn’t mean to say or imply that new ideas only come from other sources and do not come from natural history.

        If by the “core” of ecology you mean “things every ecologist *has* to know in order to be a competent ecologist, not matter what their specialization or subfield”, then I think the core is comprised of very basic stuff that every ecologist *does* know and that is at absolutely no risk of being lost because of changes in ecology curricula. Put another way, I don’t think it’s plausible to define the “core” of ecology in such a way as to imply that some ecologists don’t know it.

        Sorry for pushing back harder than you probably expected. So let me just emphasize that natural history–like lots of other things–can be hugely valuable to individual ecologists, and to ecology as a whole, and that ecology certainly would be much worse if no ecologists knew any any natural history. But I think that’s importantly different than saying that natural history–or *any other topic or subject*–is the “core” of ecology. I don’t think anyone needs to try to crown their own favorite topic as the “core” of ecology in order to argue cogently that more ecologists should learn more about it. I think good ecology can spring from different foundations, different cores, different motivations. I don’t think ecology has “a” core.

  12. Last bit, I promise! Fisher’s own sex ratio “model”, Elton’s original complexity = stability “model” (that book contains a lot of careful observational natural history) were verbal arguments: who needs math?!? teach less of that (kidding). Hamilton, Charnov and Godfrey’s later mathematical sex ratio models were especially informed by natural history (Hamilton clearly knew the life-history of his wasps – and thrips! – quite intimately; I don’t think one would argue that paper would have ever come about without intense natural history study). And of course, these are the theoretical papers I quite adore.

    So yes, that’s my experience, and the “core” of any matter is a subjective matter – I dare you to advocate a different “core” :).

    But the question was: What should we teach more of? What should we teach less of?

    I think my second two points about why natural history should be taught more stand. And I suspect that you don’t feel that natural history should be taught less, I think you just took umbrage with a particular term (its not the “core” for you). In that vein, I don’t think pedagogy forms a huge part of ecology, but I think it should be taught more (for mostly pragmatic reasons).

    Of course, ability to learn on your own doesn’t really solve the time budget problem, and I’ve had some fantastic math and stats teachers that taught me things that would have taken far longer to learn on my own.

    I guess to actually figure this out, we use a weight coefficient (0-1) for value to one’s research multiplied by difference in time needed to learn the same subject in a class or on one’s own and compare that to every other subject across all ecologists (or at least a random sample of sufficient size). Wait… that doesn’t fit with my argument.

    Instead we can observe the ecologist in his/her natural environment and observe their success (remaining, of course, at a safe distance as to not interfere with their behavior) correlating it with the courses we observed them in to create a series of models, select the best-fitting one, draw from the posterior prediction distribution for each parameters to estimate of the effect of each course on their success.

    I’ll probably need another stats and programming class for that…

  13. In my experience, I would’ve liked to have more math introduced to me earlier on – in other words, I think most programs should offer a Mathematical biology/ecology course*, which could be combined with (or precede) a course in theoretical ecology. Taking a course with Stephen Ellner and John Gukenheimer not only got me interested in theory, but also had a big effect on how I thought about ecological problems. For example, the inverse of mortality (or emigration, or both) – admittedly with big assumptions – can give you an idea of the residence time for individuals in the population. I think the other major benefit of such a course to students is that it would open the door to much/most of the literature in theoretical ecology.

    For me, I could’ve easily forgone a 2nd semester of organic chemistry or physics (even though I enjoyed both) in favor of a MathBio course. However, I think I know too little of what courses ecology students have taken on average (both as undergrads and graduates). That, I think, would be a very interesting poll for this wonderful blog – for example, the number of courses in math, chem, stats, genetics, nat. history, etc. taken – split among the undergraduate and graduate degrees. Making a note of the year that the respondents completed their degree might be useful too.

    *-note that by Mathematical Ecology I do not include statistics – which I think is important, but perhaps should be its own thing.

  14. Pingback: Don’t be so quick to recommend “best practices” in science or academia | Dynamic Ecology

Leave a Comment

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 )

Facebook photo

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

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.