The Art of Modelling poses a question to mathematically-inclined readers: can you build a model of individual movement that explains sheep cyclones?

Even if you’re not a modeller, you should click through to find out what a sheep cyclone is.

Inferring causality is hard. Especially in a world where lots of factors, some of them unknown, causally affect the response variable of interest (and each other), and where there are causal feedbacks (mutual causation) between variables. It’s even harder when, for whatever reason, you can’t do a properly controlled, replicated experiment. What do you do then?

One standard answer is to rely on what Jared Diamond (and probably others) have called “natural experiments”.  The basic idea is as follows. If you think that variation in variable A causes variation in variable B, compare the level of B across systems that vary in their level of A. So instead of manipulating A yourself, you’re relying on the “manipulations” (variations) in the level of A that nature happens to provide.

Unfortunately, natural experiments are infamously unreliable, not just compared to “real” experiments but in an absolute sense. As my PhD supervisor Peter Morin liked to say, “The problem with natural experiments is that there’s no such thing as a natural control.” That is, systems that vary in their level of A often vary in lots of other ways as well, some of which probably also affect the level of B. You can of course try to address this by statistically controlling for the levels of those other variables, assuming you can identify them. And you can try to simply collect lots of data from a large range of systems in the hopes that surely some of the among-system variation in variable A will be independent of all confounding variables. And you can try to get rid of any causal feedbacks from B to A by praying to the god of your choice…

Or maybe there’s a better way. Economists have to deal with all the same challenges in inferring causality that ecologists do. If anything, economists have it even worse because doing relevant experiments often is harder in economics than it is in ecology. In response, economists have come up with an interesting and potentially-powerful approach to inferring causality from natural experiments, the method of “instrumental variables” (IV).

Here’s the basic idea (for details, click the link above, which goes to the very good Wikipedia page on IV). An instrumental variable, call it X, is a variable that causally affects B only via its effect on A, and that is not itself causally affected (directly or indirectly) by B or A. Economists summarize the latter assumption by saying that X is “exogenous”. So you can estimate the causal effect of A on B by using, not just any natural variation in A, but only that natural variation in A that can be attributed to natural variation in X. Changes in X are perturbations that propagate to B via only one causal path, that running from A to B, so variation in the instrumental variable X allows you to estimate that strength of that causal path. The approach can be generalized to multiple causal paths, as long as you have multiple instrumental variables.

One thing I find interesting about IV is that they highlight how “more data” is not always helpful. Tempting as it is to think that, if only you had enough data on A from enough different systems, you could reliably infer the causal effect of A on B, it’s not true. What you need is not more data on the variability of A, you need the right sort of data on the variability of A (namely, that generated by an instrumental variable). Indeed, more of the wrong sort of data on variability in A can actually be harmful to inferring the effect of A on B.

The nice thing about the IV method is that it doesn’t require you to know anything about the rest of the system, such as other variables that might affect B while also covarying with A. All you have to know (and this is the hard part) is that X is what economists call a “good instrument”–that it satisfies the assumptions that make it an instrumental variable.

Which may limit the applicability of IV in ecology. In economics, IV are often policy changes. For instance, an increase in cigarette taxes should affect health only via its effect on how much people smoke. So you can use changes in cigarette taxes to estimate the effect of smoking on health, thereby getting around the fact that lots of factors may affect both health and smoking, and that people’s health may affect their inclination to smoke. Weather events like droughts also tend to make good instruments in economics.

I’m unsure whether ecologists will often have good instruments available to them. Weather is exogenous to ecological systems as well as to economic systems. But the problem is that weather changes typically affect any variable of interest via multiple causal pathways. And many policy changes certainly have ecological as well as economic effects. But the problem with many policy changes affecting ecological variables is that they’re not exogenous–the policy changes are made in response to observed changes in the variable which the policy change is intended to affect. So if ecologists want to use policy changes as instrumental variables, they may want to focus on policies with unintended ecological consequences. And even there you still might have the problem of unintended consequences propagated via multiple causal paths.  But we won’t know if IV can be useful in ecology if we don’t try them out.

And if you do try out IV and get them to work, I hope you’ll submit the paper to Oikos.

Last week was the biggest week ever for the Oikos blog. No surprise, since I did a bunch of posting. But still: 3972 views, including 1124 syndicated views! That’s 567 views/day for those of you scoring at home.

It was also the biggest week ever just counting non-syndicated views (2848), even though our many of our non-syndicated views have been replaced by syndicated views since we started putting full posts rather than teasers in our RSS feed.

It’s a bit of a pain to add up the syndicated views since you have to do it by hand from the stats on individual posts. But assuming that the proportion of syndicated views this week was typical (it’s actually probably a bit higher than usual since we had a bunch of posts this week, but whatever), then in a typical week we’re getting 2800-3400 views, or well over 400/day.

Thanks for reading everybody!

Friend of Oikos Blog Chris Klausmeier (“lowendtheory”) writes with details of a series of one-week summer courses on Enhancing Linkages between Mathematics and Ecology (ELME), to be offered at Kellogg Biological Station (MI, USA). I know all the instructors, they’re all excellent, ranging from the world-famous (Hal Caswell) to the someday-will-be-world-famous (Colin Kremer and Don Schoolmaster). Details below.

***********************************

ELME is a summer educational program at the Kellogg Biological Station devoted to Enhancing Linkages between Mathematics and Ecology.

ELME 2012 will be a sequence of three courses covering: Week 1) Maximum Likelihood Estimation, week 2) Structural Equation Modeling, and week 3) Matrix Population Modeling. In this hands-on environment, students will learn the basics in a lecture setting and cement their knowledge with independent and collaborative modeling projects using the computer program R.

Dates: June 4-22, 2012

Hours: Mon-Fri 9-5

Instructors: Week 1) Colin Kremer (Michigan State University), week 2) Don Schoolmaster (National Wetlands Research Center / USGS), and week 3) Hal Caswell (Woods Hole Oceanographic Institute)

Target audience: 12-18 graduate students or exceptional undergraduates

Prerequisites: At least one semester of statistics, undergraduate calculus, and familiarity with basic matrix manipulations Previous exposure to theoretical ecology and R useful but not required.

Format: A mixture of lecture, guided computer labs, and independent/team projects

To apply, email elme2012@kbs.msu.edu the following:

- your CV

- a statement of research interests and why you’d benefit from the course (< 1 page)

- a statement of relevant educational/research experience, including related coursework (< 1 page)

- the name of a reference who you’ve asked to email a letter of support

Deadline for applications: March 15, 2012

Preference will be given to students interested in all three courses.

Financial support to cover room and board and help defray transportation costs is available. Let us know if this is not necessary.

Academic credit is available, students of MSU and affiliated schools are encouraged to enroll.

For more info see <http://www.kbs.msu.edu/education/elme> or email elme2012@kbs.msu.edu.

Amy Hurford, an ecology graduate student working a recent ecology PhD who worked with the brilliant Troy Day and Peter Taylor at Queen’s University, has a new blog called Just Simple Enough: The Art of Mathematical Modelling. It’s great stuff, you totally need to check it out. She’s thinking out loud, and very articulately, about what makes a great model, why build a model at all, and how coming up with a simple model is often a matter of seeing the problem from the right angle (a topic on which I’ve commented myself).

Seriously, what are you doing still reading this blog? Click the links already!

My northern neighbor Mark Lewis, Canada Research Chair in Mathematical Biology, will be offering a course on “Models in ecology” for advanced undergrads and grad students at Bamfield Marine Station. Marty Krkosek is the co-instructor.

The course runs Apr. 30-May 18, 2012.

You have to apply to be admitted. Application deadline is Mar. 1. To apply, go here.

The course description is below. It sounds awesome. I especially like how the course is suitable for students from both empirical and theoretical backgrounds. And Mark is one of the very best people in the world at linking math and data, as well as a great teacher and a great guy. So if you want to learn how to do ecology the way I for one think it should be done, this is the course for you.

**************

This course develops the methods, models and tools for quantitative ecology. Students learn to formulate, analyse, parameterize, and validate quantitative models for ecological processes and data. Applications include population dynamics, species interactions, movement, and spatial processes. Approaches involve classical hypothesis testing, computer simulation, differential equations, individual-based models, least squares, likelihood, matrix equations, Markov processes, multiple working hypotheses, and stochastic processes. A computer lab covers simulation and programming methods. Course discussion entails evaluation and appraisal of current literature. This course is open to graduate and undergraduate students.

Prerequisites: Introductory calculus, and statistics/biostatistics, or permission of the instructor(s).

This course is suitable both for field-based biology students and for mathematical/theoretical students who are interested in learning about how to connect models to data in an applied ecological setting.

Not “proportion of variance explained”! At least, that’s not the most precise gloss. Nice discussion here.

HT Jarrett Byrnes (via Twitter)

The International Network of Next-Generation Ecologists is surveying ecologists about their knowledge of mathematics and their views on how to incorporate mathematics into the training of ecologists. It’s a short survey (it took me less than a minute), go take it here.

Just make all the usual judgment calls and conduct all the usual “exploratory” analyses that scientists conduct all the time!

The linked paper is the best paper I’ve read in a long time. It’s essential reading for everyone who does science, from undergraduates on up. It’s about experimental psychology, but it applies just as much to ecology, perhaps even more so. It says something I’ve long believed, but says it far better than I ever could have.

One partial solution to the problems identified in this paper is for all of us to adhere a lot more strictly to the rules of good frequentist statistical practice that we all teach, or should teach, our undergraduates. Rules like “decide the experimental design, sampling procedure, and statistical analyses in advance”, “don’t chuck outliers just because they’re ‘outliers’”, “separate exploratory and confirmatory analyses, for instance by dividing the data set in half”, “correct for multiple comparisons”, etc. Those rules exist for a very good reason: to keep us from fooling ourselves. This is not to say that judgment calls can ever be eliminated from statistics–indeed, another one of my favorite statistical papers makes precisely this point. But those judgments need to be grounded in a strong appreciation of the rules of good practice, so that the investigator can decide when or how to violate the rules without compromising the severity of the statistical test.

Basically, what I’m suggesting is that, collectively, our standards about when it’s ok to violate the statistical “rules” may well be far too lax. Of course, if they were less lax, doing science would get a lot harder. Or rather, it would seem to get a lot harder. In fact, doing science that leads to correct, replicable conclusions would remain just as hard as it always has been. It would only seem to get harder because we’d stop taking the easy path of cutting statistical corners. And then justifying the corner cutting by making excuses to ourselves about the messiness of the real world and the impracticality of idealized textbook statistical practice.

The linked paper discusses another solution: to report all judgment calls and exploratory analyses, so that reviewers can evaluate their effects on the conclusions. Sounds like a great idea to me. They also note, correctly, that simply doing Bayesian stats is no solution at all. The paper is emphatically not a demonstration of inherent flaws in frequentist statistics.

Further commentary from Andrew Gelman here.

Here’s an issue which I’ve encountered occasionally as a referee over the years (though not recently, and not as a handling editor as far as I can recall). It concerns manuscripts for which a student is the lead author, and their supervisor is a co-author. Once in a while I find that such a manuscript contains one or more serious mistakes, such as confusion about basic concepts, an experimental design that completely confounds key factors, failure to measure important response variables that obviously should’ve and easily could’ve been measured, or serious statistical errors such as analyzing a nested design as if it were a factorial design. The nature of the errors is such that I would not expect to encounter them in papers lead-authored by the supervisor.

So my assumption (and I emphasize that it is an assumption) is that one of two things is going on.* Either the supervisor didn’t really read the paper carefully before it was submitted, and so wasn’t fully aware of the mistakes or of their seriousness. Or else the supervisor was fully aware of the mistakes, but decided that “it’s the student’s paper, let him make his own mistakes”. And of course, these possibilities aren’t mutually exclusive, since a supervisor who gives his students a lot of freedom and lets them make their own mistakes is the sort of supervisor who might let students submit an ms without first reading it carefully.

My question to you is: are you bothered by this? Because I am, but I’m not sure if that’s just me. I’m bothered for several reasons. First, either possibility I’ve described would seem to be a violation of the published rules of most journals, which require that all authors take responsibility for everything in the manuscript. Second, even if those journal rules didn’t exist, wouldn’t you still want to make sure that any science with your name on it was correct? Third, I’m most bothered by the apparent willingness of some supervisors to effectively force reviewers to do the training that the supervisors ought to be doing.

Note that the situation is totally different if the supervisor isn’t a co-author. As a reviewer, I’m not the least bit bothered if I’m reviewing a manuscript sole-authored by a student and find serious mistakes that a more experienced author probably wouldn’t make. Note also that I’m all in favor of allowing students a lot of freedom, including the freedom to make mistakes. But that freedom does not extend to the freedom to make serious, clear-cut mistakes with my name on them.

But then again, maybe I shouldn’t be bothered by this. One could take the view that it’s the job of reviewers to identify mistakes, no matter what the source of those mistakes. Further, even very experienced people do sometimes make serious mistakes (like believing in zombie ideas!), so maybe my annoyance here is based on the false premise that there are some mistakes that should just never happen in any paper with an experienced co-author.

What do you think? Should supervisor co-authors let student lead authors make serious mistakes? Should reviewers care if they do? Or is this whole post just based on a false premise?

*Actually, I suppose there are at least two other possibilities: the supervisor is aware of the mistakes and their seriousness, but either hopes the reviewers won’t notice or care, or else hopes to be given the opportunity to fix the mistakes in a revision. But I ignore these possibilities, because considering them is too depressing.

p.s. to my own students: this post was not inspired by you!

One of my favorite comics asks whether we shouldn’t just let the zombies win.

My fellow editor Mark Vellend just emailed me with the fruits of his research on how to formally cite blog posts. While standards are still evolving and many ecology journals have no official policy, you can find guidance here and here. (UPDATE: second link fixed)

The latter link includes advice on how to cite pseudonyms, which uses the unintentionally amusing example of citing the Dalai Lama. Mark points out that someone following the linked guidance would list the author of my posts as “oikosjeremy” rather than Jeremy Fox. Which I don’t really think is a big deal–it’s not like it’s a secret who “oikosjeremy” is. Mark suggests that I may want to stop blogging under a pseudonym, in anticipation of being cited. On the contrary, my thought is to start blogging under a different, more entertaining pseudonym. Like “Charles Darwin”. Or “thegreatestecologistintheworld”, like in the old Calvin and Hobbes cartoon where Calvin signs his homework “Calvin, Boy of Destiny”. Or maybe “Author’s Name”, so that the citation would read “A. Name”. Kind of like when two British football (soccer) players a few years ago bought a couple of racehorses and named them “Some Horse” and “Another Horse” in the hopes that one day an announcer calling a horse race would be forced to say “And down the stretch they come, it’s some horse leading by a length over another horse!” (UPDATE: I’ve chickened out and changed my display name to my real name).

I know you’re going to do it anyway, so go ahead and suggest new pseudonyms for me in the comments. Indeed, I predict that top commenter Jim Bouldin is going to spend hours thinking up suggestions on this.

Joan Strassman has a nice post at Sociobiology about how to choose a PhD program. I agree with most but not all of what she has to say.

I don’t agree that you just avoid M.Sc. programs if you think you might want a PhD. Unless you’re sure you want a PhD, doing an MSc is a good way to hedge your bets. It’s a much smaller commitment, both on your part and on the part of your advisor. If you’re unsure exactly what you want to work on, an M.Sc. can be a good way to find out, and gives you a natural way to change directions as your interests evolve. Lots of people do an M.Sc. in one lab and a Ph.D. on a somewhat different topic in another lab. But to do this you have to start, and finish, your M.Sc. first. If you start a Ph.D. and then for whatever reason decide you don’t want to finish it, you can often take an M.Sc. instead, but that’s colloquially known as a “terminal” M.Sc. If you do that, you’ll typically have a very hard time convincing anyone to take you on for a Ph.D. After all, you already tried and failed to finish one Ph.D.–why should anyone think you’ll succeed on your second try, or choose you over a competing applicant who hasn’t failed to finish? You may not think that’s fair, but that’s the reality.

There are other reasons to do an M.Sc. first. You’ll get a paper or two out of your M.Sc. thesis, which will make your CV stronger when you eventually finish grad school and enter the academic job market. Doing an M.Sc. and then a Ph.D. does extend your total time in grad school, but often not by much because you get all your coursework out of the way during your M.Sc. An M.Sc. also can qualify you for various jobs (e.g., certain environmental consulting positions, some technician/lab manager positions) for which a bachelor’s would not qualify you, so it’s not as if you’ve wasted your time if you end up deciding not to go for a Ph.D. And as for funding, while some universities don’t provide funding for their M.Sc. students, many do.

Note that I’m not saying you should do an M.Sc. to find out if you want to go to grad school at all. You should do your homework and figure that out before you start applying to graduate programs. You should be doing an independent study or honors research project, taking research assistant positions, and talking to your TAs to find out what doing research, and graduate school, is like.

UPDATE: Zen Faulkes has a nice post on why grad students fail. Many of these pitfalls can be avoided by doing your homework, and honestly assessing your own background and motivations, before you start applying.

Other advice I’d add:

Choosing the right advisor is more important than choosing the right program. I went to Rutgers because I wanted to work with Peter Morin, even though the Rutgers EEB program wasn’t on any lists of the top graduate programs in EEB. I’ve never regretted the choice. Don’t get me wrong, it’s fun and stimulating to be part of a top graduate program because you have the opportunity to interact with so many really good faculty and students. But your relationship with your supervisor is going to be much more important to your grad school experience. If you get on well with your supervisor and your labmates, you’ll have a good experience in grad school. If not, not. And once you leave, who your supervisor was counts for more than which program granted you your degree. When I was looking for jobs, I wasn’t viewed as a Rutgers graduate–I was viewed as Morin lab graduate.

The exception to the above is that you definitely do not want to do your PhD at the same university where you got your bachelor’s (and your MSc, if you got one). If you don’t leave the nest, people will assume you can’t fly. Getting a bachelor’s in one place and both your graduate degrees somewhere else is fine. And getting a bachelor’s and master’s in one place and a Ph.D. someplace else is fine too. But getting your Ph.D. and all your pre-Ph.D. degrees in one place looks bad, even if you do different degrees under different advisors.

Here’s some advice I really can’t emphasize enough, because it concerns a really common mistake prospective grad students make. Before you contact any prospective advisor (at least in N. America), do your homework. Have a good look at their website, and read a couple of their papers. Then write each prospective advisor a personal email which is addressed specifically to that person and only to that person. Describe your background, interests, and long-term (i.e. post-grad school) goals, and say specifically why you want to join their lab (which doesn’t mean having a specific project in mind, of course). If you don’t do that, you’ve already gotten off on the wrong foot, and with many supervisors (including me) you’re already pretty much doomed. Most every decent advisor is very busy, and receives many, many inquiries from prospective students, many of them obviously bulk emails (many but not all of which come from students in developing countries). Every professor I know deletes such emails without reading them. If you can’t be bothered to take the time to do your homework before contacting me, why should I take the time to reply to you? Rather than signaling that you’re seriously interested in my lab, you’ve just signaled to me that you’re the kind of student who likes to cut corners and who doesn’t show initiative. Plus, if you don’t do your homework before contacting me, I’m just going to reply by asking you about your background, interests, long-term goals, and what specifically interests you about my lab. I mean, how else am I supposed to reply? So why not just save us both some time and send me a detailed, personal email to start with?

And then once you’ve corresponded with a few prospective advisors and narrowed the field to a few top choices, make sure to visit those labs before you make your final decision, and ideally before admission decisions are even made. Besides meeting your prospective advisor, you’ll get to meet their current grad students, see the facilities, and check out the city and the surrounding area. I actually insist on meeting prospective students face to face, with rare exceptions for students who are highly recommended by close colleagues whose judgment I trust. And no, skype or a phone call isn’t really a substitute. Most prospective advisors will at least encourage a visit even if they don’t insist on it, and will be happy to pay for it. Both you and your prospective advisor are considering making a big commitment to each other. It’s to the advantage of both of you to be as sure as possible that you’re a good match.

Here. The diagram on the left is true. The diagram on the right isn’t, but it often feels like it is.

I can’t possibly comment on how true this is.

HT Denim and Tweed.

Biological Posteriors asks a good question: how far down the [mechanistic] rabbit hole should one go to get an answer to any question? For instance, if you want to understand plant distributions, do you need to study plant physiology? Or even plant biochemistry?

Briefly, I’d say it depends on how you’ve framed the question, what sort of answer you’re looking for (e.g., a quantitative vs. a qualitative answer), and whether there’s anything comprehensible at the bottom of the rabbit hole.

But here I want to respond by asking a question of my own: why assume that you can only find the right mechanistic “level” by starting at a high level and then drilling down? Why not go the other way? Why not scale up? That is, start with a (possibly very detailed) “low level” mechanistic description of the physiology, life history, and behavior of individual organisms, and then ask about its higher level implications for density-dependence of population growth rate, coexistence, ecosystem function, etc.? There are lots of successful examples of this approach, indeed too many to list.

Note that this approach need not restrict you to building and simulating very computationally-intensive individual-based models. For instance, it may well be possible to derive a tractable analytical, high level approximation to your individual-based low level simulation. Importantly, that high level model, although simple, may well be different than the simple high level model you would’ve invented if you hadn’t first done the low level model and then scaled up. The work of Drew Purves, Steve Pacala and colleagues approximating the famous SORTIE model of forest dynamics is a fine example (Purves et al. 2008, Strigul et al. 2008).

So how do you decide whether to start high and drill down, or start low and scale up? Well, it’s often good to start at a level at which you already know, or can easily find out, a fair bit. In other words, don’t think about whether to drill down or scale up, think about starting from what you know and then working (upwards or downwards) towards something you don’t know.

It’s also worth noting that, if you don’t know how to drill down, you often won’t know how to scale up either, and vice-versa. This is something I wish a lot of macroecologists would take to heart. Macroecologists often argue that we don’t know how to scale up from individual- and population-level mechanisms to their macroecological consequences. Which is true enough. But they seem to take that as an argument for starting at the macroecological level and then drilling down. Which I confess I don’t understand. For instance, writing in the most recent issue of Oikos, Gotelli and Ulrich argue that we don’t know how to specify and parameterize system-specific process-based models of species interactions and dispersal.* But they present this as a reason to focus on null models that test for certain non-random patterns in presence-absence matrices (data matrices indicating which species are present at which sites). But if we don’t know how to build and parameterize low-level process-based models, why should we be at all confident in our ability to build high level null models that omit the effects of certain processes (such as interspecific competition)? Especially null models that putatively apply, not just to one specific system, but very generally? Because take my word for it, it is really easy to come up with very plausible low-level competition models in which competition generates presence-absence matrices that look nothing like those tested for by any of the standard null models. And conversely, it’s surprisingly difficult to come up with generally-applicable low-level process-based models that produce some of the high-level patterns that null models often test for (such as “checkerboard distributions”, where sites contain species A or species B, but never both). To be fair, I think Gotelli and Ulrich are aware of this issue, although they don’t put it quite this starkly. But I’m not sure even they have fully taken to heart the notion that, if we don’t how to scale up from microecology to macroecology, we don’t know how to drill down either.

*Grouchy aside: I also don’t understand why macroecologists harp on the purported impossibility of specifying and parameterizing low-level models for many species. First of all, as the example of SORTIE (and other examples) shows, it’s perfectly possible to build and parameterize very detailed process-based individual-level models of entire communities, or of dynamically-sufficient subsets of those communities. Second of all, why would anyone think that scaling from microecology to macroecology is totally impossible unless we have a fully-specified and parameterized model of the low-level microecological processes? For instance, you don’t need to build such a model to show experimentally that local communities are effectively closed to colonization (e.g., Shurin 2000). Which is all you need to show in order to refute the once-common macroecological claim that linear local-regional richness relationships imply that local communities are highly open to colonization. I guess I must be missing something here, because very smart macroecologists whose work I really respect keep emphasizing the claim that we can’t build and parameterize low-level process-based models of community dynamics. Which just seems like such an obvious straw man. Hopefully folks will weigh in in the comments on this and set me straight.

 A very nice reporting of the work by Jeff Ollerton forthcoming in Oikos.  The newspaper is The Guardian and here is the link.

The experimental test of the study is described very well near the end of the article.  I will post a link to the Oikos article as soon as I get it.  I think this is both a nice piece of journalistic reporting and a novel, useful study.  Good stuff (like honey).

 

The Buell and Braun awards respectively to the best student talk and poster at the Ecological Society of America Annual Meeting. They’re nice awards: besides the prestige, you get $500 plus travel reimbursement to the following year’s meeting.

To win the awards, the first thing you have to do is register. Unfortunately, there should probably be an award to students who can figure out how to do this. Buried in the “About ESA” section of the ESA website (not in the section about the upcoming Annual Meeting) is a page describing all the ESA awards. Click on “Buell & Braun Awards” to see the application rules and download the application form. The deadline for submitting the application form is Mar. 1.

Note that registering for the awards is a separate process from submitting your abstract (the deadline for which is Feb. 23, 5 pm US ET), and registering to attend the meeting. You have to do all of these things to be considered for the Buell or Braun award.

Note as well that the application form asks for a statement of up to 250 words describing how your research will advance the field of ecology. That statement is in addition to your abstract. Note however that 250 words is only an upper limit. You could just write a couple of sentences, even ones just pulled from your abstract. Without getting into specifics, if the judging works the way it did last year, this approach would not affect your chances of winning.

If all this seems like an unnecessarily complicated process to you, I don’t disagree. Indeed, every year many very good students decide it’s not worth all the bother. In a typical year, less than 20 students register to be considered for the Braun award, and only a few dozen register to be considered for the Buell. Both are small fractions of the total number of students giving posters and talks, respectively. And I know from personal experience last year, as a judge for the Buell and Braun as well as for other student awards handed out by the ESA sections, the many of the very best students only choose to register for the section awards, for which the registration process typically is much easier. That’s even though the section awards are less lucrative.

Probably the silliest part is asking students to write an extra statement about how their research will advance the field of ecology. As a student commenter on last year’s post about the Braun award notes, that’s what your abstract is supposed to do. Students are quite rightly annoyed by application procedures that take up some of their scarce time while serving no obvious purpose. Frankly, I’m surprised that extra statement is still required. Last year I sat on the committee that chose the Buell and Braun winners, and we discussed the application procedure and agreed that the requirement for the extra statement should be dropped. I don’t know why there’s no change in the application procedure this year, but I’ll be asking and will update with any information I find out.*

Despite all that, I strongly urge all interested students to register for consideration for the Buell and Braun awards. The potential payoff is worth the effort, even though much of that effort probably shouldn’t be necessary. Particularly because many of your fellow students aren’t going to bother registering, thereby increasing your odds of winning!

*There is a sense in which these hurdles serve a “purpose”: by holding down the number of applicants, they make it easier to find enough judges. In my view, there are much better ways to ensure that the judges aren’t swamped by too many applicants. Just discouraging people from applying by making the application process unnecessarily complicated has the unfortunate side effect of reducing the quality of the applicant pool, since as far as I can tell there’s no positive correlation between “willingness to apply” and “competitiveness for the award”. Obviously, the powers that be could try to recruit more judges, but they already work pretty hard at that, so it is necessary to find some way to either hold down the number of applicants, or judge them more efficiently. They could reduce the number of judges per presentation. Currently, it’s 4-6, which seems like twice as many as necessary–we only get 2-3 referees on our peer-reviewed papers! They could get a few people to pre-screen the posters in the morning each day and then only send judges to meet with the top candidates during the evening poster session. They could even pre-screen the posters in advance by asking students to submit an image of their poster a couple of weeks before the meeting. As for talks, I suggest only allowing students to apply for the Buell award twice. That way students will only apply for consideration when they feel they have their best stuff to present (in most cases, a nearly-complete MSc or PhD project), and you’ll have a manageably-sized applicant pool that likely includes most of the strongest presentations. I don’t claim any of these solutions is perfect, merely that they’d be better than the status quo. Bringing in some combination of these changes, so that the application process can just be reduced to a checkbox during the abstract submission process, seems like the way to go to me.

In the comments, please provide your own suggestions for how to arrange the application process for the Buell and Braun awards.

Friend of Oikos Blog* Peter Adler and colleagues are seeking applicants for a postdoc in plant population ecology. The ad is below. Peter’s a terrific plant population ecologist and this sounds like a neat project.

In the past I’ve only posted job ads for myself and other Oikos editors, so I’m stretching a bit here. But I decided I’m ok with posting the occasional job ad as a favor to a close colleague, as long as the ad is likely to be of interest to a sufficient number of blog readers, and as long as I don’t feel like the ads are “diluting” the other content of the blog. But I’d welcome feedback in the comments as to whether Oikos Blog ought to be posting job ads, and if so, under what circumstances.

*FOOB

***********************

Plant ecologist/population biologist

We anticipate hiring a post-doctoral researcher for a two-year position with possibility of extension working primarily with Drs. Jeremy James (Oregon State University), Elizabeth Leger (University of Nevada Reno), and Peter Adler, (Utah State University) on a USDA-NIFA funded project. The broad goal of the project is to quantify variation in the demographic processes and ecological conditions that limit native plant establishment along major environmental gradients in the Great Basin. Major duties of the position include: 1) Supervise collection of demographic data by field crews in Oregon, Idaho and Nevada 2) Compile and analyze data, and work with project scientists to build and interpret population models 3) Design and implement additional studies and analyses that complement project objectives 4) Prepare and submit papers for publication.

This project provides an exciting opportunity to ask important questions about native plant recruitment and population dynamics in relation to environmental variation and environmental change. The post-doctoral researcher will have substantial creative latitude to develop complimentary lines of inquiry and also will have numerous opportunities to collaborate with a diverse project team including ecologists, sociologists, economists, and education specialists.

The ideal candidate will have a PhD in ecology or a related field, excellent field skills in plant demography, and experience or interest in population modeling, as well as a demonstrated ability to lead project teams. The permanent work site is negotiable (the position could be based out of Burns, OR, Reno, NV or Logan UT) but the post-doctoral researcher will spend a substantial amount of time overseeing and participating in data collection during the growing season at field sites in Oregon, Idaho and Nevada. The proposed starting date is June 2012, lasting through June 2014, though the start date is flexible. Salary is competitive, and includes benefits. Consideration of interested applicants will begin April 15, 2012, and continue until the position is filled. To be considered, please email a CV, a description of your research interests and background, as well as the names and emails of three references as one pdf to:jeremy.james@oregonstate.edu.

Please feel free to contact Dr. James (Jeremy.james@oregonstate.edu) Dr. Adler (peter.adler@usu.ed) or Dr. Leger (eleger@cabnr.unr.edu) with any questions.

 

Economics, even ecological economics, isn’t something I’d ordinarily write about on the Oikos Blog–it’s not really the blog’s purpose, and it’s not something I’m really qualified to write about. But I’m making an exception to plug a very interesting exercise in ecological economics by my Calgary colleague M. Scott Taylor.

From 1870 to the late 1880s, the American buffalo (bison) population declined from about 10-15 million to 100. The decline itself isn’t a shock–settlers were spreading west, and hunting buffalo and grazing cattle as they went. But the decline was much less steep before 1870. Why the sudden crash? Various explanations have been proposed, to do with things like changes in Native American hunting practices, hunting by the US Army, and the expansion of railroads. None of these explanations are especially convincing.

In a new paper in the Dec. 2011 issue of American Economic Review, Scott uses a combination of theoretical modeling, empirical time series analysis, and historical research to develop a much more convincing explanation, to do with technological innovation and international trade. It turns out that, just before the crash began, tanners in England came up with a way to tan buffalo hide into leather, which previously had not been possible. This instantly created a massive new international market for buffalo hides; previously there wasn’t a big market for any buffalo product. Scott uses old trading records to infer that around 6 million buffalo hides, representing a kill of about 9 million buffalo, were exported from the US from 1871-1873. The combination of a technical innovation, a huge international market, open access buffalo hunting, and fixed world prices (buffalo weren’t a large enough fraction of the world leather supply for their decline to drive up prices) appears to be what ultimately drove the American buffalo to the brink of extinction.

The echoes of this crash still reverberate today. The buffalo slaughter of the 1870s was widely deplored at the time as wanton and wasteful. Some of those who witnessed it first hand, including Teddy Roosevelt, John Muir, and William Hornaday, founded the conservation movement in the US. One of the first and greatest successes of that movement was the creation of the national park system, including Yellowstone and its tiny remnant buffalo herd. And there’s a lesson for today as well: when small countries worry that the combination of technological innovation and global markets will decimate their natural resources, Americans ought to be willing to listen–because the US was once in the same boat.

I’ve seen Scott give a talk on this work, and can attest that it’s a really nice piece of science. I really like the fact that Scott put in the effort to develop every possible line of evidence, including combing through historical records and building a theoretical model, rather than just relying on those bits of evidence which he found easiest to access or develop. Too often in ecology, and probably every field, investigators follow the path of least resistance and focus narrowly on whichever lines of evidence which they find most convenient or congenial to work with. And then ignore or argue with people who’ve come to different conclusions based on different lines of evidence. I also like Scott’s effort to be as quantitative as possible, for instance by estimating how many buffalo hides were exported. It turns what otherwise would’ve been a theoretical model plus suggestive historical anecdotes into a quite convincing story.

See also further discussion at Conversable Economics (on which this post is based).

 

My fellow Oikos editor Andre de Roos passes on word that the University of Amsterdam is hiring a tenure-track theoretical/computational ecologist. Details here.

Sarcozona resumes her series of interviews from last year’s ESA Meeting (better late than never!) with a chat with plant population ecologist Julian Mulroy from the ESA Historical Records Committee.

The best of last month’s online evolutionary writings, here. Get ‘em while they’re hot!

Over at NeuroDojo, Zen Faulkes has been doing a lengthy series of posts on how to give a good presentation. The latest one, on the need to avoid “shortcuts to credibility” (like trying to talk differently than you usually talk), is here. The whole series is recommended for students.

Lots of interesting papers coming out in Oikos in the next little while. I wanted to highlight a few that particularly caught my eye.

In the most recent (Feb. 2012) issue:

• Barto & Rillig on dissemination biases in ecology. This is a really important study. Barto & Rillig analyze the citation rates of almost 4000 papers included in over 50 ecological meta-analyses. Studies that report unusually strong effects (compared to other studies on the same topic) tend to be published first, published in higher-impact journals, and get cited more frequently–even though they also have the smallest sample sizes and so are objectively provide the least reliable estimates of effect size and direction. The overall picture is that we suffer from confirmation bias and theory tenacity–we tend to cite the studies that appear to confirm what we (think we) know, and that appear to support existing theory. I suspect the bandwagon effect also contributes here; everybody tends to cite the stuff everybody else cites. Which is an especially serious problem when, because of confirmation bias, those studies are the least-reliable ones. You wonder how zombies first arise? This is how! (UPDATE: Check out Mike Fowler’s discussion of this paper, including his own personal run-in with an anonymous reviewer who seems to be guilty of the sins Barto & Rillig quantify)
• Gotelli & Ulrich on null models. In an old post I argued that ecologists should refight the null model wars. I didn’t realize when I wrote that post that leading null model proponents Gotelli & Ulrich had already decided to fire the first shot! Although their title suggests a focus on purely statistical issues (such as testing for significant deviations from the null model), in fact the ms actually engages with deeper conceptual issues, such as how we figure out what our null expectation should be in the first place. I’ll try to do a longer post responding to Gotelli & Ulrich at some point, as I disagree with some of what they have to say. But it’s great that they’re bringing out into the open issues that were basically swept under the rug many years ago, and so aren’t familiar to many younger ecologists.
• Valladeres et al. on how forest fragmentation leads to food web contraction. The authors compile a truly massive set of plant-herbivore and host-parasitoid food webs for 19 Argentinian forest fragments of varying area. Smaller fragments harbor a subset of the food webs in larger fragments, specifically the highly-connected “core” species. This is interesting for a couple of reasons. First, it shows that extinction risk isn’t just a matter of a species’ own “traits”, it’s also a matter of the food web context in which the species is embedded. The same species could be more-connected (relative to others in the web), or less-connected, depending on which other species happen to comprise the web. Second, the results may have implications for the links between community stability and “complexity”. If “complexity”, in the form of high connectance, is destabilizing, then you might expect food webs to collapse by losing highly-connected species, the opposite of what these authors found. However, even their most highly-connected webs were still pretty-low connectance in absolute terms. For that and other reasons the results are merely intriguingly suggestive in terms of their implications for complexity-stability theory.
• Fox & Kerr on an extended Price equation partition. Yes, a shameless plug. In nature, species composition often exhibits turnover along environmental or spatial gradients. Ecosystem-level properties and “functions” also vary along those gradients. In this paper, Ben Kerr and I use Ben’s clever extension of the Price equation to partition between-site variation in ecosystem function into components attributable to different effects (e.g., changes in species richness vs. changes in species composition vs. environmental effects on the functioning of individual species). One interesting insight is that, when there is compositional turnover, there’s no single “species richness effect” and no single “species composition effect”. Rather, there are two of each. So all those arguments ecologists have had about how to separate “the” effect of species richness from “the” effect of species composition are badly framed.

We have many more interesting articles in the pipeline, so look for more “highlights” posts in the near future.

Ace philosopher of science and statistics Deborah Mayo has moved her blog. It’s now here (and now looks much sharper).

The previous post referred to a philosophy talk about Robert MacArthur, his observations of feeding warblers, and the competition models which his warbler work helped inspire.The speaker apparently drew some general lessons about the conduct of ecological science from MacArthur’s example.

Maybe I’m just grouchy today, but I have to ask: is it really healthy for ecology, and for those philosophers and historians who study ecology, to place so much weight on so few historical examples? Does the example of Robert MacArthur and his warblers really have anything left to teach us about ecology or how to do it? I mean, I know he was and remains hugely influential, and rightly so. But should he, or any ecologist, really be treated like Shakespeare or the Bible, an inexhaustible source of inspiration and insight? MacArthur has been dead for forty years, and it’s only in the broadest and loosest sense that any ecologist these days actually does “MacArthurian” ecology, however you might define that. Seriously, do you think any journal today would even publish MacArthur’s warbler data? Much less take those data as strong evidence for his competition model? And it’s not just that our standards of evidence are higher today, it’s more profound than that. Ecology today, even as practiced by those who are self-consciously influenced by MacArthur and take him as a role model, is not just “MacArthur, only with better data.” And by that I don’t just mean that today we worry about forces other than competition. For instance, West et al. (1997), hugely influential and co-authored by two ecologists who hold MacArthur in the highest regard, strikes me as very far from the sort of thing Robert MacArthur himself would ever have done. Unless you regard as “MacArthurian” any paper concerned with explaining general patterns, which I really don’t think you should.

In continuing to pay so much attention to MacArthur, we necessarily ignore other voices from ecology’s rich past, we misunderstand our present (including the influence of MacArthur on the present) by viewing it through an outdated lens, and (as Peter Kareiva has argued) we misdirect our future efforts. MacArthur was an important ecologist and in many ways remains a fine role model. But he’s far from the only important ecologist, and far from the only role model. Just for starters, I’d suggest that modern students of ecology can learn at least as much from Gause as they can from MacArthur.

I should emphasize that, not actually having seen the talk Joan saw, I have no idea why MacArthur was chosen as an example and I’m totally not criticizing the talk or the speaker. The subject of the talk just happened to prompt the above thoughts, which otherwise probably would’ve been prompted by something else at some point.

UPDATE: Can’t believe I forgot to mention this in the original post, but I count myself among those who had to get over MacArthur. In grad school, I did a side project on the propagation of indirect effects in food webs, which I ended up publishing in Oikos (Fox and Olsen 2000). This genesis for this project was that I’d read a famous but quite odd paper of MacArthur’s on complexity and stability (MacArthur 1955). I decided that figuring out what the heck MacArthur was talking about and then testing it would make for a good side project. In the end, I think my paper was perfectly fine. But the starting point (“I’m going to identify an interesting question by doing textual exegesis on an old paper of Robert MacArthur’s”) was not one I would ever choose again.

UPDATE #2: Jay Odenbaugh, the philosopher who gave the talk referred to in the previous post, actually has done a lot of interesting-looking work in philosophy of science, as applied in the context of ecology (a field of science that philosophers until recently have not given much attention). As someone who’s always encouraging other ecologists to read more philosophy, I’m embarrassed that I wasn’t already familiar with all of his work (update: especially since I learned that he got his PhD at Calgary, where two of my closest colleagues were on his committee!) I’m looking forward to rectifying that. Interestingly, he is working on a paper on philosophical issues raised by Hubbell’s neutral theory. So Jay is not someone whose attention is drawn only to a limited range of famous historical figures in ecology.

UPDATE #3: Jay Odenbaugh himself has popped up in the comments with some very thoughtful remarks. Thanks for stopping by Jay!

Nice post from Joan Strassman at Sociobiology on the art of seeing the big picture, the forest for the trees. The way to do that is to have “blurry vision”, so that you can’t see individual trees at all. Read the whole thing.

Joan’s thoughts were prompted by attending a philosophy seminar. For a while now I’ve been meaning to do a post on a sort of implicit, running theme of this blog, the ways in which thinking philosophically can help you be a better ecologist. Must get around to writing that post at some point…

 

Interesting discussion thread over at statistician Andrew Gelman’s blog, about time series analysis of the lynx-hare cycle. Standard phenomenological statistical models (autoregressive moving average models) don’t fit or predict these data all that well. Andrew links approvingly to a recent statistical paper which does much better by fitting a simple mechanistic model–indeed, a laughably simple model, the original Lotka-Volterra predator-prey model! Ecologists Eric Pedersen, Ben Bolker, and myself then showed up in the comments to point out that ecologists have been mechanistically modeling the lynx-hare cycle and other ecological time series for 20 years now. I think this work–by Royama, Turchin, Wood, McCauley, Ellner, Harrison, Rees, Kendall, Bjornstad, Grenfell, King, Keeling, Ben himself, and numerous others*, is some of the best and most important ecological work of my generation. This work has always involved collaboration between ecologists, applied statisticians, and people who wear both hats, but apparently it’s not widely known among statisticians. Hopefully the linked thread will help to change that, as Andrew’s blog is very widely-read by statisticians.

Which does raise the question, are there other ways to increase cross-talk between statisticians and ecologists? NIMBioS is doing its part, and IIRC the Canadian statistical society had a special session on mechanistic time series models at its annual meeting a couple of years ago. What else could be done? Certainly sounds like there’s an opportunity for somebody (not me) to write some sort of perspectives-type review paper for a stats journal, highlighting the success of mechanistic time series modeling in population and disease ecology.

And while we’re at it, how about more cross-talk between the population and disease ecologists who’ve been doing most of this work, and ecologists working in other areas? Ecologists as a whole are increasingly statistically sophisticated. But often that sophistication is at the service of fitting and testing purely phenomenological statistical models. This is problematic, because translating the predictions of mechanistic ecological models into a form that can be evaluated by phenomenological statistical models often is tricky. Unfortunately, this translation process often seems to be based on nothing more than intuition and arm-waving, leading to an unfortunate trend towards using rigorous, sophisticated methods to test shaky, unsophisticated hypotheses. Why not try to follow the lead of the best population and disease ecology and fit mechanistic models directly to our data? And don’t say “we can’t do that because we don’t have enough data”, because one thing ecologists should not be in this LTER-NCEAS-NutNet-NEON-CIEE era is data-limited.

*My own efforts along these lines are so far quite modest and not without problems. But I hope to do better in future. A long-term ambition of mine is to apply this approach to many-species communities.

This kind of sums up my experience of this debate:

Well, either that, or this:

UPDATE: I’m kidding, of course.

Back in the blog’s early days I did what I still think is a quite nice little post on whether we should expect species’ phenotypic traits to predict or explain their abundances. I’m going to be fleshing that post out into a short perspectives-type review paper, mixing simple theory and some illustrative data. Given all the interest in “trait-based” community ecology these days (I’ve even heard it called a “new paradigm”*), this seems like a useful and timely paper to write.

Just for fun** I decided to crowdsource the data. So in the comments, please share your favorite examples of papers using species’ traits to predict their abundances. I’m specifically interested in community ecology papers, so really what I want is papers taking trait data for a bunch of species, and using those data to statistically predict or explain the abundances of those species either within a single site, or across sites (say, along an environmental gradient). I’m particularly interested in microbial examples, but non-microbial examples are great too. Note that the predictions or explanations need not be successful. I plan to cite papers reporting both successful and unsuccessful predictions and explanations, because the whole point of the ms is to argue that we only expect species’ traits to correlate with their abundances in certain circumstances. I of course have some papers in mind, but I’m genuinely curious to see what examples y’all suggest.

Afraid I can’t offer any reward here except my thanks. So, thanks in advance for your help!

*Crotchety old guy rant: Whenever I hear someone call some scientific idea or research approach a “new paradigm”, I immediately think of this line from one of my (and most people’s) favorite movies. This is followed by the wish that I had the power to do what Woody Allen does in Annie Hall (2:10 mark) and pull Thomas Kuhn out from behind a signboard. /crotchety old guy rant

**For “Just for fun” read “Because I am lazy”

Modern science is increasingly a collaborative enterprise. In ecology, NCEAS was very influential in driving the shift towards collaboration. But The Curious Wavefunction asks a good question: what if one side effect of this shift is to devalue the scientific contributions of introverts? It’s not just that some good scientists prefer to work alone, perhaps because they aren’t comfortable (scientifically and/or socially) in groups. By virtue of their independence, those sorts of scientists may be the creatives and contrarians, the sources of really new ideas. In future, is it going to be harder for such people to establish themselves and make their mark in science? Is it already getting harder?

I suppose I tend to appreciate this question because until recently I’ve mostly worked solo, or with a very few collaborators. And I certainly know some very good ecologists who are also very shy, and I wonder how that’s shaped their careers.

Following on from the previous post, another way in which community ecologists often misinterpret neutral models is by mixing up neutrality with dispersal limitation. This leads to mistakes like testing for neutrality by testing for a community ecology equivalent of “isolation by distance”, where more widely-separated communities are more different in species composition, independent of any environmental differences.

I honestly have no clue how this zombie idea could’ve gotten started. Sewell Wright’s classic paper defining isolation by distance talks at length about how migration rate among subpopulations affects their isolation, implying that isolation by distance is not a signature of neutrality. Even in a neutral world, different subpopulations can be as different or similar as you want, depending on the migration rate. Neutral models can perfectly well be panmictic. Wright’s paper also has a whole section on how selection affects isolation by distance, showing that you can still have isolation by distance even in a non-neutral world. Much the same is true in the neutral model with which community ecologists are most familiar, Hubbell’s neutral model. That model remains neutral even if the migration rate is set so high that every newborn individual in the local community is an immigrant from the “metacommunity”, so that there’s no “dispersal limitation” at all. More broadly, ecological theorists have for decades considered all sorts of non-neutral models with all sorts of dispersal rates, from completely closed systems (=zero dispersal) to highly open ones. And that’s before we even start talking about things like sophisticated habitat selection behavior on the part of migrants, which can generate isolation by distance in species composition even in non-neutral, drift-free systems. So how did anyone ever get the idea that low migration rates or isolation by distance are synonymous with, or even tend to be associated with, neutrality?

Don’t get me wrong, dispersal limitation is an interesting and important phenomenon. But it has nothing to do with neutrality vs. non-neutrality.

HT to Brian McGill for beating me to the punch on this in a comment on the previous post.

Recent interest in neutral theory in community ecology has given rise to a zombie idea: that “neutral” and “stochastic” mean the same thing. They don’t.

It’s easy to see where this zombie idea comes from: neutral drift in evolution is a stochastic process, while natural selection is a deterministic process. The problem here is confusing “neutral” with “drift”. It’s perfectly possible for a dynamical system to be both neutrally stable and deterministic. In evolution, a population experiencing no selection evolves in a neutral fashion, even if it also has infinite population size and so experiences no drift. The initial gene frequencies will simply persist forever, and if perturbed, they’ll subsequently persist forever at their post-perturbation frequencies. In ecology, a Lotka-Volterra competition model in which intra- and interspecific competition are equally strong on a per-capita basis is neutrally stable, but deterministic. And the original Lotka-Volterra predator-prey model (exponentially-growing prey consumed by a predator with a linear functional response and density-independent per-capita mortality rate) produces neutrally stable oscillations, the amplitude of which is set by initial predator and prey densities. Conversely, a world in which the strength or direction of selection fluctuates randomly over time is a stochastic world, but not a neutral world (although in certain respects it might have similar dynamics to a neutral world).

Confusing “neutral” with “drift” has led community ecologists to use some problematic tests for neutrality, which basically test for random or unexplained variation in species composition or other community properties. At best, these are tests for drift (and they may not even be tests for that, depending on the details). They’re not tests of neutrality (=zero selection). And it is simply not true that, in nature, neutrality and drift always and everywhere go hand in hand, so that there’s no practical harm in not distinguishing them.

Now, if you want to argue, as some ecologists have, that the most interesting or important distinction to draw is between deterministic vs. stochastic processes, rather than between neutrality vs. non-neutrality, that’s fine.* But you do need to recognize that those are two different distinctions.

*Although I would argue that even drawing that distinction is likely to cause you miss all the interesting ways in which stochastic and deterministic processes can interact and produce dynamics that are qualitatively distinct from those that would be produced by either deterministic or stochastic processes alone. Quasi-cycles (Pineda-Krch et al. 2006 Oikos) and stochastic resonance, for instance. In general, I think ecologists often are too quick to either treat non-mutually exclusive possibilities as mutually-exclusive alternatives, or else to treat them as ends of a linear continuum. The latter mistake is more pernicious because it’s so intuitive and accords so well with our everyday experience. In between tall and short are people of intermediate height. In between loud sounds and silence are sounds of intermediate volume. In between the two ends of a line segment is the halfway point. In between black and white are various shades of grey. But “in between” determinism and stochasticity is all kinds of cool, wonky stuff that cannot be understood by simply thinking of it as a “mix” of determinism and stochasticity. It’s as if mixing black and white made, not shades of grey, but all the colors of the rainbow.

Following on from previous posts on reforming peer review (see here, here, and here), I wanted to note a new peer review service, Peerage of Science. PoS is a private company founded by a trio of Finnish scientists, which combines several proposed peer review reforms into one package and then offers them as a service to participating journals, for a fee. Member scientists, called “peers”, can submit mss to the service, which other peers can review (anonymously). Members are obliged to review in appropriate proportion to how much they submit, an element of PoS which echoes PubCreds. Participating journals can see the mss and the reviews, and notify the authors if they want to accept the ms. Authors then can choose their preferred offer of acceptance, an element of PoS which echoes ExpressO. To provide additional incentive for scientists to join (which they can do for free, at the invitation of a current peer), and to perform reviews, peers can score one another’s reviews for quality. So if you’re a good reviewer, you can build up a good score which, at least in theory, is an objective number that you can put on your cv. There are also plans to publish the best reviews in a commentary-type journal, to provide an additional incentive to review. Incentives for review have of course been widely discussed.

PoS is just getting off the ground. They currently have about 500 peers, many of them Scandinavian ecologists. Our sister journal Ecography is currently the only participating journal.

Neither I nor Oikos endorse PoS (full disclosure: I was invited to become a peer, but declined for personal reasons), but I do find it interesting to think about and so wanted to post on it.

As with other proposed peer review reforms, I think PoS will succeed or fail depending on whether a sufficient number of the “players” (authors/reviewers, journals, publishers) think it’s to their benefit to use the service. Any serious proposal to reform peer review either has to respect, or have some plausible way to change, the incentives faced by authors, journals, and publishers. For authors and reviewers, I think the incentive to join depends very much on the number and identity of the participating journals. Prospective peers are going to be asking themselves, “Will joining PoS let me publish more papers, faster, in better journals, than I otherwise could?” Frankly, I don’t think the possibility of accumulating a good review score, or of having some of your reviews published in a “journal of peer reviews”, is going to do much at the margin to attract those who wouldn’t otherwise join. What would be attractive would be the possibility of having many journals “bid” on your mss, based on only one set of reviews. But there may be a catch-22 here, because I’m not sure the service will be attractive to journals unless there are many, many peers submitting and reviewing many, many mss, so that journals feel like they need to be able to tap into that ms stream. This is hard to judge, though, as much depends on the fees that PoS charges. How much is it worth to journals or publishers to essentially outsource their reviewing? PoS also is going to be competing with the “cascading review” services that publishers have begun to offer. Publishers are more than happy to have effectively simultaneous submissions–to their own journals–and to share reviews–among their own journals.

It’s also possible that some scientists will be uncomfortable joining PoS because they will feel like they’re “working for free” for a private company. But on the other hand, reviewing for any journal except a non-profit journal amounts to “working for free” in the same sense, as does submitting to any for-profit journal that charges author fees. There are complex issues here which need to be unpacked.

UPDATE: Mike Fowler pops up in the comments with some trenchant thoughts, and a link to an excellent post on PoS at his own blog. Mike’s clearly thought much harder about PoS than I have!

In science, we “match” manuscripts to journals by submitting to them one at a time. If the journal declines the ms, we resubmit elsewhere. This is arguably an inefficient system. Leading journals like Oikos decline the vast majority of mss they receive, and it is not uncommon for authors to resubmit the same ms to several journals before it is finally accepted. That’s a lot of rejections per “match” (accepted ms). What if there were a way to achieve roughly the same (or even a better) “match” between mss and journals without all the rejections and associated effort on the part of authors, reviewers, and editors?

There are various ways in which the efficiency of the system might be improved. Here’s a particularly radical one: let journals “bid” on mss. The basic idea is that mss would be submitted to many journals at once, and any journal that wanted to could “bid” on the ms by offering to accept it. The author would then choose among the “bids”.

Before you start in with all of the objections that I’m sure immediately occurred to you (and to me, too), know this: this is already how legal journals (law reviews) operate. ExpressO is a service in which more than 750 law reviews participate. It began a few years ago; before that law journals operated the way science journals do (except I think many law journals still were taking paper submissions). Now, authors upload their mss to ExpressO, which submits the mss to whichever participating journals the author wishes. Any journal that wants to accept the ms uses ExpressO to notify the author, who chooses among these “bids”. The service is free to journals. Authors either pay a small fee ($2.20 USD per journal to which they want their ms submitted), or their institution pays on their behalf via an institutional subscription. If you’re curious about other details (and there are many–ExpressO seems to be quite a sophisticated and refined service), click the link.

Neither I nor Oikos endorses ExpressO. I don’t know enough about it to endorse it or even argue for it, and I don’t speak for the Oikos board. But I do find it intriguing, and so wanted to bring it to your attention. It appears to solve a problem that many legal authors and journals wanted solved. In contrast, many proposed reforms of the scientific peer review system, including one I’ve stumped for, aim to solve problems that many argue don’t even exist. And ExpressO seems to be narrowly tailored to solve the problems it solves, which should reduce the risk of unintended negative side effects. And it seems to take advantage of, rather than ignore, the incentives that authors, journals, and employers face. Food for thought.

Amusing video here. But it’s mostly not sh*t ecologists or evolutionary biologists say (well, except “stochastic”, “got any virgin flies?”, and “more work is needed”). Feel free to suggest sh*t ecologists or evolutionary biologists say in the comments.

 

It’s not actually a response to my post asking if even the best lectures are bad, but it might as well have been: Joan Strassman at Sociobiology has a nice post arguing that lectures absolutely have their place. Includes some tips on how to make your own lectures better.

Recent changes in the grant application procedures of the US National Science Foundation have prompted much discussion, and have renewed the debate over the best way for governments to fund scientific research. I have argued in favor of the Canadian NSERC system. Briefly, their approach, which is almost unique in the world, is to fund long-term research programs rather than individual projects, and fund them at a relatively low average level (though with high variance around that average) so as to maintain a high success rate (because it’s not “long term” funding if most people’s research programs get cut off every few years in favor of other people’s).

Anecdotally, this system seems attractive to a lot of non-Canadians. But some non-Canadians hate it. In particular, I have the anecdotal impression that some US researchers think the Canadian NSERC system was invented because we Canadians are too weak and/or lazy to compete properly. So we’ve invented a collectivist, everybody-gets-some-money-and-nobody’s-allowed-to-get-too-much system that funds lots of weak science, underfunds the best science, and breeds further laziness in Canadian scientists. Success rates at the US NSF may be low, they say or imply, but it’s only losers who complain–the very best projects get all the money, as they should. As a result, US research is way better than Canadian research, not just in aggregate (because the US is a much bigger country that spends much more on research), but person-for-person and dollar-for-dollar.

Ahem. See in particular Fig. 5. No, it doesn’t break out NSERC-funded research from other Canadian research. And yes, I know these data are old (but is there any reason to think the picture’s changed hugely in the last decade?) I’d welcome links to more recent and NSERC-specific data. But NSERC is not a trivial fraction of Canadian research spending, and NSERC researchers are not a trivial fraction of Canadian researchers. Indeed, see Fig. 3 in the linked publication for data showing that Canada’s areas of greatest research strength are areas which are NSERC-funded. If NSERC’s approach was so terrible, do you really think Canada would crush the US (and beat every other member of the G8 except for the UK) in publications per researcher, citations per researcher, and citations per unit GDP, and equal the US on citations per unit higher education R&D funding?

Canada’s national animal is the beaver, the choice of which tends to draw a lot of snarky comments from our southern neighbors, who made a different choice. But beavers are some of the best non-human engineers in the world, and on a per-capita basis they have far more impact on their environment than any bird of prey. I can’t think of a better symbol for a nation that punches well above its weight scientifically. The beaver is a noble animal.

You want to argue that there are constraints that prevent your country from switching to anything like the NSERC system? Fine (I don’t know that I believe you, but fine). You want to argue that there other funding systems, quite different from NSERC, that can also produce a lot of bang for the buck? Hey, I totally agree (the data in the linked paper put the UK at the top of the scientific productivity league table) But if you want to argue that the NSERC system can’t or doesn’t work? Sorry, you’d better back up your claims with data, rather than stereotypes, anecdotes, and baseless a priori assumptions. Because I wouldn’t have thought that any hyper-competitive US researcher would ever be so lazy as to rely on that kind of thing.*

*To my many US colleagues: the snark here isn’t aimed at you. The snark here is borne out of a frustrating series of exchanges I had with a US blogger, who has some thoughtful and interesting views on science funding, but who sees arguments in favor of the Canadian system as too stupid to respond to with anything other than silly jokes, and who retreats to the safety of her own blog crying “censorship” when others complain that her silly jokes are derailing a productive discussion. She’s apparently convinced that I’m just a humorless old guy who can’t take a joke or laugh at himself and who isn’t aware of all internet traditions regarding commenting. Or that I’m such a crappy writer that she can’t even tell what I’m talking about and so doesn’t feel able to respond to my comments with anything other than silly jokes. Please excuse the public vent. I feel better now.

SciCurious has a poll up asking readers how many papers they read per week, and whether they think they read enough (so far, most respondents don’t think they do). Which prompted this rather peeved reaction from DrugMonkey, about how the number of papers one reads is meaningless, and certainly not something one should brag about. Reading is a means to end, and one reads very differently depending on the purpose for which one is reading. That often means not reading full papers, but perhaps just skimming the figures.

As a grad student, I read a lot, by which I mean I read the full text (often while making marginal notes) of every paper that interested me in every leading journal (and I have pretty broad interests). I also read the full text of many, many older papers. I really took to heart Steve Stearns’ “modest advice” to read and think exhaustively if I wanted to make a success of my graduate program. I eventually became my lab’s walking, talking EndNote database; anyone who was trying to remember a citation would come ask me. I tried my best to continue the habit as a postdoc and later as a faculty member, but it’s gradually been crowded out by other demands on my time, even though I can’t quite bring myself to admit it (I have a growing folder of unread pdf’s on my hard drive called “To Read”). And yes, reading that much was something that I was proud of, probably in part for the wrong reasons (the sort of reasons that got DrugMonkey annoyed). But then again, as I’ve remarked elsewhere on this blog, I definitely think I’m a better scientist for having read that much, that broadly, and in that much detail. So even if my reading habits did in part (but only in part) reflect a rather silly desire to just read lots of stuff for its own sake, as if whoever dies having read the most words wins, well, all that reading still had beneficial effects. Put it this way: had my reading been solely motivated by a ruthless calculation as to how much and what sort of reading would best develop my “scientific chops” (as DrugMonkey puts it), I’d have chosen to read the same way.

Which is not to say that’s the right way for everybody. Brad Anholt, for instance, once told me that he reads the Introductions, and just the Introductions, of every paper published in every leading ecology and evolution journal. He said that that’s how he keeps up with current thinking in the field–what are the big questions, what do we know about the answers, and what are people doing right now to increase that knowledge? I was impressed with this, both because it sounded like a really good approach given the goals of his reading, and because I couldn’t imagine myself finding the time to do it, even as a grad student!

These days, I read a fair number of abstracts (basically, as many abstracts as I once would’ve read full papers). And I read–or plan to read!–the full text of a much smaller number of papers that look really interesting or are directly relevant to my own work. This lets me keep up with the field (not as well as Brad does!), and ensures that I’m very familiar with the stuff I really need to be familiar with.

I encourage you to click through and check out the linked posts, especially the one by DrugMonkey. DrugMonkey is right that your reading should be tailored to its purpose, and that faculty give students the impression of erudition simply because they’ve had more cumulative time to read a lot of stuff (so don’t feel like you need to have read everything your supervisor has read by the time you graduate). Not sure I entirely agree that Discussion sections should just be skipped, though. DrugMonkey’s post is titled “I don’t give a flying fig about your interpretation of your data”, because what he cares about is his own interpretation of your data. I can see where he’s coming from–I certainly read critically and don’t just take the author’s word for what the data mean. But I do read Discussion sections, for various reasons. For instance, they sometimes make the paper much easier to understand. It’s particularly helpful to be walked through difficult math by an author who’s a gifted explainer, like Steve Ellner or Robin Snyder. And even if I don’t agree with an author’s interpretation of his or her own data, I often want to know what his or her interpretation is, so that I’m aware that I disagree. Such disagreements can be good fodder for one’s own research (and one’s own blog!) Work that convincingly undermines a prevailing view can be very important.

So how, and how much, do you read?

As an editor, reviewer, and committee member, I have seen many authors and students give the following rationale for their chosen research approach (e.g., choice of study design, analytical method, or response variable):

“This approach has proven useful in other contexts, so we chose to use it in our context.”

This rationale is weak. A skeptic might rephrase it as follows:

“This hammer has proven useful for hammering nails, so we tried hammering on other things to see if any of them were also nails.”

It is not enough to argue that a particular approach is useful in general, or even that it is useful in contexts that vaguely resemble your context. Lots of things vaguely resemble nails, such as screws and one’s own fingers. You should be able to make a very specific case that you have chosen the right tools for your particular job. Choosing your tool first, independent of the job, is a good way to end up using the wrong tool for the job. Your rationale for your research approach should be “We needed to hammer a nail, and so we chose to use a hammer.”

Like most profs, I haven’t had much formal teacher training. I’ve had a bit, and I’ve had some informal training (some of it from my wife, who is a trained teacher). I do care about my teaching, and I’m aware of the large body of research that says that just standing up in front of a class and lecturing is an ineffective way to teach most students. But I still mostly lecture, though I’ve slowly cut back on it over the years. I still lecture in part because I think I’m a solidly above-average lecturer and so I’ve always questioned whether that body of research really applies to me.*

But reading this post over at Crooked Timber gave me pause. It relates how a really good lecturer at Harvard–so a really good lecturer lecturing to really good students–found out that his students just weren’t grasping the material nearly as deeply as he thought they were. He was lecturing as well as possible, under the best circumstances possible–and it still wasn’t working. Now, I do wonder a little how he could’ve failed to realize it wasn’t working (the post includes some thought-provoking speculation on this point). But still, a pretty sobering story for someone like me.

The post goes on to talk about some radical alternative approaches. I don’t know that the most radical ones–like having students do the lecturing!–are feasible for the sort of material I, and probably many of you, teach. But recently my gradual ramping down of the amount of lecturing I do, in favor of things like spending one class session a week having students work in pairs on practice problems and thought exercises, has started to bear fruit. So even if you feel, with justification, that you’re an excellent lecturer, you may still want to try lecturing a bit less. There are ways to do this even in a massive class, and even if you’re not (yet) prepared to give over entire class sessions to non-lecturing activities. For instance (to pick just one example of the many non-lecturing things you can do), you can break up your lectures by posing questions to the class and have the students talk to their neighbor for a minute to come up with the answer. Then call on a few students, who will probably give different answers if the answer is non-obvious, or if there is no single right answer. Those different answers then provide a jumping off point for another minute of discussion-in-pairs, or for the next bit of the lecture.

I encourage you to share your best tips for how to teach without lecturing in the comments. And do check out the comment thread for the linked post as well–comment threads at Crooked Timber are lengthy and excellent.

*Yes, I know that most everyone thinks they’re above-average at everything they do. But in my own defense, I do have some independent reasons for thinking I’m an above-average lecturer.

Dear Oikos and Nordic folks,

Thank you so much for your feedback on the editorial ‘Money for nothing and referees for free’ published in Ideas in Ecology and Evolution in December. The most compelling and common question I was asked was is there a referee crisis in ecology (or tragedy of the ‘reviewers common’ as Hochberg et al. proposed). This is an excellent question. I propose that whilst there are more perfect ways to test this (total up # of submissions and then estimate total pool of referees, tricky), an interesting indicator would instead to be calculate the decline to review rate (d2rr) in ecology. I envision the following two primary data streams to calculate this rate: a per capita estimate derived from each of us personally and a mean estimate of rate from the publishing portals (journals). Hence, let’s do it. Only you know your decline to (accept doing a) review rate across all requests whilst journals track their own net rates and your specific rate with them too.

So, please take 30 seconds and fill in this short survey, and we can then assess to an extent whether there is a referee crisis in ecology.

https://www.surveymonkey.com/s/VD3K36W

I have also compiled a long list of emails for every editor I could find for all ecology journals and have contacted them to see if they would share the rate at which individuals decline for each of them, i.e. do they have to ask 5 or 6 people to even secure two reviews? I will not share the journal names etc. and protect their rates as I recognize the implications. I would just like to know what our overall mean is from a journal perspective too.

Thanks so much for your time and help with these discussions. I hope you think they are important too, but I also want to assure you that this is my penultimate post on the subject.

In a previous post I linked to some advice from sociobiologist Joan Strassmann on how to write a good NSF preliminary proposal. But that’s not the only thing Joan Strassmann has good advice on. Her blog is an awesome resource, packed with good advice about succeeding in academia. The writing is excellent too–warm, wise, and humane. Wish she’d been blogging when I was starting out! (not that I didn’t get good advice when I was starting out, but more is always better)

The NSF branch that funds ecology & evolution now requires preliminary proposals, and the deadline is fast approaching. Here is some very good advice, from an experienced NSF panelist, on how to write a good preliminary proposal.

Even if you’re not writing an NSF preliminary proposal you should read the linked advice, because it includes some very interesting discussion of where good research ideas come from (protip: read a lot, and read very broadly).

HT: David Inouye at Ecolog-L

From NewsBiscuit, the British equivalent of The Onion.

Following up on my previous post looking back at the Oikos Blog’s first year and ahead to 2012, here are some ideas I had for new things we might do with the blog. Please vote for the ones you’d like to see happen. You can vote for more than one, and also write in your own ideas. Feel free to elaborate in the comments.

Personally, I like all of these ideas. Especially since they wouldn’t involve much additional work on my part (well, unless I was the one assigned to conduct the interviews)

In seriousness, I’m quite keen on the idea of a new journal section devoted to publishing our best posts (which might of course include guest posts, if we were to start doing them). I do think we have posts that make a sufficiently substantial contribution that they are worth publishing in a citable, indexable, archivable format. But because they’re blog posts it’s probably worth separating them from the rest of the journal’s content, in order to allow them to maintain their distinctive style, and in order to make clear that they haven’t gone through the usual peer review process. If we did this, we’d actually be returning to something Oikos used to do, in the form of John Lawton’s View from the Park column. But I’d really like to hear what you think, and I hope Chris Lortie (who’s in charge of the blog) and the new EiC will want to as well.

UPDATE: The poll will remain open, but we have enough votes in that it’s worth summarizing the results. With this kind of poll I can’t actually tell how many people have voted (save that it’s at least 35 and <144), but so far the most popular idea is invited guest posts presenting opposing or provocative views. Next is guest posts by authors–which I’ve just learned is something Chris has been trying to arrange for months, but no authors have taken up the offer! Next most popular option is more posts by other editors, which is already happening and we’re going to try to do more on this front. Bringing up the rear are invited posts responding to posts by me or the editors (not surprising, really; that’s what comments are for), author interviews (kind of surprised folks aren’t more keen on that; guess folks would rather hear from authors in a more open-ended way?), and publishing our best posts in the journal (I can understand the lukewarm support for this, even though I personally like the idea).

Sing About Science is a massive, free, searchable online database of science and math songs. There’s a huge range of professional-quality material, aimed at various ages, which is updated regularly. You can listen to, watch videos of, and/or purchase many of the songs online, via links the site provides. It’s the brainchild of, and curated by, a buddy of mine, University of Washington malaria researcher Greg Crowther, who also writes and performs his own science songs. Whether you’re looking for some science-y music for your kids, your students (songs can be a fun and effective teaching aid), or yourself, it’s one-stop shopping. And there’s a bunch of ecology and evolution stuff on there. Check it out!

At the EEB and Flow. Lots of good stuff, as always.

The Oikos Blog began in March, so while we’re not yet one year old the start of the new year seems like an appropriate time to look back at how we’ve grown.

We wrote 170 posts and got just over 58,000 views in 2011. Note that that doesn’t count syndicated views and so is a substantial understatement of the total number of views (probably by more than 15%). WordPress doesn’t provide stats on the number of unique visitors, so we don’t have those numbers.

The biggest single day, with 948 views (again, not counting syndicated views) was May 27, which was the day we first advertised the blog on Ecolog-L, the widely-read ecological listserv. After that, the blog readership jumped, and its grown steadily since then. In November and December, we got over 9100 non-syndicated views/month. By way of comparison, the Ecological Society of America Annual Meeting is attended by about 1/3 that many people. I know that’s an apples-to-oranges comparison, but neither is it apples-to-bricks. The Oikos Blog really does have a substantial readership for an academic blog aimed at a fairly narrow audience.

WordPress doesn’t provide super-detailed stats of where our visitors come from, but they basically come from places where there are lots of English-speaking ecologists: The USA, followed by Canada and the UK, with Germany, Spain, and Australia trailing behind. But we have non-trivial numbers of readers from Asia, Central and South America, and even Africa.

The most viewed post was, not surprisingly, “Zombie ideas in ecology“, with 1460 views (including syndicated views). That was also the most-commented post (39 comments), although those comments included numerous trackbacks. Other popular posts were my compilation of resources on Bayesian vs. frequentist statistics (1167 views), my takedown of Gould and Lewontin’s “Spandrels of San Marco” (1149 views), my story of how I almost quit science (874 views), and my argument that ecologists should refight the null model wars (862 views). And of course, none of those totals count all the people who read those posts by visiting the blog’s homepage, or by scrolling down from newer posts. Without meaning to be cocky, I wouldn’t be at all surprised if those numbers compare favorably to the typical number of people who read the full text of a typical article in a leading ecology journal.

I’ve said it before and I’ll say it again: I’m tremendously pleased and flattered that the Oikos Blog has built such an audience, so thank you all for reading. I enjoy writing for the blog, but I don’t just do it for fun, or because I’m tenured so I can whatever the heck I want (which I can’t, not if I want to keep getting grants). Blogging doesn’t take me much time (less than an hour a week the majority of weeks, a few hours a week max if I decide to write multiple lengthy posts), and I’m fully convinced that it’s a good use of my time. I’m way more well-known in the field than I was nine months ago, which has some very concrete benefits. For instance, this year I attracted many more really strong applicants for the Killam postdoc than I ever have before, and I have good reason to believe that’s not just random chance or a reflection of the weak job market. I’m also starting to get inquiries from prospective grad students who found me through the blog. And this year I expect to get at least two papers out of blog posts, or out of collaborations that were spawned by blog posts. That’s two more papers than I would’ve had had I not been blogging.

So in 2012 you can expect more of the same from me, and hopefully more from other folks. Chris Lortie will be starting up the popular “Editor’s Choice” posts again, providing detailed insight into some of the best papers coming out in Oikos each month. And we have some other ideas for how to add some new voices to the blog.

2012 is going to be an exciting year for the journal as well. A new EiC will be coming on board, with some new ideas for both the journal and the blog. And the journal will be moving to the Manuscript Central system for handling submissions. There was a period when Oikos was slow to take advantage of new technologies, but I’d like to think we’re past that stage and that we’re becoming a leader in this area. I don’t know of any other journal in ecology that blogs about such a wide range of topics, as opposed to using a blog largely or exclusively as a way to promote the journal’s own content.

In the comments, we’d welcome your ideas for both the blog and the journal: what topics you’d like to see covered, what changes you’d like to see–or even just a +1 if you think we’re on the right track.