Ask Us Anything: teaching evolutionary ecology, what statistics to learn, and what devices to read on

A while back, we invited you to ask us anything. Here are our answers to our next three questions, from sagitaninta (first two questions) and Matt Ricketts:

  1. Should evolutionary ecology be incorporated into introductory ecology courses, or left for advanced courses?
  2. What statistical tools should ecologists learn in order to avoid “defaulting” to simple, familiar tools like linear regression and t-tests?
  3. Do you read the literature on paper, electronic devices, or both? What tools do you like for this purpose?

Jeremy’s answers:

  1. If by “evolutionary ecology” you mean the broad idea that organisms evolve and are adapted to their environments, then yes absolutely! If you mean something more specific like life history evolution or eco-evolutionary dynamics, then I think it depends on the interests of the instructor and other factors such as how the course in question fits in with the other courses students will be taking as part of their degree programs. Meghan has a related post on what ecological concepts should be included in an introductory biology course.
  2. We have an old AUA post on this. Also remember that you don’t have to do a fancy statistical analysis just because it’s fancy; see Brian’s old posts on statistical machismo.
  3. I read pdfs on my laptop and desktop. I stopped printing pdfs out and taking notes in the margins a number of years ago, probably after most (not all) of my colleagues had. I sometimes toy with the idea of getting a tablet and a stylus to try to replicate the old experience of reading and making marginal notes on paper. But the sad truth is that I no longer read enough papers carefully enough to make that investment worth it.

Brian’s answers:

  1. Yes evolutionary ecology should be taught early. By which I mean the notion of optimization subject to constraints and frequency/density dependence. And specific examples – explaining things like theories of why humans age or why cowbird parasatism is evolutionarily maintained excites students a lot.
  2. I guess I would echo Jeremy’s two links. I will just repeat my one previous answer in that I think we spend way to much energy learning the relatively complex art of mixed models and would be better learning quantile regression, path analysis and machine learning, and even just plain old getting better at simple multivariate regression.
  3. Depends how seriously I am going to read it. If it is a skim I’ll do it on a PDF. If it is a paper I am reviewing or want to deeply understand for my research, I still print it out and use margin notes. This is perhaps a function of my spending most of my time on a so-called ultraportable with a 12″ screen but also for the fact that nothing substitutes for hand writing notes in the margins for me (and studies have shown that hand writing improves memory over typing). When I am travelling I do sometimes just use PDF margin notes on screen, but I don’t feel like I do my best job. When it is a paper I am co-authoring I dive straight into it in Word.


5 thoughts on “Ask Us Anything: teaching evolutionary ecology, what statistics to learn, and what devices to read on

  1. My impression is that Evolutionary/Behavioral Ecology has been pushed out of the list of topics included in beginning ecology classes [ perhaps beginning evolution classes too]. For example Molles first edition did not include life histories….I got him to put a new chapter on it in later editions.
    I think some of E/B ecology is quite exciting to many students,… life histories (including aging, age at maturity, & so forth), sex ratio, male/female differences ( sexual selection), foraging, kin-selection & altruism: having introduced them into my own introductory ecology classes over many yrs I vote that they work quite well to capture student’s interest, and represent some of ecologies’ and evolutions’ great scientific success stories. GC Williams took the position [ over drinks , on several occassions] that sex ratio theory ought to be taught in introductory biology classes, if simply to introduce a near universal rule [ 1:1] that is WIDELY known by the students and for which most students probably never saw as a puzzle to be understood.
    So I agree with Brian. In general, if my examples differ from his.

  2. Thanks for the answers!

    What I mean is something more specific like eco-evolutionary dynamics indeed. I kinda think it may confuse the students to push them to integrate all the things in the intro course. But the examples in behavioral ecology is indeed nice for undergraduate course. I’ve read all those posts, but I guess I have to dig deeper on Meghan’s post for this.

    For the fancy stats, it is difficult when the reviewer asked for it, which is perhaps because I didn’t write a good justification for my statistical methods. After reading those posts and the comments, I came to think that using statistics are more subjective than I thought!

    • In the preface to the 1992? reprinting of his great book ADAPTATION & NATURAL SELECTION GC williams says that sexratio evolution and David Lack’s clutch size hypothesis were the turning points in his realization that individual/gene selection was the correct way to think. Simple versions of these arguments require only algebra, and careful logic. they also have the advantage that they are the original ESS [ game theory] and constrained optimization exemplars , and thus capture the essential 2 forms of normalizing selection used by behavioral ecologists.

  3. 1. pass
    2. to me it’s less what to learn than how to learn it. Jeremy suggested ANOVA and regression in the linked post. Much of my generation was taught by Sokal and Rohlf or Zar or similar, which emphasized the ANOVA table. As a consequence most papers report F and P values (and sometimes the whole ANOVA table!). I just don’t see any value to an ANOVA table. What I find more valuable is the coefficients of the model and their SEs (or CI or t or P, whatever your preference). But then, if the emphasis is on the coefficients and their SEs, then an analysis with categorical X is just a regression. Or better yet, replace both regression and ANOVA with “linear model”. So yes classical ANOVA is important for reading the literature but it constrains the way we learn statistics and analyze data. I am still working on cleaning my ANOVA way of thinking with its endless debates on fixed v. random effects and type I vs II vs III SS. Learning and doing stats with the linear modeling mindset makes the path to contrasts (in ANOVA!), generalized linear models and linear mixed models/multi-level modeling much easier. What I like about R is that it encourages the linear modeling mindset – and abandoning the regression vs ANOVA mindset. For example, it is the coefficient table but not the ANOVA table is reported in the summary.
    3. I like having instant access to PDFs so I use Zotero, which is free and synced on all my devices (Papership on an iphone or ipad). When I review, I take notes with a text editor (free BBedit) and just refer to the line number). This also works when simply studying or commenting on a paper for my own work. I can keep the notes with the pdf in zotero. I was going to say printing to paper is not very green but that may be an ignorant statement. Maybe if all my electricity came from solar or wind I could state that!

    • Jeff I completely agree with you on #2. And fortunately for us I guess, I think the whole field of ecology has moved strongly in that direction. Whether that was driven by popularity of R or helped to drive R’s popularity is an interesting question. Something people often don’t appreciate is that the linear model approach is driven by the ability of computers to handle the computations required in the maximum-likelihood approach that lies behind it. But at least the basic approach has moved in the right direction towards linear models.

Leave a Comment

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

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

Google photo

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

Twitter picture

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

Facebook photo

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

Connecting to %s

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