Over at Nothing in Biology Makes Sense, newly-minted PhD evolutionary biologist David Hembry reflects on the biggest changes in evolutionary biology and ecology since 2005. It’s a thoughtful piece, reflecting on some less-noted aspects of widely-noted trends. For instance, it’s not just the increasing availability of sequence data that makes synthesis and reanalysis of other people’s sequence data attractive, it’s also the fact that it’s cheap to do (particularly important in an era of rising fuel costs and increasing competition for funds). The same could be said of any database-based work, really, and also of theoretical work and laboratory microcosm work. It will be interesting to see how patterns of training, hiring, and publication shift in decades to come*, and if there aren’t frequency-dependent forces that will limit how far these directional trends can go (At some point, will really good field skills become highly prized precisely because of their scarcity, while good bioinformaticists/meta-analysts/theoreticians/programmers/etc. will be a dime a dozen?)
David also identifies some less-noted trends, such as the increasing focus of evolutionary biologists on “field model organisms” like sticklebacks and anoles, and how this poses problems of system choice for grad students who want to go on to academia. Do you choose the same model system as everyone else, thereby making it easier to ask big questions (after all, there are good reasons why sticklebacks and anoles are model systems!), but harder to stand out from the crowd? Or do you choose the road less traveled, which might make it harder to address big questions but also really impress people if you succeed? (sounds a bit like the handicap principle…)
Anyway, click through and read the whole thing.
*At least in ecology, there’s not yet much indication of a radical shift towards people publishing data collected by others.
I’ve wondered exactly that re: the relative value of field skills vs. data-jockey skills. Sooner or later, when we can have all the genome-scale data we want and there’s a suite of off-the-shelf software to crunch it, it’s going to matter most who can identify which genome-scale data sets will be interesting.
Yup. As I said back in my very first post on Oikos blog, in the future we’re increasingly not going to be “data limited” (at least not for *certain kinds* of data; that’s actually another issue I should post on at some point), we’re going to be “ideas limited”. There’s going to be an increasing premium on people who can ask the right questions. And certainly, one source of good questions is field-based natural history knowledge.
Speaking from my experience, once you have a lot of data, it’s easy to come up with all kinds of questions that are not necessarily that interesting. The challenge is to determine out of all the things you could do, which ones are actually very interesting. I suspect Jeremy Fox is right that we may be “ideas limited” in the near future, and I think navigating this “thicket of not-very-interesting ideas” is going to be a major challenge for students and their advisers. Given this, one thing I’ve been very impressed by is how well people who use large amounts of free data are at using these data to test hypotheses we are all interested in.
I didn’t state it explicitly in my post, but I also agree that we are only going to be data-limited for certain kinds of data. I look forward to seeing your thoughts on that in a future post.
“Speaking from my experience, once you have a lot of data, it’s easy to come up with all kinds of questions that are not necessarily that interesting.”
Good point. The more data you have to explore, the easier it is to get lost in pointless exploratory analyses.
Re: only being data-limited for certain kinds of data, this is something I’ve hit on in the past in specific contexts (e.g., https://oikosjournal.wordpress.com/2011/10/20/thoughts-on-nutnet/). But I’ve never tried to come up with a list of the specific kinds of data we’re accumulating rapidly, and the kinds of data we aren’t accumulating so rapidly. Might be a fun exercise.