Saturday, February 16, 2013

Eco-evolutionary semantics

What is in a name? Presumably a name should be clear and unambiguous, thereby nicely delineating what it does and does not include. After all, we should call a spade a spade, right? Or is it that simple? Perhaps a spade is instead a shovel or a trowel or a digger or a spud or a geotome. “Now wait a minute,” you might say, quoting Wikipedia: “the words spade and shovel should be held in contradistinction (piercing and digging [spade] versus scooping and moving [shovel]).” “Sure, but,” I might retort, quoting the next line in Wikipedia: “Natural language does not widely follow these prescriptions; it more often treats spade and scoop as contradistinguished subsets under shovel.” Thus, even if a shovel is not always a spade, a spade is often a shovel and, regardless, we can call either the other and still get the job of digging done.

What, then, are eco-evolutionary dynamics? I would venture the following definition: eco-evolutionary dynamics are interactions between ecology and evolution that play out over contemporary time scales, with “contemporary time scales” intended to represent time scales ranging from years to centuries. (See more details here.) This definition is intended to be inclusive, thus providing an umbrella framework for understanding how ecology and evolution interact on short time scales. But it doesn’t satisfy everyone – as I will elaborate below.

Cruising the Netherlands at dusk.

For the past week, I have been at a workshop on “Eco-evolutionary dynamics in a changing world” held in Leiden, Netherlands, at the Lorentz Centre. (Much to the surprise of almost everyone present, we learned on the first day that this was not the guy with the geese – that would be Lorenz.) Organized by Stephanie Jenouvrier, Thomas Reed, and Marcel Visser, the workshop had plenary talks, break-out discussion sessions, and group discussions. These activities were interspersed with ample time for social interaction – call them break-out drinking sessions if you will. Antics ensued.

At one dinner, a group of us decided to form an “Institute of This Table (ITT)” with each of us holding a research chair endowed by one of the others. I can’t remember all the resulting chairs, but I do recall that Richard Gomulkiewicz christened Nelson Hairston the “Dick Chair of Cryptic Dynamics” and Mike Kinnison christened himself the “Hendry Chair of Paleontological Genomics.” (I later gave him two Euros.) Later that night, we ended up at an English bar with a truly amazing collection of Scotch whisky. I tried these out on my palate while Dick tried out his Dutch pick-up lines on Mike Kinnison – stay tuned for the youtube video. On another night we took a cruise through the canals around Leiden. After much debate, Chris Thomas and I agreed to reunite for dinner 36 years hence so that he (I) could gloat over me (him) at having proven (disproven) the predictions from his 2004 Nature paper that “on the basis of mid-range climate-warming scenarios for 2050, that 15–37% of species in our sample of regions and taxa will be ‘committed to extinction’.”

The ITT.

For the first two days of the meeting, most of the arguments – and certainly the most animated ones – were about what eco-evolutionary dynamics are and are not. Here is a restrictive sampling of the many uncertainties.

1. If phenotypic differences have an effect on some ecological variable (such as population size or species richness or nutrient cycling), but you haven’t isolated or confirmed that these effects have a genetic basis, then are you studying eco-evolutionary dynamics? The excluders might argue that, if you haven’t demonstrated evolution, then you can’t say you are studying eco-evolutionary dynamics, and that you would be better off saying eco-phenotypic dynamics or some such. The includers might argue against this precision on several grounds. First, if we go down this slippery slope, then the field of evolution quickly gets much smaller. As Mike Kinnison pointed out, we would have to kick out all of the paleontologists given that none of them has yet proven that the changes they study have a genetic basis. Second, we often know that phenotypic differences have a strong genetic basis even if we haven’t proven it in a particular experiment. This is why the eco-evolutionary pantheon includes those great studies of the ecological effects of alewife divergence, guppy divergence, and stickleback divergence – even though all of those studies used wild-caught fish for the experiments. As some of the perpetrators of those studies were present (Ron Bassar and Mike Kinnison), I was tempted for a vote on whether or not we should – right then and there – drag them out of the room and throw them off the eco-evolutionary island – but then I realized I wouldn’t be far behind. Third, phenotypic plasticity evolves, and so plastic responses in a given generation are the product of evolutionary change in a past generation. In this sense, plasticity is the gift that evolution in one generation keeps giving to future generations. Fourth, it is phenotypes and not genotypes that interact with ecology; genotypes interact with ecology only indirectly through their effects on phenotypes. In short, phenotypes – even without information on genotypes – are an integral part of the eco-evolutionary dynamic framework. (Just as every modern evolution text book has a chapter or major section on plasticity, and numerous examples that focus only on phenotypes, so too should any eco-evolutionary textbook. Of course, investigators should make every attempt to explore the genetic basis underlying phenotypic effects and shouldn’t assert genetic change until they have explicitly confirmed it.)

2. If the ecological variable of interest does not change, then are eco-evolutionary dynamics occurring? The excluders might argue that if no change has occurred, then neither have any dynamics – by definition. The includers might reply that stability itself is likely the product of ongoing evolution. Such cryptic eco-evolutionary dynamics could occur in several basic ways: (a) changes in genotypes cause changes in phenotypes that cause stability in ecological variables, (b) changes in genotypes cause stability in phenotypes that cause stability in ecological variables, and (c) selection causes stability of genotypes that cause stability in phenotypes that cause stability in ecological variables. An example of (a) comes from the work of Nelson Hairston’s group, where changes in the frequencies of two different clones of algae that are differentially susceptible to a predator are necessary to maintain algae population size. An example of (b) comes from a number of studies showing “counter-gradient variation”, such as when environmental conditions (e.g., low food) that decrease growth are offset by evolutionary changes that maintain growth. The phenotype will here remain unchanged only because of evolution. Support for (c) comes from the realization that selection is constantly weeding out maladaptive genetic variation that arises owing to mutation, gene flow, and genetic drift. Without this selection, fitness would decline and the population would soon decline too. Stability is thus driven by the continual action of selection. For all of these reasons, it seems clear that the absence of ecological – or even phenotypic or genetic change – does not mean that eco-evolutionary dynamics are absent.

3. If you haven’t demonstrated a FEEDBACK (e.g., a trait influences an ecological variable that then influences selection on that trait), then are you studying ecological evolutionary dynamics? The excluders might argue that only true feedbacks should be considered dynamics but the includers feel this criterion as too restrictive. First, it is very hard to demonstrate feedbacks even when they are present; doing so usually requires manipulative experiments. Second, feedbacks of one sort or another likely occur in nearly all instances – at least in the broad sense. By this I mean that changes in any trait in some focal organism will likely influence some other organism in the ecosystem and thus have ecological effects. Overall, I would argue that eco-evolutionary dynamics should include all interactions between ecology and evolution, even one-way effects, whereas the term eco-evolutionary feedbacks should be reserved for situations where a feedback (ecology to evolution to ecology; or evolution to ecology to evolution) has been explicitly demonstrated or is explicitly hypothesized.  

4. If the interactions between ecology and evolution are happening only on long scales (e.g., millennia) are they eco-evolutionary dynamics? Hanna Kokko argued that such interactions often occur on quite long time scales, and this is certainly the case. Think of all the ecological and evolutionary changes that depended on the much earlier evolution of photosynthesis.  Although I tend to be an includer in all of the above debates, I will caucus with the excluders. The Oxford Dictionary states that dynamics is “a process or system characterized by constant change, activity, or progress”. This would seem to me to exclude long time scales but, oops, this definition is for its use as an adjective. As a noun, the definition is “a force that stimulates change or progress within a system or process.”  This one seems harder to invoke in order to exclude long term effects. And – after all – the Oxford Dictionary clearly calls only a spade a spade: “a tool with a sharp-edged, typically rectangular, metal blade and a long handle, used for digging or cutting earth, sand, turf, etc.” So do I really have any solid reason to exclude longer time scales from the umbrella eco-evolutionary dynamics? I guess I would hand-wave that interactions between ecology and evolution on long time scales have long been recognized but it is only recently that they have been appreciated on short time scales – and it is this latter appreciation that has driven the emerging field of eco-evolutionary dynamics. This is why definitions of eco-evolutionary dynamics have tended to focus on short time scales and this distinction currently delineates the field. And, after all, who wants to have to say “contemporary eco-evolutionary dynamics” (although CEED sounds cool) or, Darwin forbid, “rapid eco-evolutionary dynamics” (REED?).

So where do these semantics leave us? I would like to suggest that eco-evolutionary dynamics is a FRAMEWORK for understanding interactions between ecology and evolution on contemporary time scales, and it therefore needs to be inclusive. Or, as Hal Caswell pointed out, eco-evolutionary dynamics can be considered a social identifier that places one comfortably into the group of researchers applying evolutionary thinking in ecological studies. I would suggest that this group is inclusive with respect to the first three debates (and many others) but is (arguably) exclusive for the fourth.

These semantic issues largely subsided toward the end of the meeting given that we really could communicate quite clearly without going to further trouble. This left the rest of the meeting to argue over more substantive issues, like whether or not evolutionary (or even ecological) models can reasonably predict what will happen in the future, how important is evolution in nature, can Lande Land accommodate density- and frequency-dependent selection , which deep-fried Dutch croquette was the best (or worst), and how could Heineken simultaneously be the worst beer and the best wine in the same restaurant.


  1. I, for one, am in favour of switching to the REED acronym...


    I like your explanation of what can or should be included in the scope of eco-evolutionary dynamics, though I argue that the term should be more than a “social identifier”. “Words matter,” as my friends in the humanities like to point out. Without clarity we end up with confusing and nearly meaningless terms: take the degraded meaning of “keystone species” as an example.

    That comment aside, your final “Darwin forbid” raised something I’ve always wondered about. Did Darwin really think that evolution had to be ponderously slow, or did he just argue that evolution had plenty of time to produce the diversity of life that he was explaining? After all, he started out the Origin discussing the enormous variation that selective breeding had produced in pigeons in a short amount of time. As a supremely keep observer of nature, he surely understood what this might mean for rate of change in natural populations as well. Invoking the great spans of time available for evolution no doubt made it easier to argue for the role of natural selection in producing biological diversity, but did he honestly think that it had to take a long time? In rereading some of the Origin recently I came across the following passage (Darwin 1859, p. 183): “But it is difficult to tell, and immaterial for us, whether habits generally change first and structures afterwards; or whether slight modifications of structures lead to changed habits; both probably often change almost simultaneously.” Since by “habits” Darwin meant behavior (which can be ecological in nature), and by “structures” he meant evolving traits, to me his final phrase sounds remarkably like eco-evolutionary dynamics. Steve Ellner, Monica Geber and I quoted it at the start of our 2011 paper in Ecology Letters. Maybe I am just hoping that this is an indication that Darwin understood that adaptive evolution in natural could be fast. It is possible that he thought that behavior (“habits”) could change only slowly, but this seems harder to argue. Perhaps Darwin scholars have already explored his writings and voluminous correspondence to get a sense of his thinking, in less guarded moments, about the rate of evolution. If not it would seem a topic worth exploring.

    1. I agree with your comment on Darwin. Here is something from my forthcoming book:


      Darwin (1859) seemingly argued that natural selection was very weak and evolution very slow: “we see nothing of these slow changes in progress until the hand of time has marked the long lapse of ages” (p. 84) and “she can never take a leap, but must always advance by the shortest and slowest steps” (p. 194). Despite a few exceptions that will be noted below, this view of evolution as a sedate force prevailed for the next hundred years and more. Before considering the sea-change that followed, it is worth nothing that Darwin’s “slow” evolution could be closer to our “rapid” evolution than is normally thought. For instance, Darwin (1859, p. 120-123) describes the origin of 14 new species, as well as considerable variation within each, over less than 14,000 generations. Although we now that evolution works even more quickly than this, Darwin clearly realized that evolution could accomplish quite a bit on modest time scales.