Evaluating the fitness consequences of signalling…

Robert Kurzban, over at the Evolutionary Psychology blog, has a post up which caught my attention because it deals with an area of behavioral ecology that I happen to know a little bit about:  signalling.  My master’s degree was done on animal aggressive communication models, under the supervision of Pete Hurd.  So I was intrigued to see what Robert had to say.

The post itself was fairly underwhelming for me, because Robert only seems interested in using a (somewhat overstated) terminological dustup over signals versus cues.  But about halfway through the post, he decides to throw a rather large stone in a surprisingly glass house:

Let’s turn to the substance of the matter. Maynard-Smith and Harper (2003) define a signal as “any act or structure that alters the behaviour of other organisms, which evolved owing to that effect, and which is effective because the receiver’s response has also evolved” (p. 3).

Their first example is distinguishing two ways a stag might make another stag retreat: push him or roar at him. Pushing, they argue, isn’t a signal – it does alter the behavior or the other organism, but the response, moving backwards, didn’t evolve as a response to pushing – it’s simply a physical consequence. In contrast, retreating in response to a roar, they argue, makes the roar a signal. Roaring evolved because retreating from roars is an evolved response, the argument goes.

Note how casually Maynard-Smith and Harper make this strong claim. Labeling the stag’s roar a signal is an adaptationist claim, that the behavior in question has a function, in this case, conveying information – signaling – to a rival which, in turn, is useful in the context of intra-sexual conflict. I find it worthwhile to reiterate, at the risk of undue repetition, that this illustrates how biologists routinely make adaptationist claims based on observed patterns of behavior, rather than measuring fitness consequences, heritability, and the like.  [Emphasis mine].

I was astounded by this claim for two reasons.  First, an evolutionary psychologist is telling biologists that they don’t measure fitness consequences?  When was the last time an evolutionary psychologist did an experiment in which they manipulated a trait and measured fitness consequences, i.e. survival and / or babies made as a result?  I would love to be on the ethics committee that processes that application, I really would.  Every EP study that I have ever read which deals with fitness in any way is either a correlational study (e.g. using historical birth record data or doing a cross-sectional analysis) or uses some proxy  for fitness, like asking women to rate the attractiveness of male body odours that they sniffed from used t-shirts to assess the “fitness consequences” of MHC (major histocampatibility complex) preferences.  Don’t get me wrong, I find a lot of this work interesting, but to criticize biologists for not assessing fitness consequences is quite the santicmonious move.

In behavioural ecology we often do as the MHC researchers do and use proxies for fitness, by which I mean that we measure a trait that we argue (or assume) is correlated with fitness.  For instance, in foraging research, we assume that food intake rates will correlate with fitness and so suggest that the adaptive value of a behavioural trait can be measured – at least indirectly – by manipulating the trait in some way and seeing how that impacts food intake.  Would we prefer to measure fitness directly?  Sure we would!  The reasons we use proxies usually boil down to the difficulty, or downright impossibility, of measuring fitness in that way.  Primatologists, for instance, are going to be about as likely to perform such direct fitness studies as evolutionary psychologists are, and even those of us who work with smaller and easier to handle animals are going to find such research challenging.

But that brings me to my second objection, which is that biologists do measure fitness consequences (and heritability, and the like, but I’ll focus on fitness here).  Some examples:

–  The obvious:  twisting Drosophila into various shapes and seeing the consequences (fitness or otherwise) of that is practically an industry by now.  It took me about four seconds and a single Google search to find a nice-looking study on sexual selection and fitness outcomes in Drosophila by Promislow et al. (1998), and the related links lead to a flood of more recent research.
–  The awesome:  many examples are found in sexual selection research, for obvious reasons.  Frogs have some good examples of this, like work on the Australian frog by Jeremy Robertson that showed that female choice of male calls was adaptive by showing the consequences to female clutch fertilization by a mismatch between call and male body size.  And I’m pretty sure that Mike Ryan, who has done a lot of incredible studies on the signalling system in túngara frogs, would be surprised to find out that biologists don’t assess fitness consequences.

–  The vaguely frightening:  ever seen hermit crabs fight?  It’s a bit scary.  Check out this video from the Royal Society in which one crab thoroughly kicks the snot out of another and then steals the other guy’s house:

Research on the fitness consequences of these fights goes back decades, both intra- and interspecifically.  The signalling system used by hermit crabs (like shell rapping) is also a fascinating area of study.

–  The avian:  birds are a common target for this, too.  A particular example I like comes from an area close to home, the zebra finch.  Remember the assumption I was talking about above, using foraging intake rates as a proxy for fitness? William Lemon decided to test this directly, and so he manipulated the feeding rates of four zebra finch populations to determine the adaptive value of energy maximization (and its suitability as a proxy for fitness) by directly measuring survival and reproduction in the populations.  Show me the EP study that does that.

This is just a short list cobbled together from what I can think of off the top of my head and some quick Googling, and I even confined myself to just studies looking at reproductive benefits;  many more have done work on the survival component of fitness, and to cover even a fraction of those would require an inconveniently long book.  Even the examples I’ve chosen are slanted towards behaviour, for obvious reasons, and people skewing towards the genetics side of things have done a lot more work on these types of questions than we have in behavioural ecology.

When I began my undergraduate career in psychology, I was attracted to evolutionary psychology.  I felt then, as I do now, that the core concept of taking evolution’s effects on homo sapiens into account is true and needs attention.  It’s what inspired me to shift to biology, so that I could study the tools of biology and bring them back to the sort of questions that EP studies.  And if Robert’s feelings are shared by other current practitioners of EP, I can’t help thinking that the field needs more people who will do the same.

Tagged ,

2 thoughts on “Evaluating the fitness consequences of signalling…

  1. Rob Kurzban says:

    Thanks for your remarks on my post. I think I was unclear, and I would like to clarify. You highlight and remark on my claim that “this illustrates how biologists routinely make adaptationist claims based on observed patterns of behavior, rather than measuring fitness consequences, heritability, and the like.” You take this to be a criticism of what biologists do, but that was not my intended meaning at all. Quite to the contrary, my intention was to applaud the application of the logic of adaptationism, whether by biologists or psychologists. The reason that I pointed this out – and I probably should have been explicit about this – was that some have criticized evolutionary psychologists for failing to measure fitness consequences and heritability. (I have recently written about this issue.) As you correctly point out, evolutionary psychologists, rather than measuring fitness, make inferences from behavior, as a general matter, just as behavioral ecologists do. This is my point, the evolutionary psychologists apply the same logic biologists do. It had nothing whatever to do with a criticism of the way that biologists conduct their work. And certainly I don’t deny that biologists measure fitness. My claim is a much narrower one, that biologists draw on other sorts of evidence to make claims of function.

    In any case, I’m not quite sure why my wording led you to infer I was criticizing biologists, but I want to emphasize that that was not my intention. I did not say that biologists don’t ever measure fitness; I was pointing out that they make adaptationist claims without doing so, just as evolutionary psychologists do. My broader point is that the same logic applies to humans as to other species, a point about which I think we agree. In fact, other than my not writing clearly, I think we agree on most of the points at issue here, and again I apologize for the lack of clarity on my part.

  2. stevenhamblin says:


    Hmm. Well, I guess I took that the wrong way, and I’m sorry if I mis-characterized your intentions. In truth, I probably took it that way because I’m not much of a strict adaptationist myself. I would prefer the direct assessment of fitness as the gold standard, and inferring from behavioural measures to be a necessary evil subject to hard scrutiny. It honestly pains me that we don’t have the resources or techniques to do more direct measurement of the effect of trait manipulations on fitness, and it worries me that EP *never* gets the chance to do it at all.

Leave a Reply

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

WordPress.com Logo

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

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