Here is a rant that I have had on and off over the last few years, and a recent discussion in my own group has prompted me to post this on the blog, just to try the argument out on a slightly wider audience….
One approach to studying evolution is to focus on a single organism or group of organisms, and try to understand different facets of their biology, in order to understand processes that must be more widespread. Such a focus is absolutely necessary – we do not have the resources or time available to study all processes in all species, and a deep understanding of ecology, behaviour and genetics is needed to study processes such as speciation. So we need to study some taxa in great detail (and of course I think that the Heliconius butterflies are a great group in which to study adaptation and speciation). However I hate the term ‘model species’ for these taxa, and I don’t think that the term is helpful in evolutionary biology. The term ‘Model’ comes from the medical literature, where, for example, a ‘mouse model’ of diabetes would be an experimental procedure or strain of mice that has been developed to mimic the human disease. The ‘model’ is just that – a tractable, toy version of the real thing (this question of how useful such models actually are for developing disease treatments is a separate issue which I will not tackle here). We use the term in a similar way to describe mathematical models that are simplified versions of the real world, that can be used to infer general processes.
My problem is where the term has taken on a new meaning, to refer to a particular species in which biological processes are studied. The intended implication seems to be that whatever is discovered can be inferred to be more generally true across the tree of life. This approach has of course been rather extraordinarily successful in cellular biology, where basic cellular processes uncovered in the fruit fly have proven crucial in understanding cancer in humans, for example. But as evolutionary biologists surely we are interested in diversity, so the assumption that processes can be inferred from one group of organisms to many others seems deeply problematic to me.
Some of the processes that we uncover in one group of organisms will turn out to be rather unusual, whilst others will be more widely shared across many organisms. There are many examples that illustrate this, but one of my favourites is the circadian clock. This is a molecular mechanism that regulates the daily rhythm of animals. Studies in mice and fruit flies seemed to suggest that there were major differences in the way that the clock functioned between insects and vertebrates. However, when the circadian clock was studied in a butterfly, it was realised that in fact the fruit fly is unusual. The fruit fly has lost components of the clock, that are otherwise shared between mice and butterflies (Zhu et al., 2008). This suggests that most insects are probably much more like vertebrates in this particular aspect of their biology than was thought from studying the fruit fly.
So even for basic cellular processes that are shared across the animals, such as the circadian clock, we can be seriously misled by studying one or two ‘model’ species. This must be inherently more true of evolutionary processes such as speciation – where some of what we learn from one group of organisms will turn out to be generally true, while other processes will turn out to be unique. Whilst as biologists we certainly hope that we can draw general conclusions from our work, the reality is that some aspects will turn out to be rather arcane and specific. Only once we have studied many different kinds of organisms will the general patterns become clear. Of course it is this diversity of life that is the primary fascination of the evolutionary biologist, so it is best not to get too downhearted about this. Let’s celebrate the diversity of organisms being studied, rather than try to compete to be the best ‘model’ system.
Some people have argued that the term is so embedded, that perhaps we shouldn’t worry and live with it. But I think it is misleading, because it gives a false sense of security. We really need to go out and test whether our conclusions derived from one group of organisms are general to the wider diversity of life. We shouldn’t hide behind the term ‘model’ organism to give us a false sense of complacency about the generality of our findings.
Finally, many people have said that they like the term simply because it is useful as a handle to describe the systems that we study intensively, but actually it isn’t hard to avoid it and say more straightforwardly what you actually mean. So instead of ‘Heliconius is a model for the study of speciation‘, why not just say ‘speciation has been widely studied in Heliconius‘ – see it uses less words and states the case more succinctly…
Okay, rant over for the moment, just don’t get me started on ’emerging model systems’…
Zhu H, Sauman I, Yuan Q, Casselman A, Emery-Le M, Emery P, Reppert SM. 2008. Cryptochromes Define a Novel Circadian Clock Mechanism in Monarch Butterflies That May Underlie Sun Compass Navigation. PLoS Biol 6: e4.