News & Views | Published:

Evolutionary biology

Life imperfectly imitates life

Nature volume 483, pages 410411 (22 March 2012) | Download Citation

Some species evolve to resemble another species so as to protect themselves from predation, but this mimicry is often imprecise. An analysis of hoverflies suggests why imperfect imitation persists in the face of natural selection. See Letter p.461

One hundred and fifty years ago, the English naturalist Henry Walter Bates1 discovered a phenomenon that he regarded as “a most powerful proof of the theory of natural selection”. Charles Darwin and Alfred Russel Wallace, who had proposed natural selection as the engine of evolution only a few years earlier, agreed. Indeed, Darwin2 considered Bates's manuscript to be “one of the most remarkable and admirable papers I ever read”.

Bates had uncovered a spectacular type of adaptation, now known as Batesian mimicry3, in which a species (the mimic) that is edible to predators evolves to resemble a conspicuous, inedible species (the model) that predators avoid. These lookalikes are selectively favoured, Bates argued1, because predators confuse them for the inedible model and thereby avoid them. This idea was so persuasive that Batesian mimicry is still widely used to illustrate the power of natural selection for producing adaptations3,4. However, mimics are often poor replicas of their model, and this inexact mimicry poses a challenge for evolutionary theory3. On page 461 of this issue, Penney et al.5 discuss possible explanations for why mimics are frequently imprecise.

As Penney and colleagues describe, several non-mutually exclusive hypotheses have been proposed to explain imperfect mimicry. They include: the 'eye-of-the-beholder' or sensory-limitation hypothesis, which asserts that imprecise mimicry is an artefact of human perception; the 'jack-of-all-trades' or multimodel hypothesis, which posits that imperfect mimics persist because they are under selection pressure to resemble more than one model; and the 'kin-selection' hypothesis, which asserts that imperfect mimicry is maintained because it provides benefits for genetically related individuals. Specifically, if mimics are imperfect, then predators will be more discriminatory and occasionally spare a mimic by mistaking it for a model. This will also spare relatives that share the same imperfect trait.

Another theory to explain imperfect mimicry is the 'relaxed-selection' hypothesis, which holds that there is little selective benefit in refining mimetic resemblance beyond a certain point, for example if the model is common or especially noxious. Finally, the 'constraints' hypothesis proposes that either imperfect mimics lack the genes to produce refined mimicry, or that a selective trade-off exists between predator-mediated selection favouring more precise mimicry on the one hand and other agents of selection (such as mate choice) favouring divergence on the other.

Until now, no study had rigorously evaluated these multiple hypotheses within a single system. Penney and colleagues5 did precisely this by comparing the degree of mimicry in dozens of species of hoverfly. Mimicry is a dominant feature of this large family of harmless Diptera (Fig. 1). About 5,600 species have been described, at least a quarter of which resemble stinging wasps and bees (Hymenoptera)6. Although some hoverflies are remarkably accurate mimics, converging both morphologically and behaviourally on their hymenopteran model, many others seem to bear poor resemblance6. This variation in mimetic fidelity makes hoverflies ideal for studying imperfect mimicry.

Figure 1: A fly in wasp's clothing.
Figure 1

The harmless marmalade hoverfly, Episyrphus balteatus, shown here, resembles stinging wasps. Although some species of hoverfly closely resemble wasps or bees, others are poor lookalikes. Penney and colleagues5 show that the lower selection pressures that come with decreasing abundance and small body size may provide the explanation for how such imprecise mimicry can persist.

The authors report that birds (which are hoverfly predators) and humans seem to agree on the resemblances between hoverflies and hymenopterans. Thus, in this case, the eye-of-the-beholder hypothesis does not hold. Furthermore, Penney and colleagues' statistical analysis reveals that the mimics' characteristics do not fall somewhere between those of different models, so the jack-of-all-trades hypothesis receives no support either. The kin-selection hypothesis predicts that mimetic precision should decrease with an increasing abundance of mimics, but the authors observed the opposite trend. In fact, this finding is consistent only with the constraints and relaxed-selection hypotheses. In addition to predicting that mimics' precision should increase with their abundance, these two hypotheses predict that mimetic precision should increase with the body size of the mimic, and this is exactly what the authors find.

To understand how the relaxed-selection hypothesis applies when both the abundance and body size of mimics increase, let us consider the problem of discriminating between tasty mimics and nasty models from the predators' perspective. If not all mimics are equally deceptive, a predator must take risks when attacking its prey — if it strikes a mimic, it will reap more reward than regret. When mimics are abundant, the likelihood of attacking a model by mistake decreases. In such conditions, predators should be willing to sample all but the best mimics, which will push selection towards more precise mimicry.

By the same token, when mimics have a larger body size, their potential benefit to predators is greater (larger prey are generally more profitable for predators), so there will be greater selection pressure on them than on their smaller counterparts to become better mimics. Although other studies have found that mimetic precision increases when mimics are abundant7, Penny and colleagues' demonstration that this pattern also holds when mimics increase in size is an elegant affirmation of an old idea.

We still do not know whether hoverfly mimicry is imprecise because of an absence of selection for refinement once mimicry is 'good enough' (as in the relaxed-selection hypothesis), or whether there is active selection pressure against further refinement because of the costs of producing better mimicry (as in the constraints hypothesis). For example, constraints may be imposed by competition between mimics and their models (over shared resources8 or reproductive opportunities9), which would favour divergence between them and, hence, imprecise mimicry10. Future studies are needed to tease apart these two hypotheses.

As we celebrate the 150th anniversary of Bates's discovery of mimicry1, the topic continues to fascinate the public and scientists alike3,4. Penny and colleagues' findings help us to understand why selection sometimes produces precise mimicry, but often does not, and further clarification of this puzzle promises to provide additional insight into the evolutionary process.


  1. 1.

    Trans. Linn. Soc. 23, 495–556 (1862).

  2. 2.

    , & (eds) Evolution: Selected Letters of Charles Darwin 1860–1870 (Cambridge Univ. Press, 2008).

  3. 3.

    , & Avoiding Attack: The Evolutionary Ecology of Crypsis, Warning Signals & Mimicry (Oxford Univ. Press, 2004).

  4. 4.

    Dazzled and Deceived: Mimicry and Camouflage (Yale Univ. Press, 2009).

  5. 5.

    , , , & Nature 483, 461–464 (2012).

  6. 6.

    in Insect Evolutionary Ecology (eds Fellowes, M. D. E., Holloway, G. J. & Rolff, J.) 231–288 (CABI, 2005).

  7. 7.

    & Proc. R. Soc. Lond. B 274, 1955–1961 (2007).

  8. 8.

    et al. Nature 469, 84–88 (2011).

  9. 9.

    & J. Evol. Biol. 21, 749–760 (2008).

  10. 10.

    & Curr. Zool. (in the press).

Download references

Author information


  1. David W. Pfennig and David W. Kikuchi are in the Department of Biology, University of North Carolina, Chapel Hill, North Carolina 27599, USA.

    • David W. Pfennig
    •  & David W. Kikuchi


  1. Search for David W. Pfennig in:

  2. Search for David W. Kikuchi in:

Corresponding authors

Correspondence to David W. Pfennig or David W. Kikuchi.

About this article

Publication history



Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing