Evolutionary biology

Communication and speciation

An electrifying evolutionary radiation has evidently occurred among elephant fish in Africa's Ivindo basin. An implication is that open niches for communication can result in species diversification.

Groups of organisms that have diversified to produce species adapted to a variety of ecological niches have attracted increasing attention in recent years1,2,3,4. A widely held view is that such adaptive radiation is triggered by the availability of under-exploited resources or the absence of predators, as may occur after colonization of an island or a mass extinction1,5,6. Such ecological opportunity allows organisms to explore new ecological niches, the result being the evolutionary flowering seen in radiations such as those of Darwin's finches, Hawaiian honeycreepers, rift-lake cichlid fish and Caribbean anole lizards. However, adaptive radiation involves not only ecological divergence, but also divergence into different species. Ecological opportunity is widely implicated as the prompt for speciation, as well as for ecological divergence, but this explanation does not account for species-rich groups that exhibit little ecological differentiation. How exceptional species diversity is produced in such cases has long been discussed1,7.

Writing in The American Naturalist, Arnegard et al.8 propose that, in a situation analogous to ecological opportunity, groups that communicate using a signal modality that is not used by other species, and is not detected by predators, may diversify in the signals used to communicate species identity. They may thus speciate to a much greater extent than species that have to contend with a more occupied signal space and more attuned predators.

Arnegard and colleagues investigated the evolutionary radiation of African mormyrid fishes, commonly known as elephant fish because of their proboscis-like snouts (Fig. 1). To a casual observer, elephant fish seem bland — they have no flashy colours, a mysterious life history and little morphological diversity. However, like many weakly electric fish, elephant fish have for decades intrigued neuroethologists because of their ability to communicate by sending and receiving electric-organ discharges (EODs); distortions of their self-generated EODs are also used to locate prey and to navigate their murky habitat. These signals are part of courtship displays, and the fish can modulate the frequency of the EOD during social interactions. Moreover, EOD waveforms differ greatly among species, allowing the use of EODs as species-recognition signals.

Figure 1: The elephant fish Campylomormyrus rhynchophorus.


Mormyrid species such as this were the subject of Arnegard and colleagues' study8. The length of the specimen is about 14 cm.

The Ivindo basin in west–central Africa is known for its rich diversity of elephant fish; up to 20 species may be found at a single locality. Such high diversity raises the question of what factors promote species diversity, particularly when most of the species are ecologically and morphologically very similar. Arnegard et al. tackled this question by combining decades of research on the evolution of EODs with state-of-the-art phylogenetic comparative methods, which allowed them to infer the contributions of ecology, morphology and signal modality to the formation of species assemblages.

This window onto the past revealed two unexpected results. First, species-specific EODs have evolved at a faster rate than have morphology, size and feeding ecology. Second, EODs diverged during early stages of the radiation, whereas morphological and ecological traits diverged later. The authors' interpretation of the results is that the ability of the fish to enter an open communication niche triggered species diversification, which was followed by a lesser degree of ecological and morphological divergence. Furthermore, they conclude that the empty communication niche allowed for species-specific signal divergence to occur in the absence of morphological or ecological divergence, resulting in species assemblages in which communication is the major axis of diversification.

But why does diversification in signal structure occur? A popular hypothesis invokes ecological speciation, in which adaptation to different ecological niches produces, directly or indirectly, divergence in species-recognition signals, leading to reproductive isolation and, hence, speciation. Arnegard et al. suggest that this is unlikely for species inhabiting the Ivindo because there is no evidence of divergence in EOD waveform as a result of ecological conditions. A second possibility is that genetic drift has occurred — perhaps populations randomly diverged in signal structure, eventually becoming so different that they no longer recognize each other. Although possible in principle, this hypothesis is difficult to test and currently not in fashion.

A third explanation is that sexual selection is responsible for waveform divergence. The authors favour this view on the basis of the fact that, within species, individual differences in components of EODs are used for mate discrimination9,10. How intraspecific sexual selection would produce speciation is not clear, however. Presumably it would involve different populations evolving, for some reason, differences in mate choice. The result would be divergence in sexual signals that incidentally led to reproductive isolation between populations. Differences in mate preference, mediated by differences in male mating calls, among populations of the Amazonian frog Physalaemus petersi may provide one such example11.

Several studies have linked sexual selection and rates of speciation. But just because a trait is subject to sexual selection within a species, it does not necessarily follow that interspecific differences arose through this process. Rather, we have known since the time of the early naturalists that many signals are used to allow an organism to distinguish individuals of different species from those of the same species. Such signals can arise in many ways that do not involve competition among members of one sex for reproductive success, unless one is willing to broaden the definition of sexual selection to encompass all mating decisions, including those involving individuals of other species. Reinforcement and reproductive character displacement, in which selection favours individuals that mate preferentially with members of the same species, are two well-established ideas about how the ability to recognize a member of a different species may arise to prevent interspecific mating without the operation of sexual selection (Arnegard et al.8 do acknowledge this possibility).

Regardless of whether signal evolution was driven by sexual selection or by selection for species recognition, Arnegard and colleagues' general point is an excellent one. Evolutionary radiation, whether adaptive or non-adaptive, requires speciation. It is certainly plausible that taxonomic groups with a greater ability to diversify in species-recognition signals, and that are thus more likely to speciate, may be those that diversify to the greatest extent.

Technological advances are opening a window onto many hitherto little-appreciated communication channels. For example, frogs can communicate using ultrasonic calls12, and polarized light is used for mate recognition in butterflies13. Findings such as those of Arnegard et al. provide fertile ground to further explore the contribution of using a previously empty communication niche in evolutionary radiations. The constraining effect of eavesdropping predators also remains to be determined. The observation14,15 that weakly electric fishes in South America exhibit a great diversity of species and EODs, even though they have evolved in the presence of predators that can perceive electric discharges16, suggests that partitioning of signal space along the communication axis is the more potent evolutionary force. Examples from other sensory modalities are needed to test the generality of this hypothesis.


  1. 1

    Schluter, D. The Ecology of Adaptive Radiation (Oxford Univ. Press, 2000).

    Google Scholar 

  2. 2

    Losos, J. B. Am. Nat. 175, 623–639 (2010).

    Article  Google Scholar 

  3. 3

    Losos, J. B. & Mahler D. L. in Evolution Since Darwin: The First 150 Years (eds Bell, M. A., Futuyma, D. J., Eanes, W. F. & Levinton, J. S.) 381–420 (Sinauer, 2010).

    Google Scholar 

  4. 4

    Glor, R. E. Annu. Rev. Ecol. Evol. Syst. (in the press).

  5. 5

    Simpson, G. G. The Major Features of Evolution (Columbia Univ. Press, 1953).

    Google Scholar 

  6. 6

    Yoder, J. B. et al. J. Evol. Biol. 23, 1581–1596 (2010).

    CAS  Article  Google Scholar 

  7. 7

    Rundell, R. J. & Price, T. D. Trends Ecol. Evol. 24, 394–399 (2009).

    Article  Google Scholar 

  8. 8

    Arnegard, M. E. et al. Am. Nat. 176, 335–356 (2010).

    Article  Google Scholar 

  9. 9

    Hopkins, C. D. & Bass, A. H. Science 212, 85–87 (1981).

    ADS  CAS  Article  Google Scholar 

  10. 10

    Hanika, S. & Kramer, B. Behaviour 142, 145–166 (2005).

    Article  Google Scholar 

  11. 11

    Boul, K. E., Funk, W. C., Darst, C. R., Cannatella, D. C. & Ryan, M. J. Proc. R. Soc. Lond. B 274, 399–406 (2007).

    Article  Google Scholar 

  12. 12

    Feng, A. S. et al. Nature 440, 333–336 (2006).

    ADS  CAS  Article  Google Scholar 

  13. 13

    Sweeney, A., Jiggins, C. & Johnsen, S. Nature 423, 31–32 (2003).

    ADS  CAS  Article  Google Scholar 

  14. 14

    Fernandes, C. C., Podos, J. & Lundberg, J. G. Science 305, 1960–1962 (2004).

    ADS  CAS  Article  Google Scholar 

  15. 15

    Turner, C. R., Derylo, M., de Santana, C. D., Alves-Gomes, J. A. & Smith, G. T. J. Exp. Biol. 210, 4104–4122 (2007).

    Article  Google Scholar 

  16. 16

    Stoddard, P. K. Adv. Study Behav. 31, 201–242 (2002).

    Article  Google Scholar 

Download references

Author information



Rights and permissions

Reprints and Permissions

About this article

Cite this article

Leal, M., Losos, J. Communication and speciation. Nature 467, 159–160 (2010). https://doi.org/10.1038/467159a

Download citation

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.


Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

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