Fish found in flagrante delicto

Article metrics

Genetic analysis of cichlid fish in Nicaraguan lakes reveals a possible case of repeated sympatric speciation: the creation of two species from one in the same environment.

A species of vertebrate gives rise to another species on average once every few million years. That is somewhat longer than the time span of a typical research grant, which is one reason why speciation is a tough subject to study. But sometimes the process can be caught in the act, giving us a window onto the origin of biodiversity.

A paper by Wilson et al.1, just published in Proceedings of the Royal Society, reports on just such an apparent case in cichlid fishes from Nicaragua. The contribution is all the more interesting because it seems that sympatric speciation is taking place. This is the controversial hypothesis which holds that one species can split into two without the benefit of geographical barriers to prevent interbreeding.

The celebrated evolutionist Ernst Mayr built a persuasive case that speciation can occur only when geographical barriers enforce non-random mating while new species are emerging2. This is allopatric speciation, and Mayr's view became the orthodoxy for many years. There is now growing evidence, however, that sympatric speciation can happen in the right circumstances3. The study by Wilson et al. breaks new ground by providing the first example of replicated sympatric speciation: one species splitting in two several times independently. If confirmed, these natural replicates would offer a unique opportunity to identify the factors that trigger sympatric speciation.

The cichlid fishes are particularly attractive for speciation studies. The 700 species are remarkable for their diverse colours, shapes and sizes, and behaviours (which include complex courtship and parental care). Cichlids can speciate rapidly: the 300 species in Lake Victoria, east Africa, may have descended from a single ancestral cichlid in only the past 12,400 years4. Further, these fish may be prone to sympatric speciation. A remarkable case involves cichlids from isolated crater lakes in Cameroon5. Genetic data show that nine species living in Lake Bermin are each other's closest relatives, implying that they speciated there. The lake is only 0.6 km2 in surface area and 14.5 m deep, making it implausible that geographical barriers were involved.

Wilson et al. 1 studied cichlids in four lakes in Nicaragua. Each lake has two types of fish that differ in colour — a 'normal' morph and a 'gold' morph. In two of the four lakes, Wilson et al. found statistically significant differences between the nuclear and/or mitochondrial gene frequencies of the morphs. Further, mitochondria from the different morphs within each lake are statistically more similar to each other than to those from different lakes, supporting the authors' interpretation that sympatric speciation has occurred (Fig. 1). Other explanations are that allopatric speciation occurred on a local scale within each lake; or that parallel invasions of the lakes by two species that originated in different lakes were followed by hybridization between the two, and the spread of mitochondria from one species to the other. These alternatives may be less likely, but they serve to show how hard it is to prove sympatric speciation conclusively, even with solid molecular data.

Figure 1: Molecular signatures of speciation: possible outcomes of genetic analyses of cichlid fish in two Nicaraguan lakes.

a, Allopatric speciation, in which species form while a geographical barrier is present and then come into contact secondarily. b, The corresponding pictures for 'replicated sympatric speciation', where new species form independently within each lake in the absence of geographical barriers. At the bottom is the gene tree for genes from gold (G) and normal (N) individuals sampled from the two lakes that would be consistent with each course of events. The gene trees reflect the historical relations of the populations. Under allopatric speciation (a), genes from the gold forms in different lakes will be more similar than genes from the gold and normal forms within each lake. The opposite is true under sympatric speciation (b). Wilson and colleagues' data1 tend to support this second pattern, and the view that sympatric speciation has occurred.

Earlier studies of these fish6 found strong 'assortative' mating based on colour (that is, like-colour morphs mate with like). Population-genetic models show that assortative mating is a particularly powerful mechanism for generating the genetic differences that lead to speciation7, so this propensity may have set the stage for sympatric speciation. Even one successful hybridization per generation should prevent substantial divergence for the molecular markers studied by Wilson et al.8, so apparently there have been strong barriers to genetic mixing between the colour morphs for quite some time.

There is another striking form of variation found in these fish. Cichlids have distinctive bones in their throats called pharyngeal jaws, which have played a key role in their spectacular ecological diversification9. The Nicaraguan cichlids have two forms of these jaws, one suited to a diet of snails and the other to soft prey. Typically, speciation involves genetic divergence for ecological traits, and so one would expect one of the two forms of pharyngeal jaws to be strongly associated with each of the colour morphs. They are not. Although there is a correlation, it is only weak (about r = 0.48). Further, Wilson et al. could not find molecular differences between the pharyngeal morphs.

What gives? Two possibilities come to mind. The first one, favoured by the authors, relates to mathematical models of sympatric speciation10,11. Consider a population that is under divergent selection for two different ecological forms, say eating snails and eating worms, and that also mates assortatively, say with respect to colour. The models show that genetic variation for the ecological trait and for colour will become correlated. This correlation further reinforces divergence of both the ecological trait and colour, and under the right conditions it can cause the population to split in two completely within only a few dozen generations. The Nicaraguan cichlids could be at an intermediate step in this process, in which case the colours, pharyngeal jaws and molecular variation will eventually become tightly associated. A difficulty with that suggestion, however, is that the molecular data suggest the colour morphs have already been isolated for a considerable time.

There is another possibility: that the dramatic variation in pharyngeal jaws may not have genetic causes. The pharyngeal jaws of these fish change their form in response to diet12. The two colour morphs live in different habitats at certain times of the year, where different prey are available. Perhaps diet alone directly causes the variation in pharyngeal jaws and its modest correlation with colour. In that case, the jaws may be but an interesting sideshow to speciation.

Non-genetic effects might have a hand in yet another part of this story. In these cichlids, the young spend several weeks in the care of both parents, giving offspring the opportunity to learn the colour of their parents. Perhaps the mating preferences of the fish are influenced in part by behavioural imprinting. Indeed, lab experiments on these cichlids suggest that a fish's response to gold and normal individuals is not affected by its own colour, but is affected by the colour of the parents that reared it13. Imprinting may influence speciation in birds14, so there might be a connection between two remarkable features of cichlids — their high rates of speciation and their parental care.


  1. 1

    Wilson, A. B., Noack-Kunnmann, K. & Meyer, A. Proc. R. Soc. Lond. B 267, 2133 –2141 (2000).

  2. 2

    Mayr, E. Animal Species and Evolution (Harvard Univ. Press, 1963).

  3. 3

    Howard, D. J. & Berlocher, S. H. (eds) Endless Forms: Species and Speciation (Oxford Univ. Press, 1998).

  4. 4

    Johnson, T. C. et al. Science 273, 1091– 1093 (1996).

  5. 5

    Schliewen, U. K., Tautz, D. & Pääbo, S. Nature 368, 629– 632 (1994).

  6. 6

    McKaye, K. R. Envir. Biol. Fishes 5, 75–78 (1980).

  7. 7

    Kirkpatrick, M. & Ravigné, V. Am. Nat. (in the press).

  8. 8

    Wright, S. Genetics 16, 97–159 ( 1931).

  9. 9

    Liem, K. F. Syst. Zool. 22, 425–441 (1973).

  10. 10

    Dieckmann, U. & Doebeli, M. Nature 400, 354–357 (1999).

  11. 11

    Kondrashov, A. S. & Kondrashov, F. A. Nature 400, 351–354 (1999).

  12. 12

    Meyer, A. Biol. J. Linn. Soc. 39, 279–299 (1990).

  13. 13

    Barlow, G. W., Francis, R. C. & Baumgartner, J. V. Anim. Behav. 40, 713– 722 (1990).

  14. 14

    Irwin, D. E. & Price, T. Heredity 82, 347–354 (1999).

Download references

Author information

Correspondence to Mark Kirkpatrick.

Rights and permissions

Reprints and Permissions

About this article

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.