brief communications blood-meal sizes were also comparable for L. McCarroll*, M. G. Paton*, S. H. P. P. Second, does Doncaster et al.’s model the resistant phenotype. Karunaratne†, H. T. R. Jayasuryia‡, aid our understanding of the patterns of Artificially feeding insecticide-resistant K. S. P. Kalpage§, J. Hemingway* sexuality observed in nature? Several theo- and insecticide-susceptible mosquito colo- *School of Biosciences, Cardiff University, retical problems have been pointed out nies (PelRR and PelSS, respectively11) with Cardiff CF10 3TL, UK with the tangled bank model, and empirical blood infected by W. bancrofti to an inter- e-mail: [email protected] studies have tested its (identical) assump- mediate level of parasitaemia — which †Zoology Department, Peradenyia University, tions and predictions. should result in the infection of mosquitoes Peradenyia, Sri Lanka For example, the advantage of sex is without substantial insect mortality — pro- ‡Zoology Department, Open University, Nawala, reduced if asexual populations are made up duced stage-L3 infective parasite larvae after Nugegoda, Sri Lanka of multiple clones that fill multiple por- 12 days in 76% of PelSS females (nǃ250), §Anti-Filariasis Campaign, Colombo, Sri Lanka tions of the niche space4,5; coexistence of 4 but no larvae in any of the PelRR females 1. Hemingway, J. & Ranson, H. Annu. Rev. Entomol. 45, 369–389 sexuals and asexuals is rarely observed ; (nǃ200). Our results indicate that an (2000). plant and algae populations consisting of increase in esterase activity could affect the 2. Hemingway, J. & Karunaratne, S. H. P. P. Med. Vet. Entomol. 12, mixtures of genotypes have carrying capac- 1–12 (1998). development of stage-L1 W. bancrofti lar- 3. Raymond, M., Callaghan, A., Fort, P. & Pasteur, N. Nature 350, ities not much greater than that of the aver- vae, which may be arrested in the gut cells 151–153 (1991). age of their component genotypes, and of insecticide-resistant but not insecticide- 4. Hemingway, J. et al. Ins. Mol. Biol. 9, 93–99 (2000). rarely greater than that of the best susceptible mosquitoes. 5. Paton, M. G. et al. Biochem. J. 346, 17–24 (2000). genotype8; and results from natural popula- 6. Hemingway, J. Ins. Biochem. Mol. Biol. 30, 1009–1015 (2000). Filarial infection severely damages the 7. Jayasekera, N., Kalpage, K. S. P. & De, S. Trans. R. Soc. Trop. tions indicate that the ecological and demo- mosquito host, often killing it. The spread Med. Hyg. 85, 250–254 (1991). graphic predictions of the tangled bank of esterase-based insecticide resistance in 8. Peiris, H. T. R. & Hemingway, J. Med. Vet. Entomol. 10, 283–287 model are not met6,9. field populations of C. quinquefasciatus (1996). Third, does the model of Doncaster et 9. Nicolas, L. & Plichart, C. Parasite 4, 253–257 (1997). may therefore be influenced by selection 10.Salazar, C. E. et al. Ins. Mol. Biol. 3, 1–13 (1994). al. provide new insights into the cost of sex pressures for both insecticide detoxifica- 11.Vaughan, A., Hawkes, N. & Hemingway, J. Biochem. J. 325, (males)? The authors argue that the cost of tion and reduction of the microfilarial 359–365 (1997). males is ecology dependent, so there is not burden. Similar esterase-based insecticide- 12.Brogdon, W. G. & Barber, A. M. Pest. Biochem. Physiol. 37, necessarily a twofold advantage to be 130–139 (1990). resistance mechanisms have been selected 13.Karunaratne, S. H. P. P. Southeast Asian J. Trop. Med. Pub. Hlth recouped in adaptive payoffs. However, one in field populations of the malaria vectors 30, 460–469 (1999). of the model’s implicit assumptions is that Anopheles albimanus12 and A. culicifacies13, different genotypes use different niches Supplementary information is available on Nature’s World-Wide which could directly affect the transmis- Web site (http://www.nature.com) or as paper copy from the (genotype-by-environment interactions), sion of malaria. London editorial office of Nature. so the model is not purely ecological. The way in which this type of model allows coexistence between sexuals and asexuals has already been discussed4,5. Evolution brium at which both sexual and clonal indi- Moreover, this mechanism is more appro- viduals persist”4. priately viewed as an adaptive payoff in its Paying for sex Although the mechanisms favouring sex own right, as in previous formulations1,3–5. is not easy appear to be identical, the elegantly simple In this case, it may reduce the cost to be form of Doncaster et al.’s Lotka–Volterra paid by other mechanisms, but such inter- xplaining the maintenance of sexual model means that it requires implicit actions between models have been dis- reproduction remains one of the great- assumptions and is therefore hard to com- cussed extensively10. Eest challenges for biology, with more pare with previous tangled bank models, Nonetheless, Doncaster et al. remind us than 20 hypotheses having been advanced which were based on numerical simulations of the role that the tangled bank mecha- so far1. Doncaster et al.2 have proposed and which made more explicit assump- nism could play in a pluralist explanation another possible explanation, but we ques- tions4,5. For example, does Doncaster et al.’s of sex, possibly by interacting with deleteri- tion the novelty and importance of their model require competition between siblings ous mutations in a fashion analogous to the suggested mechanism. (as with some formulations of the tangled Red Queen10. First, does this model2 provide a new bank), or is it the special case in which sib S. A. West, A. D. Peters mechanism to help explain sex? The model competition is excluded5? Institute of Cell, Animal, and Population assumes that different genotypes exploit It is hard to test the implicit assump- Biology, University of Edinburgh, different parts of the environment, and that tions in the new model2, but it may Ashworth Laboratories, West Mains Road, an asexual clone is not able to occupy all the render some aspects of the tangled bank Edinburgh EH9 3JT, UK environmental niches that are open to a mechanism more testable by emphasizing e-mail: [email protected] sexual population. This provides an advan- two parameters: the maximum population 1. Kondrashov, A. S. J. Hered. 84, 372–387 (1993). 2. Doncaster, C. P., Pound, G. E. & Cox, S. J. Nature 404, 281–285 tage to sex, and allows coexistence between growth rate, R0, and the degree of overlap Ȋ (2000). sexuals and asexuals. between sexual and asexual niches, . 3. Ghiselin, M. T. The Economy of Nature and the Evolution of Sex However, the assumptions are the same However, the importance of these para- (Univ. California Press, Berkeley, 1974). as those of the ‘tangled bank’ mechanism meters has been discussed previously — for 4. Bell, G. The Masterpiece of Nature: The Evolution for the maintenance of sex1,3–5. The message example, it has been pointed out6 that, and Genetics of Sexuality (Univ. California Press, Berkeley, 1982). that has emerged from previous considera- analogous to variation in R0 , higher fecun- 5. Maynard Smith, J. The Evolution of Sex (Cambridge Univ. Press, tions of this mechanism is therefore the dity leads to greater sib competition and Cambridge, 1978). same as that proposed by Doncaster et al.2 so increases the advantage of sex. It has 6. Burt, A. & Bell, G. Nature 326, 803–805 (1987). 4 7. Begon, M., Harper, J. L. & Townsend, C. R. Ecology 3rd edn — namely, that competition within a fixed also been shown how the advantage of (Blackwell Science, Oxford, 1996). set of niches can provide an advantage to sex varies with a ‘competition coefficient’, 8. Bell, G. Selection: The Mechanism of Evolution (Chapman & sex, with “the success of the clone [being] which is a version of Ȋ. Furthermore, such Hall, New York, 1997). 9. Lively, C. M. Nature 328, 519–521 (1987). restrained by the narrowness of its ecologi- competition coefficients are notoriously 10.West, S. A., Lively, C. M. & Read, A. F. J. Evol. Biol. 12, 7 cal range” and leading to “a stable equili- difficult to measure . 1003–1012 (1999). © 2000 Macmillan Magazines Ltd 962 NATURE | VOL 407 | 26 OCTOBER 2000 | www.nature.com