364 Review 14 Chen, F.C. and Li, W.H. (2001) Genomic divergences between humans and other hominoids and the effective population size of the common ancestor of humans and chimpanzees. Am. J. Hum. Genet. 68, 444–456 15 Ruvolo, M. (1997) Molecular phylogeny of the hominoids: inferences from multiple independent DNA sequence data sets. Mol. Biol. Evol. 14, 248–265 16 Bailey, W.J. et al. (1991) Molecular evolution of the psi-eta-globin gene locus: gibbon phylogeny and the hominoid slowdown. Mol. Biol. Evol. 8, 155–184 17 Horai, S. et al. (1992) Man’s place in Hominoidea revealed by mitochondrial DNA genealogy. J. Mol. Evol. 35, 32–43 18 Li, Y.J. et al. (1999) Paleo-demography of the Drosophila melanogaster subgroup: application of the maximum likelihood method. Genes Genet. Syst. 74, 117–127 19 Hey, J. and Kliman, R.M. (1993) Population genetics and phylogenetics of DNA-sequence variation at multiple loci within the Drosophila melanogaster species complex. Mol. Biol. Evol. 10, 804–822 20 Takahata, N. and Satta, Y. (1997) Evolution of the primate lineage leading to modern humans: phylogenetic and demographic inferences from DNA sequences. Proc. Natl. Acad. Sci. U. S. A. 94, 4811–4815 21 Wright, S. (1969) Evolution and the Genetics of Populations: the Theory of Gene Frequencies (Vol. 2), University of Chicago Press TRENDS in Ecology & Evolution Vol.16 No.7 July 2001 22 Wakeley, J. (1999) Nonequilibrium migration in human history. Genetics 153, 1863–1871 23 Takahata, N. (1995) A genetic perspective on the origin and history of humans. Annu. Rev. Ecol. Syst. 26, 343–372 24 Slatkin, M. (1977) Gene flow and genetic drift in a species subject to frequent local extinction. Theor. Popul. Biol. 12, 253–262 25 Pannell, J.R. and Charlesworth, B. (1999) Neutral genetic diversity in a metapopulation with recurrent local extinction and recolonization. Evolution 53, 664–676 26 Whitlock, M.C. and Barton, N.H. (1997) The effective size of a subdivided population. Genetics 146, 427–441 27 Slatkin, M. (1991) Inbreeding coefficients and coalescence time. Genet. Res. 58, 167–175 28 Barton, N.H. and Wilson, I. (1995) Genealogies and geography. Philos. Trans. R. Soc. London Ser. B 349, 49–59 29 Berry, A.J. et al. (1991) Lack of polymorphism on the Drosophila 4th chromosome resulting from selection. Genetics 29, 1111–1117 30 Kreitman, M.E. and Aguade, M. (1986) Excess polymorphism at the ADH locus in Drosophila melanogaster. Genetics 114, 93–110 31 Hewitt, G.M. (1999) Post-glacial re-colonization of European biota. Biol. J. Linn. Soc. 68, 87–112 32 Ibrahim, K.M. et al. (1996) Spatial patterns of genetic variation generated by different forms of dispersal during range expansion. Heredity 77, 282–291 33 Taberlet, P. et al. (1998) Comparative phylogeography and postglacial colonization routes in Europe. Mol. Ecol. 7, 453–464 34 Hewitt, G.M. (2000) The genetic legacy of the quaternary ice ages. Nature 405, 907–913 35 Rieseberg, L.H. and Noyes, R.D. (1998) Genetic map-based studies of reticulate evolution in plants. Trends Plant Sci. 3, 254–259 36 Song, K. et al. (1995) Rapid genome change in synthetic polyploids of Brassica and its implications for polyploid evolution. Proc. Natl. Acad. Sci. U. S. A. 92, 7719–7723 37 Liu, B. et al. (1998) Rapid genomic changes in newly synthesized amphiploids of Triticum and Aegilops. I. Changes in low-copy noncoding DNA. Genome 41, 272–277 38 Bromham, L. et al. (2000) The power of relative rates tests depends on the data. J. Mol. Evol. 50, 296–301 39 Rand, D.M. (1994) Thermal habit, metabolic rate and the evolution of mtDNA. Trends Ecol. Evol. 9, 125–131 40 Chao, L. and Carr, D.E. (1993) The molecular clock and the relationship between population-size and generation time. Evolution 47, 688–690 41 Gillespie, J.H. (1989) Lineage effects and the index of dispersion of molecular evolution. Mol. Biol. Evol. 6, 636–647 42 Li, W.H. et al. (1987) An evaluation of the molecular clock hypothesis using mammalian DNA-sequences. J. Mol. Evol. 25, 330–342 Sexual selection and speciation Tami M . Panhuis, Roger Butlin, M arlene Zuk and Tom Tregenza The power of sexual selection to drive changes in mate recognition traits gives it the potential to be a potent force in speciation. M uch of the evidence to support this possibility comes from comparative studies that examine differences in the number of species between clades that apparently differ in the intensity of sexual selection. We argue that more detailed studies are needed, examining extinction rates and other sources of variation in species richness. Typically, investigations of extant natural populations have been too indirect to convincingly conclude speciation by sexual selection. Recent empirical work, however, is beginning to take a more direct approach and rule out confounding variables. SEXUAL SELECTION (see Glossary) results from Tami M . Panhuis* M arlene Zuk Dept of Biology, University of California, Riverside, CA 92521, USA. * e-m ail: panhuist@citrus.ucr.edu Roger Butlin Tom Tregenza Centre for Biodiversity and Conservation, School of Biology, University of Leeds, Leeds, UK LS2 9JT. differential mating success among individuals within a population. Competition for fertilization occurs through direct competition between members of the same sex (e.g. male–male competition and sperm competition) or through the attraction of one sex to the other (e.g. female choice). Although long recognized as important in intrapopulation evolution, sexual selection has more recently been invoked as a driving force behind SPECIATION. Speciation, the splitting of one SPECIES into two or more, occurs by sexual selection when a parallel change in mate preference and SECONDARY SEXUAL TRAITS within a population leads to PREZYGOTIC ISOLATION between populations, and when this is the primary cause of REPRODUCTIVE ISOLATION. Classic models of speciation1,2 recognized that reproductive isolation, and subsequent speciation, could be generated by differences in sexual traits (including behaviours). Divergence in sexual traits between allopatric populations was considered to result either from drift, PLEIOTROPY or adaptation to environmental conditions, or following SECONDARY CONTACT, because individuals benefited by avoiding heterospecific matings (i.e. by REINFORCEMENT). It became clear, however, that changes between populations in sexual traits could also result from sexual selection and that this might represent a distinct process of speciation3,4. Sexual selection has the potential to lead to rapid divergence between populations, it can be independent of environmental differences, and it is predisposed to generate reproductive isolation because of its direct effect on traits involved in mate recognition. It is important to point out that the rapid change between populations http://tree.trends.com 0169–5347/01/$ – see front m atter © 2001 Elsevier Science Ltd. All rights reserved. PII: S0169-5347(01)02160-7 Review TRENDS in Ecology & Evolution Vol.16 No.7 July 2001 365 Box 1. Sexual conflict and speciation In species w here fem ales have m ore than one m ate, each m ale is selected to m axim ize the proportion of the reproductive effort of a female going into his offspring. This is in his interest even if it decreases the lifetime reproductive output of the female. A conflict between the sexes is created as males attempt to both manipulate females and to out-compete other males (even if this is costly to the female) and females evolve to avoid these costs and controls. Such antagonistic coevolution has the potential to drive rapid evolutionary change, particularly in aspects of the reproductive system, giving it the potential to create divergence between allopatric populationsa. A mathematical model for the evolution of reproductive barriers via sexual conflict has recently been proposed b in which conflicts of interest between the sexes over mating rate are predicted to cause divergence, because it pays females to reduce the proportion of the male population with whom they are reproductively compatible. The model demonstrates that, under some scenarios, conflict can lead to runaway coevolution driving divergence. Also, in contrast to models where change occurs because of GENETIC DRIFT (see Glossary), sexual conflict generates direct selection on reproductive traits: divergence is expected to be more rapid when populations are larger. The generality of this model has been questioned on the basis that females will only benefit from incompatibility with a proportion of males if such incompatibilities actually prevent matings. If this is the case, males might be able to as a result of sexual selection can also play an indirect role in speciation by increasing the overall rate of change within isolated populations5. Some have argued that this indirect role might be more important than a direct role for sexual selection in speciation6. Theoretical studies have convincingly demonstrated the potential for speciation to result from sexual selection (see Turelli et al.7, this issue). Here, we focus on recent empirical evidence for SPECIATION BY SEXUAL SELECTION. Much of this support comes from comparative studies that examine differences in the number of species between CLADES. Patterns of variation in secondary sexual traits and mating preferences between populations or species also suggest that sexual selection plays a role in speciation, but these ‘signatures’ have the drawback that evolutionary processes cannot be tracked through time and thus exclude alternative explanations. We discuss the strengths and weaknesses in using the comparative method for understanding the role of sexual selection in speciation. We then consider why the common ‘signatures’ that lend support for speciation by sexual selection are not conclusive, and discuss criteria for demonstrating the role of sexual selection in speciation. Comparative evidence Important evidence showing how sexual selection affects speciation comes from comparisons across taxa. Darwin8 noted that elaborate secondary sexual characters tended to occur in groups that also had high species richness, suggesting that sexually selected ornamentation might cause taxa to diversify, but rigorous tests of this association have appeared only recently. The first such test used 20 sister pairs of http://tree.trends.com redirect their mating effort, changing the dynamics of the system c. An alternative approach to conflicts of interest has been suggested d wherein species in which males invest very little in matings might be less prone to speciation, as there will be less selection on males to avoid outbreeding. References a Rice, W.R. (1998) Intergenomic conflict, interlocus antagonistic co-evolution, and the evolution of reproductive isolation. In Endless Forms: Species and Speciation (Howard, D.J. and Berlocher, S.H., eds), pp. 261–270, Oxford University Press b Gavrilets, S. (2000) Rapid evolution of reproductive barriers driven by sexual conflict. Nature 403, 886–889 c Tregenza, T. et al. (2000) Sexual conflict and speciation. Nature 407, 149–150 d Parker, G.A. and Partridge, L. (1998) Sexual conflict and speciation. Philos. Trans. R. Soc. London Ser. B 353, 261–274 passerine bird tribes that differ in their proportions of sexually dichromatic and monochromatic species9. Sister taxa have, by definition, accumulated diversity over the same period of time. Therefore, the tribe with the greater number of extant species necessarily has the higher net speciation rate (rate of speciation minus the rate of extinction). If sexual selection promotes speciation, diversity is expected to be higher in clades with more dichromatic species, assuming that sexual differences in plumage result from sexual selection. Of 15 comparisons with marked differences in the frequency of dichromatism, 12 showed differences in species richness in the expected direction9. Similar comparative studies have since supported speciation by sexual selection: in birds, by comparing the number of species in taxa with different mating systems10, and different degrees of feather ornamentation11; and in plants, by comparing lengths of nectar spurs12. More ornamented species have more subspecies11, suggesting ongoing differentiation. These studies concentrate on the contribution of sexual selection to speciation through divergence in mating signals and preferences. Comparative evidence also suggests, however, that postmating effects promote speciation and that sexual selection influences the evolution of POSTZYGOTIC ISOLATION. If a female mates more than once, the opportunity for sexually antagonistic selection arises (Box 1). If such selection has a role in speciation, MONANDROUS clades should show lower speciation rates than would related POLYANDROUS clades. A strong pattern of this type was recently documented using data from 25 pairwise comparisons in five orders of insects13. Another comparative analysis14 demonstrated that genital morphology evolves more rapidly in polyandrous 366 Review TRENDS in Ecology & Evolution Vol.16 No.7 July 2001 Box 2. Haldane’s rule and speciation In pairs of anim al species w here prem ating barriers can be overcom e and som e hybrid offspring generated, it is com m on to find that one sex is inviable or infertile. In 1922, J.B.S. Haldane noted a striking pattern in these crosses: the sex that is inviable or sterile is alm ost invariably the heterogam etic (XY) sex. This is one of the strongest ‘rules’ in evolutionary biology. It is obeyed in 151 out of 157 crosses in Drosophila and m am m als (m ale heterogam etic) and 91 out of 98 crosses in Lepidoptera and birds (fem ale heterogam etic)a. Exam ining the tim e course of speciation in Drosophila suggests that Haldane’s rule is a nearly obligatory ‘first step in the evolution of postzygotic isolation’ a . How does Haldane’s rule im plicate sexual selection in speciation? There are several com peting explanations for the pattern a, one of w hich, the ‘faster-m ale’ theory b , invokes sexual selection. The argum ent is that m ale sterility m ight be m ore com m on than fem ale sterility in hybrids either because sperm atogenesis is inherently m ore sensitive than oogenesis, or because sexual selection and sexual conflict cause m ore rapid divergence of genes expressed in m ales. Divergent loci interact poorly in hybrids, leading to sterility. If sperm atogenesis is a sensitive process, this is presum ably because sexual selection or sexual conflict favours a high rate of sperm production. Clearly, this cannot be the only mechanism underlying Haldane’s rule because it is unlikely to explain asymmetrical inviability and because it works against the rule in taxa with heterogametic females. However, evidence that it contributes to sterility in taxa with XY males would demonstrate a role for sexual selection in the origin of postzygotic isolation. The faster-male theory is now supported by two lines of evidence. First, chromosome regions introgressed from one species of Drosophila into another are more likely to contain male-sterile than female-sterile genesc,d. This fits the expectation of more rapid divergence at loci expressed only in males. Second, the faster-male theory is not dependent on the Y chromosome being inert and thus insect clades, implicating postmating sexual selection in the rapid evolutionary divergence of genitalia. If polyandry generates greater opportunity for sexual selection as well as for sexual conflict, the causal link to speciation remains unclear. Sex-related genes (i.e. those involved in sex determination or any aspect of mating and fertilization) show higher nonsynonymous:synonymous nucleotide substitution ratios between closely related species than do other classes of genes15. This suggests divergence resulting from sexual selection around the time of speciation. Finally, the evidence explaining HALDANE’S RULE by faster male evolution (Box 2) strongly suggests that sexual selection can promote genetic divergence underlying hybrid male sterility. Counting species is not enough Comparisons based on the numbers of species in different groups are not without problems. If taxonomists use secondary sexual traits in species definitions, the numbers of named species could be upwardly biased in dimorphic taxa. Comparative studies also tend to assume that dimorphic traits have diverged by sexual selection, often with limited direct evidence. At best, such studies demonstrate a correlation between speciation and sexual selection (across taxa or in time) rather than a causative link. Comparative analyses must deal with the problem of confounding variables: sexual dichromatism, for http://tree.trends.com recessive loci on the X chromosome being expressed in hybrids, which is the basis of the other major explanation for the rule. Taxa with non-degenerate Y chromosomes are, therefore, expected to follow Haldane’s rule for sterility but not for inviability. This is exactly the pattern observed in a comparison between two mosquito genera: Aedes, where the Y chromosome is homologous to the X, and Anopheles, where the Y is inert e. References a Orr, H.A. (1997) Haldane’s rule. Annu. Rev. Ecol. Syst. 28, 195–218 b Wu, C-I. and Davis, A.W. (1993) Evolution of postmating reproductive isolation: the composite nature of Haldane’s rule and its genetic bases. Am. Nat. 142, 187–212 c True, J.R. et al. (1996) A genome-wide survey of hybrid incompatibility factors by introgression of marked segments of Drosophila mauritiana chromosomes into Drosophila simulans. Genetics 142, 819–837 d Hollocher, H. and Wu, C-I. (1996) The genetics of reproductive isolation in the Drosophila simulans clade: X vs. autosomal effects and male vs. female effects. Genetics 143, 1243–1255 e Presgraves, D.C. and Orr, H.A. (1998) Haldane’s rule in taxa lacking a hemizygous X. Science 282, 952–954 example, could be correlated with some other variable that actually causes elevated speciation rates. The only safeguard against this possibility is to control for as many such variables as possible. A particularly thorough analysis16 attempted to minimize the problem by simultaneous analysis of six different explanations for variation in species richness among bird families: chance (i.e. by fitting models of random CLADOGENESIS), body size, life history, sexual selection, intrinsic ecological factors and extrinsic abiotic factors. Using comparisons of species numbers in 28 unequivocal sister taxa, strong support was found for an association with sexual dichromatism but not with mating system. Generalist feeding habits, high dispersal and large, fragmented geographical ranges are also associated with high diversity. Despite the limitations of comparisons based on species counts, comparative studies do suggest that sexual selection can accelerate the net rate of speciation. Divergence in secondary sexual characters might, however, generate reproductively isolated, but ecologically equivalent species unable to coexist in SYMPATRY. Some evidence suggests that the processes of extinction and speciation in birds are determined by different attributes16 and might, therefore, be uncorrelated. If this is true, it becomes difficult to interpret what net speciation rates tell us about the speciation process. In the insect study mentioned previously13, elevation in net speciation rate is much Review TRENDS in Ecology & Evolution Vol.16 No.7 July 2001 367 mating signals and preferences; these differences will be the major barrier to gene exchange and the species will differ little in other traits. If sexual selection is a common cause of speciation, these patterns should be encountered frequently. Signatures from population differences Fig. 1. Exam ples of sexual selection. (a) A m ale from the centre of a hybrid zone betw een subspecies of the m eadow grasshopper Chorthippus parallelus. A steep cline in fem ale preference across the zone indicates that preference is under selection 24. (b) A m ating pair from a laboratory m ate choice test in the m eadow grasshopper. There is sufficient divergence betw een allopatric populations of this species in Europe to generate significant assortative m ating. In som e cases, this divergence m ust have been accum ulated since the end of the last glaciation 23. Sexual selection m ight explain such rapid evolution. greater in comparisons within than between families (relative rate 3.3 and 1.1, respectively), suggesting only a short-term change in the balance between speciation and extinction. This implies that comparative analyses should focus on recently diverged sister taxa where the signal of elevated speciation rate will be strongest. Although sexual selection might promote speciation, ecological processes are more important in determining the numbers of species that can coexist and, over longer time scales, might obscure the effect of increased speciation rate. Even if the comparative evidence does point to a role for sexual selection in speciation, it does not directly distinguish between acceleration of divergence in ALLOPATRY, initiation of divergence in sympatry, or involvement in reinforcement. Some attempts have been made to address this problem. Further analysis17 of earlier data9, focusing on the number of species of each taxon living together rather than the total species richness, could not show that more diverse, sexually dichromatic taxa have greater numbers of species in sympatry than do the less diverse, monochromatic taxa. This suggests that sexual selection promotes allopatric divergence rather than promoting SYMPATRIC SPECIATION and that the allopatric species are ecological equivalents, prevented from expanding their ranges into sympatry by competition. In flocks of cichlid fish species, however, frequent sympatric occurrence of different colour morphs suggests that sexual selection can promote sympatric speciation despite a lack of ecological separation18. Are there conclusive signatures of speciation by sexual selection? Via19 (this issue) refers to patterns that provide indirect evidence of sympatric speciation as ‘signatures’. Two signatures might be recognized for speciation by sexual selection: (1) within species, populations will vary in sexually selected traits and associated preferences and this might generate partial premating isolation between populations; and (2) closely related species will differ markedly in http://tree.trends.com There are many examples of variation among populations in potentially sexually selected traits20,21. In some cases, the variable traits can be shown to be under sexual selection within populations (e.g. guppies; Table 1), but, even in these cases, the cause of divergence among populations is usually unclear. In other examples22,23, partial premating isolation has been demonstrated among populations; however, the signal traits involved are unknown. Variation in female preference has only been studied rarely, using either mating tests (e.g. grasshoppers24 : Fig. 1) or playback experiments (e.g. lacewings25). Table 1 illustrates the extent to which some recent studies of geographical variation in sexually selected traits have covered these aspects. Clearly, the signature is strongest if all components are included, but this has yet to be achieved. It is possible that divergence between populations in sexually selected traits and/or preferences could result in speciation. Speciation, however, is just one of several evolutionary outcomes of the sexually selected differences. Upon secondary contact, the partial barrier to gene flow between populations might diminish if postmating isolation is absent and if the differences in behaviour are insufficient to prevent hybridization. For example, isolation could be asymmetrical (i.e. only one population exhibits female preference for males of their own population) as in the case of Drosophila melanogaster populations from Zimbabwe (Table 1). These populations exhibit behavioural premating isolation from other non-African populations, but this is the result of asymmetrical female preference, and postmating isolation has not been detected26. This suggests that populations would interbreed in sympatry and differentiation would be lost. This type of breakdown of signal differences and ASSORTATIVE MATING is seen in some hybrid zones (e.g. fire-bellied toads, Bombina27 ), although natural and sexual selection may restrict hybridization to a narrow region. The evolutionary outcome of population differences in sexually selected traits and preferences could also depend on whether there is a mismatch between the male trait and the female preference. A mismatch could result if a population changes its male display trait (possibly as a result of environmental pressures), but not in female preference. A study28 on female preferences in the house finch Carpodacus mexicanus illustrates this phenomenon. Geographical variation in male plumage colouration and the amount of ventral pigmentation (patch size) is driven by environmental differences in regional and local access to carotenoid plumage pigments. Females, independently of their Review 368 TRENDS in Ecology & Evolution Vol.16 No.7 July 2001 Table 1. Examples of detecting signatures of speciation by sexual selectiona–c Taxa Exampled Insects Drosophila m elanogaster D. silvestris Chrysoperla spp. Chorthippus parallelus Ephippiger ephippiger Gryllus texensis and G. rubens Agrotis segetum Type of trait Trait under Preference sexual variation selectione Prezygotic reproductive isolatione,f Character Refs displacement Zim babw e fem ales prefer Zim babw e m ales over non-Zim babw e m ales Asym m etrical fem ale preference am ong different Israel populations Sex com b row num ber Substrate-transm itted acoustic signals Unknow n Yes Asym m etrical Unlikely Unknow n Yes Asym m etrical Unlikely Yes Unknow n Asym m etrical Yes Unlikely Unlikely 54 25 Unknow n Yes Yes, betw een species and ‘song m orphs’ Yes No 23 Unknow n Yes Yes, in som e population pairs Probably Unlikely 55 Yes Yes, betw een species Probably w eak No 39 Unknow n Yes, betw een species Yes Unlikely 56 M ale song, pherom ones and assortative m ating Fem ale preference functions and m ale calling song (num ber of syllables/chirp) Pulse rate of m ale song and associated fem ale preference Pherom one blend and m ale response 26,52,53 Frogs Acris crepitans M ale advertisem ent calls (independent of body size) and fem ale preference for m ale calls Unknow n Yes Probably No 57–59 Fish Poecilia reticulata M ale colour pattern and fem ale preference for m ales from their ow n locality Sexually dim orphic colouration and associated preferences Yes Yes Little Unlikely 60 Yes Yes, betw een Yes, betw een species/m orphs species/m orphs Possible 18 M ale strut displays, body size and feather m orphology differing in the Gunnison Basin, CO (USA) population (recently given species status) com pared w ith Jackson County, CO and M ono County, CA, (USA) populations M ale display traits, such as bow er structures, and fem ale preferences in at least one population M ale song betw een tw o sym patric subspecies that do not interbreed, yet are connected by a ring of populations that do interbreed Yes Unknow n Unknow n Unlikely 61,62 Yes Probably Yes Unknow n Probably, but Unlikely m ight be asym m etric Only dem onstrated No betw een subspecies Haplochrom ine cichlids of Lake Victoria Birds Centrocercus urophasianus Am blyornis inornatus Phylloscopus trochiloides 45 49, 50 aThis is not a com plete list of exam ples, but is intended only as an illustration of the phenom enon. Also see Refs 5,6,35. exam ples included have low genetic divergence as judged by allozym e, DNA or m itochondrial DNA (m tDNA) sequence divergence, or taxom onic status. Phylloscopus50 ssp. and Centrocercus populations63 show m tDNA divergence betw een subspecies/species. A Zim babw ean D. m elanogaster population show s greater population structure com pared w ith North Am erican populations on the basis of DNA sequence divergence64. cSom e degree of postzygotic isolation has been found betw een Chrysoperla25 spp. and Chorthippus23 populations. Studies have not dem onstrated postzygotic isolation betw een Zim babw e and ‘cosm opolitan’ populations of D. m elanogaster 26 or betw een Gryllus texensis and G. rubens39. Postzygotic isolation in the cichlids is inferred to be low. To the best of our know ledge, postzygotic isolation data are not available for the other exam ples in the table. d Com m on nam es: Acris creptians, Cricket frogs; Agrotis segetum , Turnip m oths; Am blyornis inornatus, Vogelkop Bow erbirds; Centrocercus urophasianus, Sage grouse; Chorthippus parallelus, Grasshoppers; Chrysoperla spp., Lacew ings; Drosophila m elanogaster and D. silvestris, Fruit flies; Ephippiger ephippiger , Bush crickets; Gryllus texensis and G. rubens, Crickets; Poecilia reticulata, Guppies; Phylloscopus trochiloides, Greenish Warblers. eUnknow n – to the best of our know ledge, data are not available. f Not necessarily com plete reproductive isolation. b M ost population, prefer the ancestral male type with the brightest and largest patches. Upon subsequent contact, this mismatch between the male trait and the female preference would probably lead to interbreeding between populations, rather than prezygotic isolation. Mismatches are also associated with the sensory bias and sensory exploitation model of sexual selection29–31. This model proposes that females respond http://tree.trends.com to certain signals because of pre-existing properties of their sensory system30,32 and hence male traits evolve after female preferences33. This phenomenon has been demonstrated in several taxa (e.g. fish31, frogs33 and spiders34)35 by revealing that females of closely related species prefer males with the heterospecific derived male trait rather than conspecific males without the trait. Population divergence in male traits might or might not occur, depending on the environmental or Review TRENDS in Ecology & Evolution Vol.16 No.7 July 2001 developmental constraints on the rapidity with which the male trait could evolve to exploit the female preference. There is also evidence that female sensory biases are rather generalized and can be satisfied by a variety of specific male traits32,36. Therefore, different male traits might evolve in response to the same sensory bias leading to divergence in signal traits without increasing isolation. This model of sexual selection seems unlikely to lead directly to speciation given the fact that females, all else being equal, will tend to interbreed with males that exploit their sensory bias, independent of their population. However, it could exaggerate already established species differences32,35. Alternatively, it could be that local environmental conditions alter the female sensory bias or make some male traits better at exploiting biases, just as specific traits might be better indicators of male condition in some environments than others37. In this case, different coevolutionary paths of signal and preference might generate speciation. A different perspective on mismatches between traits and preferences suggests that they might provide good evidence for the role of sexual selection in speciation. If female preferences evolve for ‘good genes’ or Fisherian reasons, they are expected to be open-ended, or at least to have mean preferences for trait values above the male population mean. This is because the sexual advantage of the trait is balanced by its cost to the male. Therefore, if two populations differ in the mean value of the male trait, they are likely to have even greater differences in female preference. This can be contrasted with the expectation for signals and preferences that have evolved for species recognition, where a match between male trait mean and female preference mean would be expected38. The mismatch between signal and preference can generate asymmetric isolation between populations with intermediate levels of divergence: this pattern has been observed in tests of mate choice between populations of Desmognathus salamanders22. However, this model does not explain the pattern of assortative mating observed among Chorthippus grasshopper populations23. Signatures from species differences Because we cannot see into the future, observations on divergent populations can only ever show that steps towards speciation might be promoted by sexual selection: how often the process is completed, and by what route, remains uncertain. Studying species differences suffers from the opposite problem: the difficulty of inferring the past. Many morphologically and genetically similar species differ markedly in mating signals and preferences. It might be inferred from this that sexual selection has driven speciation, but to do so requires that other modes of divergence be excluded. The possibility that species recognition has been an important force in shaping the changes seen between male traits used to attract mates or female http://tree.trends.com 369 preferences can be difficult to exclude. Species recognition is important when interpopulation or heterospecific mating is costly, producing relatively inviable or infertile offspring. Traits facilitating species recognition might then diverge under natural selection when populations or closely related species have overlapping geographical ranges2. Thus, reinforcement or REPRODUCTIVE CHARACTER DISPLACEMENT (RCD) could drive the change in male traits and female preferences, eliminating a direct role of sexual selection39. However, it has been proposed33,40 that sexual selection and species recognition are part of a continuum and that sexual selection might facilitate reinforcement41. Several studies have shown (e.g. butterflies42; fish35 and frogs30) that male display traits can be both sexually selected and used in species recognition. Contrary results have been found in two sibling species of Hawaiian Drosophila, where the sexually selected trait is not used for species recognition40. Evidence favouring speciation by sexual selection rather than reinforcement or RCD could demonstrate that: (1) the trait in question does not differ significantly in areas of sympatry versus allopatry (the opposite trend is a distinguishing feature of character displacement39); (2) there is trivial or no postzygotic isolation (i.e. the fitness reduction required to drive reinforcement or RCD is absent); and (3) the trait used in mate acquisition is used both in sexual selection and species recognition. However, it is possible that the initial divergence in traits or preferences could have been the result of natural selection to decrease hybridization and then been subject to sexual selection, or vice versa33, making a clean separation difficult. An analysis of data on 41 species of haplochromine cichlids from Lake Victoria showed that species pairs with a substantial range overlap were more likely to differ in sexually selected colouration than were pairs with little range overlap18. This is the pattern expected from reinforcement or RCD, but, in this case, there is little evidence for hybrid fitness reduction and the extremely short time scale for speciation in the lake makes the two stage process of allopatric divergence followed by range expansion and divergence in secondary sympatry improbable. The authors favoured a model of speciation as a result of disruptive sexual selection in sympatry (see Turelli et al.7, this issue). What should new studies aim to demonstrate? It is improbable that any single test will separate speciation by sexual selection from other modes of speciation. This is a problem common to all speciation research and stems from our inability to observe the whole process, forcing us either to infer the most probable future course of events or to separate different possible histories. Therefore, recent studies have aimed to overcome these difficulties by accumulating observations consistent with speciation by sexual selection and by trying to 370 Acknow ledgements We thank A. Uy for com m ents on this article and access to unpublished m aterial. Review TRENDS in Ecology & Evolution Vol.16 No.7 July 2001 rule out alternative scenarios. These studies have suggested the following major components: • Substantial differences occur among populations or species in male sexually selected traits with correlated female preferences43–45. • Changes are sufficient to result in prezygotic isolation should populations come into contact39. • There is little genetic differentiation between populations or species, suggesting rapid divergence43–45. • There is little or no reduction in hybrid viability or fertility, although hybrid fitness might be reduced by inappropriate signals or preferences. • If applicable, there is no character displacement39. Without the first element, there is no case for speciation by sexual selection. As noted earlier, sexual selection should be demonstrated directly from the effect of variation in the trait on mating success rather than simply being inferred from elaboration or sexual dimorphism. The second point is required because divergence under sexual selection does not necessarily result in a substantial barrier to gene exchange. The catch here is to demonstrate that prezygotic isolation is the direct result of the changes in sexually selected traits and not a result of other evolutionary forces (e.g. drift or natural selection) or evolutionary history. This might be difficult, but could be achieved through comparative studies using a phylogenetic approach to help separate independent evolutionary origins from historical effects resulting from common ancestry35 (see Barraclough and Nee46, this issue). Genetic ‘dissection’ of the mating traits important in prezygotic isolation has also been proposed47. This consists of hybridization studies between closely related species to analyse the genetic and phenotypic difference in male traits and female preferences. The third component might be necessary to help rule out pleiotropic effects of genetic divergence45. INCIPIENT SPECIES or sibling species could be useful systems to study speciation by sexual selection because they will, by definition, have experienced relatively little genetic divergence. Lack of postzygotic isolation is important for the same reason, as well as helping to exclude the possibility that prezygotic isolation has arisen secondarily rather than being the initial cause of speciation. The final component is important for studies of closely related species that overlap in their geographical range. By ruling out character displacement, a stronger case is made for the role of sexual selection in the speciation event39. These criteria are not applicable to all systems, but the objective should be to cover as many as possible. Some of the best examples from nature that suggest that sexual selection could be involved in the speciation process are listed in Table 1, but it is clear that each one of them falls short of being conclusive. The studies on cichlids suggest that sexual selection could be important in generating sympatric speciation18,48. Recent studies49,50 on greenish warblers (Phylloscopus trochiloides) imply that changes in the intensity of sexual selection on song are important in generating prezygotic isolation http://tree.trends.com between two sympatric subspecies forming the ends of a ring species. Character displacement is improbable because the degree of song variation between allopatric and sympatric populations is small within both subspecies. Work on two cryptic species of field crickets (Gryllus texensis and G. rubens) demonstrates that: (1) a change between species in male calling pulse rate is genetically correlated with a change in female preference; (2) the difference in male traits and female preferences leads to prezygotic isolation; (3) males are morphologically almost identical (only ovipositor length differs between females); (4) signs of postzygotic isolation are lacking; and (5) there is no evidence for character displacement39. However, even this last study was only able to conclude that the data are ‘substantially more consistent with speciation by sexual selection than they are with the alternatives’ of speciation by reinforcement or by ecological differentiation. Even when incipient species are described, speciation might not be completed by sexual selection. For example, in the Vogelkop bowerbirds (Amblyornis inornatus; Table 1) males from allopatric populations are morphologically and genetically very similar and only differ in their display traits; however, it is not known whether the female preference is symmetrical and whether it will result in prezygotic isolation45. Some have also argued that rapid radiation in the Hawaiian Drosophila can be explained by other evolutionary forces, such as drift followed by natural selection51, or that the sexually selected traits are not involved in species recognition40. Thus, sexual selection might not be as important in the origin of some of the Hawaiian Drosophila species pairs as it first seemed. Prospects There is no doubt that sexual selection has the potential to play a major role in speciation. Models of both allopatric and sympatric populations indicate that sexual selection has the power to drive rapid divergence and generate reproductive isolation (see Turelli et al.7, this issue). The evidence suggesting that theory translates to reality comes from comparative studies that repeatedly indicate higher species diversity in more sexually selected clades and from observations of patterns of divergence between populations and species that are consistent with the expectations of speciation by sexual selection. Further comparative studies are needed, examining extinction rates as well as numbers of extant species, confirming the role of sexual selection in the evolution of the traits used as indicators (such as plumage dimorphism in birds) and quantifying ecological and other sources of variation in species richness. Empirical work has taken a more direct approach to demonstrating speciation via sexual selection, but more examples are needed in which all elements of the sexual selection model have been demonstrated and that eliminate, so far as is possible, alternative factors such as character displacement and pleiotropic effects of divergence. Review References 1 Mayr, E. (1942) Systematics and the Origin of Species, Belknap Press 2 Dobzhansky, T. (1937) Genetics and the Origin of Species, Columbia University Press 3 West-Eberhard, M. (1983) Sexual selection, social competition, and speciation. Q. Rev. Biol. 58, 155–183 4 Coyne, J.A. (1992) Genetics and speciation. Nature 355, 511–515 5 Ligon, J.D. (1999) The Evolution of Avian Breeding Systems, Oxford University Press 6 Questiau, S. (1999) How can sexual selection promote population divergence? Ethol. Ecol. Evol. 11, 313–324 7 Turelli, M. et al. (2001) Theory and speciation. Trends Ecol. Evol. 16, 330–343 8 Darwin, C. (1871) The Descent of Man and Selection in Relation to Sex, John Murray 9 Barraclough, T.G. et al. (1995) Sexual selection and taxonomic diversity in passerine birds. Proc. R. Soc. London B Biol. Sci. 259, 211–215 10 Mitra, S. et al. (1996) Species richness covaries with mating system in birds. Auk 113, 544–551 11 Møller, A.P. and Cuervo, J.J. (1998) Speciation and feather ornamentation in birds. Evolution 52, 859–869 12 Hodges, S.A. and Arnold, M.L. (1995) Spurring plant diversification: are floral nectar spurs a key innovation? Proc. R. Soc. London B Biol. Sci. 262, 343–348 13 Arnqvist, G. et al. (2000) Sexual conflict promotes speciation in insects. Proc. Natl. Acad. Sci. U. S. A. 97, 10460–10464 14 Arnqvist, G. (1998) Comparative evidence for the evolution of genitalia by sexual selection. Nature 393, 784–786 15 Civetta, A. and Singh, R.S. (1998) Sex-related genes, directional sexual selection, and speciation. Mol. Biol. Evol. 15, 901–909 16 Owens, I.P.F. et al. (1999) Species richness among birds: body size, life history, sexual selection or ecology? Proc. R. Soc. London B Biol. Sci. 266, 933–939 17 Price, T. (1998) Sexual selection and natural selection in bird speciation. Philos. Trans. R. Soc. London Ser. B 353, 251–260 18 Seehausen, O. and van Alphen, J.M. (1999) Can sympatric speciation by disruptive sexual selection explain rapid evolution of cichlid diversity in Lake Victoria? Ecol. Lett. 2, 262–271 19 Via, S. (2001) Sympatric speciation in animals: the ugly duckling grows up. Trends Ecol. Evol. 16, 381–390 20 Butlin, R.K. (1995) Genetic variation in mating signals and responses. In Speciation and the Recognition Concept: Theory and Applications (Lambert, D.M. and Spencer, H.G., eds), pp. 327–366, Johns Hopkins University Press 21 Bakker, T.C.M. and Pomiankowski, A. (1995) The genetic basis of female mate preferences. J. Evol. Biol. 8, 129–171 22 Arnold, S.J. et al. (1996) The evolution of asymmetry in sexual isolation: a model and a test case. Evolution 50, 1024–1033 23 Tregenza, T. et al. (2000) The origins of premating reproductive isolation: testing hypotheses in the grasshopper Chorthippus parallelus. Evolution 54, 1687–1698 24 Butlin, R.K. and Ritchie, M.G. (1991) Variation in http://tree.trends.com TRENDS in Ecology & Evolution Vol.16 No.7 July 2001 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 female mate preference across a grasshopper hybrid zone. J. Evol. Biol. 4, 227–240 Wells, M.M. and Henry, C.S. (1998) Songs, reproductive isolation, and speciation in cryptic species of insects. In Endless Forms: Species and Speciation (Howard, D.J. and Berlocher, S.H., eds), pp. 217–233, Oxford University Press Wu, C-I. et al. (1995) Sexual isolation in Drosophila melanogaster: a possible case of incipient speciation. Proc. Natl. Acad. Sci. U. S. A. 92, 2519–2523 Sanderson, N. et al. (1992) Variation in mating call across the hybrid zone between the firebellied toads Bombina bombina and Bombina variegata. Evolution 46, 595–607 Hill, G.E. (1994) Geographic variation in male ornamentation and female mate preference in the house finch: a comparative test of models of sexual selection. Behav. Ecol. 5, 63–73 Endler, J.A. and McLellan, T. (1988) The process of evolution: towards a newer synthesis. Annu. Rev. Ecol. Syst. 19, 385–421 Ryan, M.J. (1990) Signals, species, and sexual selection. Am. Sci. 78, 46–52 Basolo, A.L. (1990) Female preference predates the evolution of the sword in swordtail fish. Science 250, 808–810 Kime, N.M. et al. (1998) Consistency of female choice in the tungara frog: a permissive preference for complex characters. Anim. Behav. 55, 641–649 Ryan, M.J. and Rand, A.S. (1993) Species recognition and sexual selection as a unitary problem in animal communication. Evolution 47, 647–657 McClintock, W.J. and Uetz, G.W. (1996) Female choice and pre-existing bias: visual cues during courtship in two Schizocosa wolf spiders (Araneae: Lycosidae). Anim. Behav. 52, 167–181 Ptacek, M.B. (2000) The role of mating preferences in shaping interspecific divergence in mating signals in vertebrates. Behav. Processes 51, 111–134 Ryan, M.J. and Rand, A.S. (1993) Sexual selection and signal evolution – the ghost of biases past. Philos. Trans. R. Soc. London Ser. B 340, 187–195 Schluter, D. and Price, T. (1993) Honesty, perception and population divergence in sexually selected traits. Proc. R. Soc. London B Biol. Sci. 253, 117–122 Butlin, R.K. (1996) Co-ordination of the sexual signaling system and the genetic basis of differentiation between populations in the brown planthopper, Nilaparvata lugens. Heredity 77, 369–377 Gray, D.A. and Cade, W.H. (2000) Sexual selection and speciation in field crickets. Proc. Natl. Acad. Sci. U. S. A. 97, 14449–14454 Boake, C.R.B. et al. (1997) Is sexual selection and species recognition a continuum? Mating behavior of the stalk-eyed fly Drosophila heteroneura. Proc. Natl. Acad. Sci. U. S. A. 94, 12442–12445 Liou, L.W. and Price, T.D. (1994) Speciation by reinforcement of premating isolation. Evolution 48, 1451–1459 Wiernasz, D.C. and Kingsolver, J.G. (1992) Wing melanin pattern mediates species recognition in Pieris occidentalis. Anim. Behav. 43, 89–94 Kaneshiro, K.Y. (1988) Speciation in the Hawaiian Drosophila. BioScience 38, 258–263 Meyer, A. (1993) Phylogenetic relationships 371 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 and evolutionary processes in East African cichlid fishes. Trends Ecol. Evol. 8, 279–284 Uy, J.A.C. and Borgia, G. (2000) Sexual selection drives rapid divergence in bowerbird display traits. Evolution 54, 273–278 Barraclough, T.G. and Nee, S. (2001) Phylogenetics and speciation. Trends Ecol. Evol. 16, 391–399 Shaw, K.L. (2000) Specific genetics of mate recognition: inheritance of female acoustic preference in Hawaiian crickets. Evolution 54, 1303–1312 Galis, F. and Metz, A.J. (1998) Why are there so many cichlid species? Trends Ecol. Evol. 13, 1–2 Irwin, D.E. (2000) Song variation in an avian ring species. Evolution 54, 998–1010 Irwin, D.E. et al. (2001) Speciation in a ring. Nature 409, 333–337 Templeton, A.R. (1979) Once again, why 300 species of Hawaiian Drosophila? Evolution 33, 513–517 Hollocher, H. et al. (1997) Incipient speciation by sexual isolation in Drosophila melanogaster: variation in mating preference and correlation between sexes. Evolution 51, 1175–1181 Korol, A. et al. (2000) Nonrandom mating in Drosophila melanogaster laboratory populations derived from closely adjacent ecologically contrasting slopes at ‘Evolution Canyon’. Proc. Natl. Acad. Sci. U. S. A. 97, 12637–12642 Carson, H.L. (1985) Genetic variation in a courtship related male character in Drosophila silvestris from a single Hawaiian locality. Evolution 36, 678–686 Ritchie, M.G. (1996) The shape of female mating preferences. Proc. Natl. Acad. Sci. U. S. A. 93, 14628–14631 LaForest, S. et al. (1997) A genetic analysis of population differences in pheromone production and response between two populations of the turnip moth, Agrotis segetum. J. Chem. Ecol. 23, 1487–1503 Ryan, M.J. and Wilczynski, W. (1988) Coevolution of sender and receiver: effects on local mate preference in cricket frogs. Science 240, 1786–1788 Ryan, M.J. and Wilczynski, W. (1991) Evolution of intraspecific variation in the advertisement call of a cricket frog (Acris crepitans, Hylidae). Biol. J. Linn. Soc. 44, 249–271 Ryan, M.J. et al. (1992) Auditory tuning and call frequency predict population-based mating preferences in the cricket frog, Acris creptians. Am. Nat. 139, 1370–1383 Endler, J.A. and Houde, A.E. (1995) Geographic variation in female preferences for male traits in Poecilia reticulata. Evolution 49, 456–468 Young, J.R. et al. (1994) Phenotypic divergence of secondary sexual traits among sage grouse, Centrocercus urophasianus, populations. Anim. Behav. 47, 1353–1362 Young, J.R. et al. (2000) A new species of Sage Grouse (Phasianidae: Centrocercus) from Southwestern Colorado. Wilson Bull. 112, 445 Kahn, N.W. et al. (1999) Molecular analysis of genetic variation among large- and small-bodied Sage Grouse using mitochondrial control-region sequences. Auk 116, 819–824 Begun, D.J. and Aquadro, C.F. (1993) African and North American populations of Drosophila melanogaster are very different at the DNA level. Nature 365, 548–550