The University of Chicago Weak Disruptive Selection and Incomplete Phenotypic Divergence in Two Classic Examples of Sympatric Speciation: Cameroon Crater Lake Cichlids. Author(s): Christopher H. Martin Reviewed work(s): Source: The American Naturalist, Vol. 180, No. 4 (October 2012), pp. E90-E109 Published by: The University of Chicago Press for The American Society of Naturalists Stable URL: http://www.jstor.org/stable/10.1086/667586 . Accessed: 14/09/2012 14:38 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. The University of Chicago Press, The American Society of Naturalists, The University of Chicago are collaborating with JSTOR to digitize, preserve and extend access to The American Naturalist. http://www.jstor.org vol. 180, no. 4 the american naturalist october 2012 E-Article Weak Disruptive Selection and Incomplete Phenotypic Divergence in Two Classic Examples of Sympatric Speciation: Cameroon Crater Lake Cichlids Christopher H. Martin* Department of Evolution & Ecology and Center for Population Biology, University of California, Davis, California 95616 Submitted February 2, 2012; Accepted May 21, 2012; Electronically published August 24, 2012 Online enhancement: appendix (PDF file). Dryad data: http://dx.doi.org/10.5061/dryad.rn30d. in Via 2000; Turelli et al. 2001; Coyne and Orr 2004; abstract: Recent documentation of a few compelling examples of Gavrilets 2004; Bolnick and Fitzpatrick 2007). This end- sympatric speciation led to a proliferation of theoretical models. Unfortunately, plausible examples from nature have rarely been used point on the speciation-with-gene-flow continuum is tra- to test model predictions, such as the initial presence of strong dis- ditionally defined geographically as individuals in a pop- ruptive selection. Here I estimated the form and strength of selection ulation within dispersal range of each other (Mallet et al. in two classic examples of sympatric speciation: radiations of Cam- 2009) or, from a population genetic viewpoint, as the evo- eroon cichlids restricted to Lakes Barombi Mbo and Ejagham. I lution of reproductive isolation within an initially pan- measured five functional traits and relative growth rates in over 500 individuals within incipient species complexes from each lake. Dis- mictic population (Fitzpatrick et al. 2008). Despite its ap- ruptive selection was prevalent in both groups on single and mul- parent rarity in nature, dependent on both the spatial scale tivariate trait axes but weak relative to stabilizing selection on other of gene flow (Kisel and Barraclough 2010) and the rarity traits and most published estimates of disruptive selection. Further- of geographic scenarios in which it can be strongly inferred more, despite genetic structure, assortative mating, and bimodal spe- (Coyne and Orr 2004), sympatric speciation holds a pe- cies-diagnostic coloration, trait distributions were unimodal in both rennial fascination due to its formidable integration of species complexes, indicating the earliest stages of speciation. Long waiting times or incomplete sympatric speciation may result when natural and sexual selection, ecology, and population ge- disruptive selection is initially weak. Alternatively, I present evidence netics (Bolnick and Fitzpatrick 2007; Fitzpatrick et al. of additional constraints in both species complexes, including weak 2008). Moreover, it illustrates the power of natural selec- linkage between coloration and morphology, reduced morphological tion to generate biodiversity without geographic interven- variance aligned with nonlinear selection surfaces, and minimal eco- tion (Darwin 1859). logical divergence. While other species within these radiations show Models of sympatric speciation by natural selection gen- complete phenotypic separation, morphological and ecological di- vergence in these species complexes may be slow or incomplete out- erally require three initial conditions for the splitting of side optimal parameter ranges, in contrast to rapid divergence of ecological subgroups within a population: (1) strong dis- their sexual coloration. ruptive selection on ecological traits, (2) strong assortative mating by ecotype, and (3) the buildup of linkage dis- Keywords: adaptation, adaptive radiation, disruptive selection, selec- equilibrium between mating and ecological loci (Kirkpat- tion gradient, diversification, ecological opportunity, fitness land- scape, sexual selection, adaptive dynamics, trophic competition. rick and Ravigne 2002; Bolnick and Fitzpatrick 2007). The central requirement is that disruptive selection on ecotypes within a freely interbreeding population is strong enough Introduction to drive the evolution of nonrandom mating with respect to ecotype, thus increasing linkage between ecotype and After 150 years of contention, the theoretical possibility mate choice loci and ultimately reproductive isolation be- and existence in nature of at least a few plausible examples tween ecotypes (pleiotropic traits affecting ecotype and of sympatric speciation is now widely accepted (reviewed assortative mating can also circumvent the obstacle of link- * E-mail: [email protected]. age; Servedio et al. 2011). Disruptive selection arises either from frequency-dependent competition for shared re- Am. Nat. 2012. Vol. 180, pp. E90–E109. ᭧ 2012 by The University of Chicago. 0003-0147/2012/18004-53625$15.00. All rights reserved. sources (Roughgarden 1972; Bolnick 2004a; Pfennig and DOI: 10.1086/667586 Pfennig 2010), performance trade-offs resulting from ad- Nonlinear Selection in Cameroon Cichlids E91 aptation to different resources (Wilson and Turelli 1986; of components and parameter ranges in these systems (Gav- Martin and Pfennig 2009), or the uneven distribution of rilets et al. 2007), relative to estimates in systems where resources in the environment (Schluter and Grant 1984; sympatric speciation has not occurred (Bolnick 2011), can Hendry et al. 2009). Genetic homogenization due to ran- help cull the many existing theoretical models and provide dom mating is dependent on the effective recombination realistic parameter values for future efforts. Ultimately, this rate (Via 2009); the rate of decay of linkage disequilibrium approach should lead to an understanding of the necessary (Flint-Garcia et al. 2003); the number of genetic loci un- and sufficient components and parameter space for sym- derlying ecotypes, mate choice, and ecotype marker traits patric speciation in nature. (Bolnick and Fitzpatrick 2007); and the strength of as- One complication of this retrospective approach is that sortative mating (Dieckmann and Doebeli 1999; Otto et speciation models estimate the initial conditions for spe- al. 2008; van Doorn et al. 2009). ciation to proceed, whereas in most empirical systems, Beyond these three necessary conditions, the impor- speciation is already under way or has completed. For tance of additional factors is unknown. For example, if example, models predict that strong disruptive selection ecotype divergence automatically results in assortative is necessary to initiate sympatric speciation, but the fitness mating (Berlocher and Feder 2002) or generates some re- surface begins to flatten as phenotypic variance increases, productive isolation as a by-product (automatic and classic resulting in weak or absent disruptive selection after phe- magic traits sensu Servedio et al. 2011), then sympatric notypic separation of ecological subgroups (Dieckmann speciation is relatively easy (Dieckmann and Doebeli and Doebeli 1999; Bolnick and Doebeli 2003; also see dis- 1999). Indeed, sympatric speciation facilitated by magic cussion of the ghost of competition past: Connell 1980). traits may be common (Berlocher and Feder 2002; Soren- Thus, interpreting these estimates of selection additionally son et al. 2003; Papadopulos et al. 2011). In contrast, requires an understanding of the progress of speciation sympatric speciation should be more difficult when eco- within each system; ideally, species in the earliest stages of type, species-specific markers, and mate choice loci are all divergence should provide the best estimates of the initial initially unlinked (Dieckmann and Doebeli 1999), but the conditions of sympatric speciation. occurrence of this form of sympatric speciation in nature Here I estimated the form and strength of natural se- is unknown because it requires ruling out the presence of lection in the flagship example of sympatric speciation, any magic traits, which may be very common (Servedio the Cameroon crater lake cichlid radiations: Coyne and et al. 2011). Orr (2004, p. 152) state, “We know of no more convincing Additional components may also be necessary for sym- example [of sympatric speciation] in any group.” Indeed, patric speciation, such as sexual selection (van Doorn et the three independent sympatric radiations of Cameroon al. 2009; Maan and Seehausen 2011), competition versus cichlids admirably fulfill Coyne and Orr’s (2004) stringent habitat preference (Bolnick and Fitzpatrick 2007), genomic criteria for sympatric speciation: each is a monophyletic islands of speciation