Geographically separated populations accumulate genetic and phenotypic differences through genetic drift and/or adaptation to local environmental conditions [1]. The roles of local adaptation and consequent divergent selection between environments on the evolution of mate preferences are widely recognized in the literature and have been documented in diverse taxonomic groups such as snails [2], crustaceans [3, 4], insects [5-8], fish [9-14], amphibians [1, 15-17], reptiles [18], birds [19, 20] and mammals [21, 22]. From previous studies on mate preference in females, which account for the vast majority of published papers focusing on this issue, we know the following. First, mate preferences are often evaluated with respect to behaviors associated with their intentions of rejecting and/or accepting males rather than actual copulations with males [18]. Second, local adaptations are more valuable than genetic novelties in most of the species studied thus far, as revealed by the fact that females of these species prefer to mate with native (local) over foreign males [2]. To our knowledge, the Trinidadian guppy Poecilla reticulata [9], the sailfin molly Poecilia latipinna [13], the striped mouse Rhabdomys pumilio [22], the Pacific blue-eye Pseudomugil signifer [14] and the rainwater killifish Lucania parva [11] represent the exceptions of species where female mate preference for native males is strong in some populations but not in others, or asymmetric in different populations. Third, comprehensive studies integrating molecular, ecological, phenotypic and behavioral approaches to show changes in mate preference correlated with local adaptation are rare. To our knowledge, such studies have been performed only in P. signifer [14], the Allegheny mountain salamander Desmognathus ochrophaeus [17] and the fire salamander Salamandra salamandra [15] where female mate preferences depend either on local adaptation rather than neutral genetic distance (S. salamandra), or on how genetically and geographically separated populations are from one another (P. signifer and D. ochrophaeus).
Divergence among populations in mate preferences is a crucial step of population differentiation and potential speciation because it may lead to premating reproductive isolation [23, 24]. Identifying the factors that contribute to such a divergence is therefore important for understanding of the manner in which premating reproductive isolation might arise and how this isolation may in turn contribute to the evolutionary process of population divergence and incipient speciation [25]. However, as the roles of individual factors in mate assessment by males or females are not mutually exclusive, disentangling their effects on mate preferences is difficult. Fortunately, the recent progress in statistical methods for spatial data analysis and the increasing availability of high-resolution geographical and environmental data layers make it possible to evaluate the relative contributions of the factors to population divergence in mate preferences.
The northern grass lizard (Takydromus septentrionalis) is a multiple-clutched oviparous lacertid lizard endemic to China and has a distributional range covering the central and southeastern parts of the country [26]. The lizard consists of three divergent lineages, with isolation by distance known to be the main cause of genetic divergence [27]. Proximate factors are less important determinants of spatio-temporal variation in life-history traits (e.g., size at maturation, adult size, clutch size, clutch frequency and egg size) than are genetic influences [28-31]. Adults can have up to three distinct color morphs within a population (Additional file 1). Morph 1 is shared by both sexes, with a yellowish-white longitudinal stripe between brown dorsal and green lateral surfaces. Morph 2 is a typical male coloration, with morph 1 decorated with numerous black speckles on the lateral surface. Morph 3 is a typical female coloration, with brown dorsal and green lateral surfaces connected directly. Although small adults do mate less frequently than large ones, the lizard does not show size-assortative mating [32]. Males are the less choosy sex, as revealed by three lines of evidence. First, they can be easily induced to mate at extreme phylogenetic distances, either with conspecific females from other populations, or even with heterospecific (but congeneric) females (e.g., T. sexlineatus) (Fig. 1a), which is an extreme example of outcrossing. Second, males are often found to mate with one another in a manner adopted by heterosexual mating partners (Fig. 1b). Third, female receptivity rather than male sexual coercion (and thus forced insemination) has a direct role in determining mating success, and females are more receptive to mate within 2 d post-laying and much less so thereafter [32]. Here, we describe a study offering each post-laying female T. septentrionalis 10 males, two from each of five geographically separated populations (see below for details), to test whether the species also shows a female mate preference for local males and, if so, whether the preferences that females display in the mating trials correlate with phenotypic adaptation to local environments, or to the neutral genetic distance measured by divergence of mitochondrial DNA sequence loci.