- Top of page
- Conflict of Interest
- Supporting Information
Mate preferences for ecologically relevant traits have important consequences for local adaptation and speciation. Females often prefer mates that are in good phenotypic condition or that express traits which indicate high genetic quality (Iwasa and Pomiankowski 1991; Kirkpatrick and Ryan 1991; Pfennig 1998; Andersson and Simmons 2006). This is because these males can provide benefits to females that are either direct, such as increased fertility or parental care (Price et al. 1993; Kirkpatrick 1996; Moller and Jennions 2001) or indirect, such as genes that confer high fitness for offspring (Fisher 1930; Zahavi 1975). Many ecologically relevant traits may be reliable signals of a male's ability to acquire resources (Snowberg and Benkman 2009). Therefore, female preferences for traits that indicate male condition or genetic quality may target traits that are directly involved in adaptation to local environments (Lorch et al. 2003; Servedio 2004). This is because well-adapted males are likely to be of the best condition and quality. In this way, preferences for ecologically relevant traits may potentially drive local adaptation (sensu Kawecki and Ebert 2004) between populations.
Female preferences for ecologically relevant traits may also be important for the evolution of reproductive isolation and speciation (Servedio 2004; van Doorn et al. 2009). When populations or species adapt to new environments, natural selection will drive divergence in ecologically relevant traits generating either direct or indirect selection on female mate preferences if they are based on traits that are important in ecological adaptation. Thus, female preferences can track ecological divergence resulting in divergent preferences (Servedio 2004). When diverging populations subsequently come into contact, preferences for ecologically relevant traits will lead females to discriminate against males which have adapted to a different environment, generating sexual isolation (van Doorn et al. 2009). Therefore, preferences for ecologically relevant traits or well-adapted mates provide a way to link divergent natural selection with divergence in mating traits and mate preferences, which greatly facilitates ecological speciation even in the face of gene flow (Gavrilets 2005; Schluter and Conte 2009; Servedio et al. 2011).
Traits that are involved in both divergent selection due to adaptation to different environments and assortative mating between species have been termed ‘magic traits’ (Gavrilets 2004). Magic traits may facilitate speciation because the two functions of these traits are pleiotropically linked and cannot become disassociated via recombination (Servedio et al. 2011). Typically, magic traits are thought to arise through divergent natural selection acting on traits that are also used as mating cues, such as color or body size (e.g., Jiggins et al. 2001; Podos 2001). Putative magic traits have been identified in a number of species pairs (Servedio et al. 2011; Nosil 2012). However, in most cases, the basis of assortative mating and origin of female preferences for these traits is unknown. Sexual selection within species for well-adapted males could provide one mechanism to generate a preference for traits under natural selection (Servedio et al. 2011). The combined force of natural and sexual selection acting on preference for ecologically relevant traits may make these preferences a powerful driver of speciation (Ritchie 2007; Maan and Seehausen 2011).
Here, we investigate female mate preferences for two ecologically important traits, body shape and body size which have diverged in the threespine stickleback species complex (Gasterosteus aculeatus spp.) (Fig. 1). In the benthic–limnetic stickleback species complex, an extensive body of work has demonstrated that size and shape differences in sticklebacks result from adaptation to local foraging and predator environments (reviewed in McKinnon and Rundle 2002; Reid and Peichel 2010). Divergent natural selection has led to local adaptation of body shape and body size at two levels. First, benthic and limnetic sticklebacks have diverged from their anadromous ancestors by adapting to freshwater environments (McPhail 1994; Taylor and McPhail 2000; Walker and Bell 2000; Aguirre 2009). In general, anadromous fish are larger, have shorter heads and larger keels than freshwater fish; adaptations for long distance migration and predator avoidance in anadromous fish which are subsequently lost upon the colonization of freshwater environments (Taylor and McPhail 1986; McPhail 1994; Walker 1997; Walker and Bell 2000; Dalziel et al. 2012). Second, in freshwater lakes with extensive littoral and pelagic habitats, benthics and limnetics have diverged from each other in body size and shape during adaptation to different niches (benthic sticklebacks are specialized for feeding on macroinvertebrates in the littoral zone and limnetic sticklebacks are specialized for feeding on plankton in the pelagic zone), thus showing fine scale local adaptation. Evidence that differences in body size and shape between benthics and limnetics arise from adapting to different ecological niches is extensive. Benthics are large, deep-bodied, with a wide gape; this shape makes them highly maneuverable (Webb 1984; Walker 1997; Blake 2004) and enables them to forage efficiently on invertebrates located on the sediment or vegetation (Schluter 1993) and escape their grasping sit and wait predators (Walker 1997). In contrast, limnetics are small, slender, with a narrow upturned mouth; this shape gives them good prolonged swimming ability (Webb 1984; Walker 1997; Blake 2004) and enables them to forage efficiently on zooplankton in the open water and escape their pursuit predators (McPhail 1984, 1992; Malmquist 1992; Schluter and Mcphail 1992).
Figure 1. The three-spine stickleback species used in our study. (A) Anadromous male, (B) benthic male, (C) limnetic male.
Download figure to PowerPoint
Body shape and body size make ideal candidates for female preferences based on ecologically relevant traits due to their role in adaptation to different foraging niches and to different suites of predators. What gives rise to female preferences for size remains poorly understood despite the fact that body size has been proposed as a putative magic trait in sticklebacks, under both natural and sexual selection (McKinnon et al. 2004; Boughman et al. 2005; Conte and Schluter 2013). Body size has been shown to be important in reproductive isolation between benthics and limnetics, with heterospecific mating being more common between similar sized individuals (Nagel and Schluter 1998; Boughman et al. 2005; Conte and Schluter 2013). However, females do not show preferences for body size within species (Head et al. 2009), suggesting that size-based assortative mating is not an outcome of sexual selection on size within species. One alternative is that size-based assortative mating may have evolved in response to selection on correlated characters such as body shape. Surprisingly, the role of shape for within or between species mate choice has not been considered in sticklebacks despite its clear role in adaptation and likely link with body size.
Our aims are threefold. First, to set the stage for further analyses, we confirm body shape and size differences between benthic and limnetic sticklebacks and their anadromous ancestor that have been found in previous studies (e.g., McPhail 1992; Taylor et al. 2005). Next, we use this information to investigate the relationship between body shape and size and determine whether this relationship is the same for all species. This analysis is important in the context of the current study, as body size and shape are tightly linked traits and one could evolve as a correlated response to selection on the other, or selection may act on combinations of these traits. Identifying the specific traits under selection aids our understanding of preference evolution and its contribution to sexual isolation (Servedio 2004; Servedio et al. 2011). Finally, we look at female preferences for body shape and size in both conspecific and heterospecific mating trials, to gain insight into the evolutionary processes that lead to reproductive isolation between the species. Thus, we ask if anadromous, benthic, and limnetic females have preferences for body size, body shape, or some combination of these traits. We test if these preferences are the same in all species or if they have diverged between species. We also investigate whether benthic and limnetic preferences lead to sexual isolation by comparing female preferences when choosing mates within and between species. In this way, we can determine if females have preferences for size and shape, traits clearly involved in adaptation, and whether these preferences may contribute to reproductive isolation.
- Top of page
- Conflict of Interest
- Supporting Information
How females use ecologically relevant traits when choosing mates helps us understand how natural and sexual selection may interact to cause speciation (Ritchie 2007; Maan and Seehausen 2011). Body size and shape in sticklebacks are known adaptations to foraging and escaping predators in different environments (reviewed in introduction). Here, we show that stickleback females have preferences for male body size, shape, or a combination of the two, but that these preferences differ depending on the females’ own species identity. When assessing conspecific males, anadromous females base their mate choice on a combination of size and shape. Conspecific preferences have diverged from this ancestral preference in both benthic and limnetic sticklebacks. Benthic females choose mates based predominantly on size while limnetics choose mates based predominantly on shape. Therefore, our results from conspecific trials suggest that stickleback females do have mate preferences for ecologically relevant traits, but these differ between populations. Our results from heterospecific trials suggest that limnetic preferences for shape are used to discriminate against benthic males and thus shape preferences contribute to sexual isolation. This preference for shape is an incomplete isolating barrier, but likely contributes to strong sexual isolation between species when combined with preferences for other traits such as nuptial color and odor.
Male body size and shape divergence between species
Our results show that there has been significant shape divergence between anadromous, benthic, and limnetic sticklebacks, confirming results of previous studies (e.g., Walker and Bell 2000; Aguirre 2009). A novel aspect of our shape analysis is that we show that male shape differences between species are not simply due to species differences in body size, and that allometric relationships between size and shape have also diverged. The differences in both the slope and intercept of allometry that we find between species suggests that divergence in body shape is unlikely to be caused by a correlated response to selection on body size (McGuigan et al. 2010). This result is important because distinguishing the patterns of variation and covariation in size and shape will help to identify the nature of selection on these traits, and how their evolutionary change affects reproductive isolation.
Species-specific relationships between size and shape also suggest there may be natural selection for certain shape/size combinations in fish, possibly due to locomotor performance. Positive allometry between DF1 and body size appears to be driven mostly by anadromous males (Fig. 4A), with larger males having relatively smaller heads and deeper caudal peduncles. This same relationship occurs in Alaskan anadromous sticklebacks (McGuigan et al. 2010; Wund et al. 2012). The fact that this allometric slope decreases in lake fish (particularly for sedentary benthics) may indicate that allometry is beneficial for long distance swimming performance (Webb 1984; Blake 2004). Such ecologically dependent relationships between shape and size may have important consequences for the evolution of reproductive isolation. Reduced hybrid performance in parental foraging niches is a key mechanism of postmating reproductive isolation in ecological speciation, and is known to be important in stickleback speciation (Schluter 1993; Hatfield and Schluter 1999; Rundle 2002). This may occur not only because hybrids have intermediate phenotypes (Schluter 1993) but also due to an uncoupling of adaptive trait combinations of body size with body shape or trophic morphology. Mismatched trait combinations reducing hybrid performance are likely to be important in a broad range of taxa including lizards (Lancaster et al. 2010) and even plants (Melendez-Ackerman and Campbell 1998).
Anadromous female preferences for body size and shape
Anadromous stickleback females had preferences for both size and shape: females disliked anadromous males that were large and limnetic shaped. We hypothesize that this avoidance of large slender males may reflect a preference for males that are in good condition (because males that are “skinny” for their size are likely to be malnourished). Alternatively, slender males may be avoided because they are at a fitness disadvantage: because large fusiform bodies increase prolonged swimming performance required for migration (Walker 1997) and deep bodies decrease predation by gape limited predators (Reimchen 1994). Anadromous female preference may also reflect a combination of the two, because condition is likely to reflect local adaptation (Weissing et al. 2011). Future work examining the benefits that anadromous females who mate with smaller, deep-bodied males gain would help to solve this question.
Anadromous female mate preferences for body size and body shape may be key to the rapid morphological evolution seen when these fish colonize freshwater habitats. Anadromous populations of sticklebacks have colonized freshwater streams and lakes repeatedly throughout the Northern Hemisphere, resulting in freshwater populations with great variation in size and shape (e.g., Walker 1997; Leinonen et al. 2006; Spoljaric and Reimchen 2008). Many of the same phenotypes are seen repeatedly in similar habitats, implicating selection in their evolution. Further, several studies have shown that divergence in body size and shape can occur rapidly (within 8–11 generations – Aguirre and Bell 2012). Ancestral preferences for size and shape can complement divergent natural selection for these same traits, hastening the pace of evolutionary change because of dual effects of natural and sexual selection (Ritchie 2007) and coupling adaptation to assortative mating (Servedio 2004). These ancestral preferences also allow mate preferences to track natural selection in novel environments. Thus, mate preferences may enhance natural selection and lead to rapid adaptation to new environments (Fricke and Arnqvist 2007). This may be particularly important for increasing the likelihood of speciation when populations are geographically isolated for short periods of time before coming into secondary contact. Evidence from other species radiations also shows ecological divergence and mate preferences for ecologically relevant traits (e.g., beak size in the medium ground finch – Huber et al. 2007; color pattern in Heliconius butterflies – Merrill et al. 2011). This may indicate that mate preferences for ecologically relevant traits may provide a general mechanism for rapid divergence.
Benthic and limnetic female preferences and sexual isolation
Benthic and limnetic mate preferences for body size and body shape have diverged from those of their anadromous ancestor and from each other. Essentially, each of these species continues to base mate choice decisions on these ecologically relevant traits, however, the relative importance of body size and body shape to mate choice have changed (benthic females focus on size, while limnetic females focus on shape). This suggests each species has diverged along a different axis of ancestral mate preference variation. There are several possible reasons for this divergence in preference between benthics and limnetics. One hypothesis is that mate preferences are for well-adapted mates and tracking traits that are most important in local adaptation in each species (Servedio 2004). Large, deep bodies play a key role in avoiding predation and navigating in benthic environments, so potentially benthic females benefit from preferring large males. Likewise, limnetic shape is an adaptation to foraging on plankton in the open water and limnetic females may benefit from preferring males with the most extreme and presumably well-adapted shape. Another reason for this difference between species may reflect differences in the reliability of these traits for signaling species identity. Benthics and limnetics live and mate in the same lakes, as such avoiding heterospecific mates is a constant problem. As benthics are larger and deeper bodied than limnetics, a preference for large males may enable avoidance of small limnetic males. For limnetics, on the other hand, female preferences based on size may lead to conflict between choice for good quality conspecific males and avoidance of heterospecifics (Pfennig 1998) if there is a relationship between quality and size and small limnetics are low quality. As such selection may target a female preference for shape which may reduce such conflict. These ideas deserve to be tested in future work, along with possible mechanisms that lead to female preferences for size and shape (e.g., phenotype matching [Conte and Schluter 2013] or learning [Kozak et al. 2011]).
Mate preferences for ecologically relevant traits may have important consequences for the evolution of reproductive isolation. This is because within-species preferences for such traits can be easily extended to between species mate choice and lead to assortative mating (Servedio 2004; van Doorn et al. 2009; Weissing et al. 2011). Limnetic female preference for limnetic-shaped males leads these females to reject benthics, particularly those that are very benthic like in shape. A preference for limnetic-shaped males therefore contributes to sexual isolation and provides the building blocks for the evolution of magic traits. However, within-species female preferences do not necessarily lead to the evolution of a magic trait. Unlike limnetic mate preferences for shape, benthic preferences for large body size did not extend to between species trials (despite overlap in the body size of limnetic and benthic males). This result contrasts to that found in previous studies, which suggested that body size is important for reproductive isolation between these stickleback species (Nagel and Schluter 1998; Boughman et al. 2005; Conte and Schluter 2013). Based on those studies, many have hypothesized that body size operates as a magic trait in benthic and limnetic sticklebacks. However, these studies used standard length (rather than centroid size) and did not control for differences in shape, as such, assortative mating based on size found previously may be confounded with differences in shape between species. Furthermore, body size may contribute to assortative mating through other mechanisms such as male mate choice (Kozak et al. 2009)) or mate choice for nest characteristics that correlate with body size (Wong et al. 2012). Investigating how traits that are correlated with body size affect assortative mating may aid in determining the role that body size plays in sexual isolation of these species as well as other sticklebacks species where manipulative experiments have shown that body size contributes to sexual isolation (e.g., McKinnon et al. 2004; Furin et al. 2012).
In sticklebacks, female preferences for multiple traits appear to be important for sexual isolation. Our result that limnetic female preferences for body shape contribute to sexual isolation adds to previous research which has found that color (Boughman 2001) and odor (Rafferty and Boughman 2006) are important for sexual isolation between benthic and limnetic sticklebacks. Interestingly, like for color and odor (Boughman 2001; Rafferty and Boughman 2006), the way in which shape contributes to sexual isolation is asymmetrical between species. Limnetic females are isolated from benthic males due to a combination of shape and nuptial color, while benthic females are isolated from limnetic males via preferences for odor. Work from other systems suggests that reproductive isolating barriers are typically asymmetrical, including both sexual isolation (Yukilevich 2012) and hybrid inviability (Bolnick et al. 2008; Schrader and Travis 2008). Thus, asymmetrical barriers may be the norm rather than the exception and reproductive isolation may typically accrues due to the combined action of many barriers.