Introgression despite substantial divergence in a broadcast spawning marine invertebrate

Understanding the relationship between reproductive isolation and time since divergence is critical to our understanding of speciation. One group for which we know little about the relationship between hybridization/introgression and time since divergence is the marine broadcast spawners. Here, we investigate the distribution of closely related cryptic species of marine broadcast spawners (Type A and B Ciona intestinalis) in areas of potential sympatry to determine whether these two types occur together and if so, whether they show evidence of hybridization and introgression. Then we combine our data with other studies to investigate general patterns of reproductive isolation versus divergence in marine broadcast spawners. We found that Type A and B C. intestinalis occurred sympatrically in 2007, and that 21 individuals show evidence of introgression in sympatry (out of approximately 500). Type A and B C. intestinalis are 12.4% divergent at mitochondrial COI (mtCOI), and in comparison with other marine broadcast spawning species at mtCOI, these two types may be near the upper limit of the range of divergence values in which introgression is still possible. However, introgression at divergence levels similar to those found in Ciona does exist, prompting questions about the strength of postmating prezygotic reproductive barriers in marine broadcast spawners.     

Sexual dimorphism and adaptive speciation: two sides of the same ecological coin

Models of adaptive speciation are typically concerned with demonstrating that it is possible for ecologically driven disruptive selection to lead to the evolution of assortative mating and hence speciation. However, disruptive selection could also lead to other forms of evolutionary diversification, including ecological sexual dimorphisms. Using a model of frequency-dependent intraspecific competition, we show analytically that adaptive speciation and dimorphism require identical ecological conditions. Numerical simulations of individual-based models show that a single ecological model can produce either evolutionary outcome, depending on the genetic independence of male and female traits and the potential strength of assortative mating. Speciation is inhibited when the genetic basis of male and female ecological traits allows the sexes to diverge substantially. This is because sexual dimorphism, which can evolve quickly, can eliminate the frequency-dependent disruptive selection that would have provided the impetus for speciation. Conversely, populations with strong assortative mating based on ecological traits are less likely to evolve a sexual dimorphism because females cannot simultaneously prefer males more similar to themselves while still allowing the males to diverge. This conflict between speciation and dimorphism can be circumvented in two ways. First, we find a novel form of speciation via negative assortative mating, leading to two dimorphic daughter species. Second, if assortative mating is based on a neutral marker trait, trophic dimorphism and speciation by positive assortative mating can occur simultaneously. We conclude that while adaptive speciation and ecological sexual dimorphism may occur simultaneously, allowing for sexual dimorphism restricts the likelihood of adaptive speciation. Thus, it is important to recognize that disruptive selection due to frequency-dependent interactions can lead to more than one form of adaptive splitting.     

Disruptive selection in a bimodal population of Darwin's finches

A key part of the ecological theory of adaptive radiation is disruptive selection during periods of sympatry. Some insight into this process might be gained by studying populations that are bimodal for dual-context traits, i.e. those showing adaptive divergence and also contributing to reproductive isolation. A population meeting these criteria is the medium ground finch (Geospiza fortis) of El Garrapatero, Santa Cruz Island, Galápagos. We examined patterns of selection in this population by relating individual beak sizes to interannual recaptures during a prolonged drought. Supporting the theory, disruptive selection was strong between the two beak size modes. We also found some evidence of selection against individuals with the largest and smallest beak sizes, perhaps owing to competition with other species or to gaps in the underlying resource distribution. Selection may thus simultaneously maintain the current bimodality while also constraining further divergence. Spatial and temporal variation in G. fortis bimodality suggests a dynamic tug of war among factors such as selection and assortative mating, which may alternatively promote or constrain divergence during adaptive radiation.     

The risk of polyspermy in three congeneric sea urchins and its implications for gametic incompatibility and reproductive isolation

Developmental failure caused by excess sperm (polyspermy) is thought to be an important mechanism driving the evolution of gamete-recognition proteins, reproductive isolation, and speciation in marine organisms. However, these theories assume that there is heritable variation in the susceptibility to polyspermy and that this variation is related to the overall affinity between sperm and eggs. These assumptions have not been critically examined. We investigated the relationship between ease of fertilization and susceptibility to polyspermy within and among three congeneric sea urchins. The results from laboratory studies indicate that, both within and among species, individuals and species that produce eggs capable of fertilization at relatively low sperm concentrations are more susceptible to polyspermy, whereas individuals and species producing eggs that require higher concentrations of sperm to be fertilized are more resistant to polyspermy. This relationship sets the stage for selection on gamete traits that depend on sperm availability and for sexual conflict that can influence the evolution of gamete-recognition proteins and eventually lead to reproductive isolation.     

Adaptive speciation when assortative mating is based on female preference for male marker traits

Adaptive speciation occurs when frequency-dependent ecological interactions generate conditions of disruptive selection to which lineage splitting is an adaptive response. Under such selective conditions, evolution of assortative mating mechanisms enables the break-up of the ancestral lineage into diverging and reproductively isolated descendent species. Extending previous studies, I investigate models of adaptive speciation due to the evolution of indirect assortative mating that is based on three different mating traits: the degree of assortativity, a female preference trait and a male marker trait. For speciation to occur, linkage disequilibria between different mating traits, e.g. between female preference and male marker traits, as well as between mating traits and the ecological trait, must evolve. This can lead to novel speciation scenarios, e.g. when reproductive isolation is generated by a splitting in the degree of assortativeness, with one of the emerging lineages mating assortatively, and the other one disassortatively. I investigate the effects of variation in various model parameters on the likelihood of speciation, as well as robustness of speciation to introducing costs of assortative mating. Even though in the models presented speciation requires the genetic potential for strong assortment as well as rather restrictive ecological conditions, the results show that adaptive speciation due to the evolution of assortative mating when mate choice is based on separate female preference and male marker traits is a theoretically plausible evolutionary scenario.     

Density‐dependent patterns of multivariate selection on sperm motility and morphology in a broadcast spawning mussel

Sperm cells exhibit extraordinary phenotypic variation, both among taxa and within individual species, yet our understanding of the adaptive value of sperm trait variation across multiple contexts is incomplete. For species without the opportunity to choose mating partners, such as sessile broadcast spawning invertebrates, fertilization depends on gamete interactions, which in turn can be strongly influenced by local environmental conditions that alter the concentration of sperm and eggs. However, the way in which such environmental factors impact phenotypic selection on functional gamete traits remains unclear in most systems. Here, we analyze patterns of linear and nonlinear multivariate selection under experimentally altered local sperm densities (densities within the capture zone of eggs) on a range of functionally important sperm traits in the broadcast spawning marine mussel, Mytilus galloprovincialis. Specifically, we assay components of sperm motility and morphology across two fertilization environments that simulate either sperm limitation (when there are too few sperm to fertilize all available eggs), or sperm saturation (when there are many more sperm than required for fertilization, and the risk of polyspermy and embryonic failure is heightened). Our findings reveal that the strength, form, and targets of selection on sperm depend on the prevailing fertilization environment. In particular, our analyses revealed multiple significant axes of nonlinear selection on sperm motility traits under sperm limitation, but only significant negative directional selection on flagellum length under sperm saturation. These findings highlight the importance of local sperm densities in driving the adaptation of sperm phenotypes, particularly those related to sperm motility, in broadcast spawning invertebrates.     

Releasing small ejaculates slowly increases per‐gamete fertilization success in an external fertilizer: Galeolaria caespitosa (Polychaeta: Serpulidae)

The idea that male reproductive strategies evolve primarily in response to sperm competition is almost axiomatic in evolutionary biology. However, externally fertilizing species, especially broadcast spawners, represent a large and taxonomically diverse group that have long challenged predictions from sperm competition theory—broadcast spawning males often release sperm slowly, with weak resource‐dependent allocation to ejaculates despite massive investment in gonads. One possible explanation for these counter‐intuitive patterns is that male broadcast spawners experience strong natural selection from the external environment during sperm dispersal. Using a manipulative experiment, we examine how male reproductive success in the absence of sperm competition varies with ejaculate size and rate of sperm release, in the broadcast spawning marine invertebrate Galeolaria caespitosa (Polychaeta: Serpulidae). We find that the benefits of Fast or Slow sperm release depend strongly on ejaculate size, but also that the per‐gamete fertilization rate decreases precipitously with ejaculate size. Overall, these results suggest that, if males can facultatively adjust ejaculate size, they should slowly release small amounts of sperm. Recent theory for broadcast spawners predicts that sperm competition can also select for Slow release rates. Taken together, our results and theory suggest that selection often favours Slow ejaculate release rates whether males experience sperm competition or not.     

A quantitative genetic competition model for sympatric speciation

I use multilocus genetics to describe assortative mating in a competition model. The intensity of competition between individuals is influenced by a quantitative character whose value is determined additively by alleles from many loci. With assortative mating based on this character, frequency- and density-dependent competition can subdivide a population with an initially unimodal character distribution. The character distribution becomes bimodal, and the subpopulations corresponding to the two modes are reproductively separated because mating is assortative. This happens if the resource distribution is unimodal, i.e. even if selection due to phenotypic carrying capacities is not disruptive. The results suggest that sympatric speciation due to frequency-dependent selection can occur in quite general ecological scenarios if mating is assortative. I also discuss the evolution of assortative mating. Since it induces bimodal phenotype distributions, assortative mating leads to a better match of the resources if their distribution is also bimodal. Moreover, in a population with a bimodal phenotype distribution, the average strength of frequency-dependent competition is lower than in a unimodal population. Therefore, assortative mating permits higher equilibrium densities than random mating even if the resource distribution is unimodal. Thus, even though it may lead to a less efficient resource use, assortative mating is favoured over random mating because it reduces frequency-dependent effects of competition.     

Assortative mating between two distinct micro-allopatric populations of Littorina saxatilis (Olivi) on the northeast coast of England

Size assortative mating is a common invertebrate mating pattern and is usually accompanied by male and female sexual selection, and these three behaviours can contribute to reproductive isolation. Two distinct populations of the marine prosobranch Littorina saxatilis, H and M, occur within 15 m of each other on the same shore. Previous studies have demonstrated that these two forms have different reproductive strategies and that the rare hybrids between the two forms show evidence of reproductive dysfunction and hence are less fit than the assumed parental forms. In both populations, female shell height was shown to be a predictor of the number of embryos contained within the brood pouch. The mean shell height of the M population was significantly larger than that of the H population, and the M population matures at a larger shell size than the H population. The two populations show complete assortative mating to type in the field, and occupy different microhabitats on the same shore. Therefore, laboratory-based experiments were performed to determine if assortative mating was maintained in sympatry and also to determine the effect of population density on mate choice. The males of both populations showed sexual selection for female size, choosing to mate with females approximately 10% larger than themselves from an assortment of female sizes. The M population showed complete assortative mating to type, irrespective of the density of H and M females, whereas at low densities the H males did occasionally mate with M females. The role of assortative mating and reinforcement (due to natural selection acting against the less fit hybrids), in maintaining the partial reproductive barrier between the two populations is discussed.     

Assortative mating by diet in a phenotypically unimodal but ecologically variable population of stickleback

Speciation with gene flow may be driven by a combination of positive assortative mating and disruptive selection, particularly if selection and assortative mating act on the same trait, eliminating recombination between ecotype and mating type. Phenotypically unimodal populations of threespine stickleback (Gasterosteus aculeatus) are commonly subject to disruptive selection due to competition for alternate prey. Here we present evidence that stickleback also exhibit assortative mating by diet. Among-individual diet variation leads to variation in stable isotopes, which reflect prey use. We find a significant correlation between the isotopes of males and eggs within their nests. Because egg isotopes are derived from females, this correlation reflects assortative mating between males and females by diet. In concert with disruptive selection, this assortative mating should facilitate divergence. However, the stickleback population remains phenotypically unimodal, highlighting the fact that assortative mating and disruptive selection do not guarantee evolutionary divergence and speciation.     

Evidence of prezygotic isolation,but not assortative mating,between locally adapted populations of Fundulus heteroclitus across a salinity gradient

Selection along environmental gradients can drive reproductive isolation and speciation. Among fishes, salinity is a major factor limiting species distributions, and despite its importance in generating species diversity, speciation events between marine and freshwater are rare. Here, we tested for mechanisms of reproductive isolation between locally adapted freshwater and brackish water-native populations of killifish, Fundulus heteroclitus, from either side of a hybrid zone along a salinity gradient. There was evidence for pre-zygotic endogenous reproductive isolation with reduced fertilization success between crosses of freshwater-native males and brackish water-native females. Exogenous pre-zygotic isolation was also present where females had highest fertilization in their native salinity. We used a replicated mass spawning design to test for mate choice in both brackish and fresh water. After genotyping 187 parents and 2523 offspring at 2347 SNPs across the genome, 85% of offspring were successfully assign to their parents. However, no reinforcing mate choice was observed. These results therefore demonstrate emerging, yet limited, reproductive isolation and incipient speciation across a marine to freshwater salinity gradient and suggest that both endogenous and exogenous mechanisms, but not assortative mating, contribute to divergence.     

The evolution of gonad expenditure and gonadosomatic index (GSI) in male and female broadcast‐spawning invertebrates

Sedentary broadcast‐spawning marine invertebrates, which release both eggs and sperm into the water for fertilization, are of special interest for sexual selection studies. They provide unique insight into the early stages of the evolutionary succession leading to the often‐intense operation of both pre‐ and post‐mating sexual selection in mobile gonochorists. Since they are sessile or only weakly mobile, adults can interact only to a limited extent with other adults and with their own fertilized offspring. They are consequently subject mainly to selection on gamete production and gamete success, and so high gonad expenditure is expected in both sexes. We review literature on gonadosomatic index (GSI; the proportion of body tissue devoted to gamete production) of gonochoristic broadcast spawners, which we use as a proxy for gonad expenditure. We show that such taxa most often have a high GSI that is approximately equal in both sexes. When GSI is asymmetric, female GSI usually exceeds male GSI, at least in echinoderms (the majority of species recorded). Intriguingly, though, higher male GSI also occurs in some species and appears more common than female‐biased GSI in certain orders of gastropod molluscs. Our limited data also suggest that higher male GSI may be the prevalent pattern in sperm casters (where only males release gametes). We explore how selection might have shaped these patterns using game theoretic models for gonad expenditure that consider possible trade‐offs with (i) somatic maintenance or (ii) growth, while also considering sperm competition, sperm limitation, and polyspermy. Our models of the trade‐off between somatic tissue (which increases survival) and gonad (which increases reproductive success) predict that GSI should be equal for the two sexes when sperm competition is intense, as is probably common in broadcast spawners due to synchronous spawning in aggregations. Higher female GSI occurs under low sperm competition. Sperm limitation appears unlikely to alter these conclusions qualitatively, but can also act as a force to keep male GSI high, and close to that of females. Polyspermy can act to reduce male GSI. Higher male than female GSI is predicted to be less common (as observed in the data), but can occur when ova/ovaries are sufficiently more resource‐intensive to produce than sperm/testes, for which some evidence exists. We also show that sex‐specific trade‐offs between gonads and growth can generate different life‐history strategies for males and females, with males beginning reproduction earlier. This could lead to apparently higher male GSI in empirical studies if immature females are included in calculations of mean GSI. The existence of higher male GSI nonetheless remains somewhat problematic and requires further investigation. When sperm limitation is low, we suggest that the natural logarithm of the male/female GSI ratio may be a suitable index for sperm competition level in broadcast spawners, and that this may also be considered as an index for internally fertilizing taxa.     

Assortative fertilization and selection at larval stage in the mussels Mytilus edulis and M. galloprovincialis

Assortative mating (prezygotic isolation) and reduced hybrid fitness (postzygotic isolation) are typically invoked to explain the stability of hybrid zones. In the tension zone model, these factors work in opposition to migration, which promotes genetic homogeneity. Many marine animals migrate over long distances through a planktonic larval stage. Therefore, strong reproductive isolation is needed to maintain stable marine hybrid zones. However, surprisingly little is known about mating preferences and hybrid fitness in marine organisms. Smooth-shelled mussels (Mytilus spp.) form a well-known species complex, with hybridization over extensive areas such as the contact zone of M. edulis and M. galloprovincialis around European Atlantic coasts. This paper reports direct experimental evidence of assortative fertilization, hybrid larval inviability, and early heterosis for growth rate in M. edulis and M. galloprovincialis. Four crosses between pure M. edulis and M. galloprovincialis were analyzed with a new polymerase-chain-reaction-based diagnostic marker. Gamete competition between taxa was allowed in two out of the four crosses. Genotype frequencies observed at an early stage (36 h after fertilization) unambiguously revealed assortative fertilization when gamete competition was allowed. A significant reduction in hybrid viability was subsequently observed during the larval stage. At the same stage an antagonistic effect, heterosis, was observed on growth rate. However, even if heterosis is observed in the F1, it is expected to vanish in subsequent hybrid generations. Although specialization for different habitats and asynchronous spawning have been mentioned as factors contributing to the maintenance of the blue mussel hybrid zone in Europe, we argue that assortative fertilization and reduced hybrid fitness are important factors that also contribute to the stabilization of this zone. These results emphasize that multiple factors may act concomitantly in a barrier to gene flow, especially in complex life cycles. Furthermore, they show that assortative mating through gamete preference, as already demonstrated for sea urchins, may play a role in speciation processes taking place in the sea.     

Life-history consequences of investment in free-spawned eggs and their accessory coats

The optimal trade-off between offspring size and number can depend on details of the mode of reproduction or development. In marine organisms, broadcast spawning is widespread, and external coats are a common feature of spawned eggs. Egg jelly coats are thought to influence several aspects of fertilization and early development, including the size of the target for sperm, fertilization efficiency, egg suspension time, polyspermy, embryo survival, and fecundity. These costs and benefits of investment in jelly result in trade-offs that can influence optimal reproductive allocation and the evolution of egg size. I develop an optimization model that sequentially incorporates assumptions about the function of egg coats in fertilization. The model predicts large variation in coat size and limited variation in ovum size under a broad range of conditions. Heterogeneity among spawning events further limits the range of ovum sizes predicted to evolve under sperm limitation. In contrast, variation in larval mortality predicts a broad range of optimal ovum sizes that more closely reflects natural variation among broadcast-spawning invertebrates. By decoupling physical and energetic size, egg coats can enhance fertilization, maintain high fecundity, and buffer the evolution of ovum size from variation in spawning conditions.     

Elevated pCO2 increases sperm limitation and risk of polyspermy in the red sea urchin Strongylocentrotus franciscanus

Anthropogenic carbon dioxide (CO₂) emissions and the resultant acidification of surface ocean waters are predicted to have far‐reaching consequences for biological processes in the marine environment. For example, because changes in pH and pCO₂ can alter sperm performance, ocean acidification may be accompanied by reductions in the success of fertilization in marine broadcast spawners. Several studies have attempted to determine the effects of elevated pCO₂ on marine invertebrate fertilization success, albeit with differing results. These conflicts may stem from the use of inappropriate sperm–egg contact times and, in several cases, the lack of measurements over a range of sperm concentrations extending from sperm‐limited conditions to polyspermy scenarios. In our study, we used biologically realistic sperm–egg contact times and a full range of sperm concentrations to assess the effect of elevated pCO₂ on fertilization in the broadcast spawning sea urchin, Strongylocentrotus franciscanus. Fertilization experiments were carried out in seawater bubbled with CO₂ to 400 (control), 800, and 1800 ppm. Using a fertilization kinetics model, we estimate that elevated pCO₂ levels both increased sperm limitation and reduced the efficiency of fast blocks to polyspermy. Thus, elevated pCO₂ decreased the range of sperm concentrations over which high fertilization success was likely. Given the inherent difficulties in achieving high fertilization success in broadcast spawners, raised pCO₂ levels are likely to exacerbate low fertilization success in low‐density populations or in areas with high water turbulence.     

Speciation by natural and sexual selection: models and experiments

A large number of mathematical models have been developed that show how natural and sexual selection can cause prezygotic isolation to evolve. This article attempts to unify this literature by identifying five major elements that determine the outcome of speciation caused by selection: a form of disruptive selection, a form of isolating mechanism (assortment or a mating preference), a way to transmit the force of disruptive selection to the isolating mechanism (direct selection or indirect selection), a genetic basis for increased isolation (a one- or two-allele mechanism), and an initial condition (high or low initial divergence). We show that the geographical context of speciation (allopatry vs. sympatry) can be viewed as a form of assortative mating. These five elements appear to operate largely independently of each other and can be used to make generalizations about when speciation is most likely to happen. This provides a framework for interpreting results from laboratory experiments, which are found to agree generally with theoretical predictions about conditions that are favorable to the evolution of prezygotic isolation.     

An analytically tractable model for competitive speciation

Several recent models have shown that frequency-dependent disruptive selection created by intraspecific competition can lead to the evolution of assortative mating and, thus, to competitive sympatric speciation. However, since most of these results rely on limited numerical analyses, their generality has been debated. Here, we consider one of the standard models (the so-called Roughgarden model) with a simplified genetics where the selected trait is determined by a single diallelic locus. This model is sufficiently complex to maintain key properties of the general multilocus case but simple enough to allow for comprehensive analytical treatment by means of invasion fitness arguments. Depending on (1) the strength and (2) the shape of stabilizing selection, (3) the strength and (4) the shape of pairwise competition, (5) the shape of the mating function, and (6) whether assortative mating leads to sexual selection, we find five different evolutionary regimes. In one of these regimes, complete reproductive isolation can evolve through arbitrarily small steps in the strength of assortative mating. Our approach provides a mechanistic understanding of several phenomena that have been found in previous models. The results demonstrate how even in a simple model, the evolutionary outcome depends in a complex way on ecological and genetic parameters.     

Plasticity of fertilization rates under varying temperature in the broadcast spawning mussel,Mytilus galloprovincialis

Oceans are a huge sink for the increased heat associated with anthropogenic climate change, and it is vital to understand the heat tolerance of marine organisms at all life stages to accurately predict species’ responses. In broadcast spawning marine invertebrates, reproduction is a vulnerable process in which sperm and eggs are released directly into the open water. Gametes are then exposed to fluctuating environmental conditions that may impact their fertilizing capacity. Using the broadcast spawning Mediterranean mussel, Mytilus galloprovincialis, as a model species, we performed blocks of factorial mating crosses to assess the variance in fertilization rates among individuals under both ambient and elevated temperatures. Overall, we found a small, but significant decline in fertilization rates with elevated temperatures. However, there was substantial plasticity in responses, with particular mussels having increased fertilization under elevated temperatures, although the majority showed decreased fertilization rates. Our results suggest possible future reproductive costs to ocean warming in M. galloprovincialis, although it is also possible that genetic variation for thermal sensitivity may allow for adaptation to changing environmental conditions.     

The effect of alternative mating tactics on the fertilization success of a hermaphroditic seabass

In the simultaneously hermaphroditic marine fish, Serranus subligarius, male role individuals are known to pair spawn, group spawn and streak spawn. While the effects of these common mating tactics on mating success in the male role have been well studied, their consequences for the reproductive success of the individuals taking the female role have received little attention. To investigate those consequences, I observed mating behaviors and quantified fertilization success in natural and experimental settings during the summers of 2005-2008 at three sites with different local population densities. I observed focal individuals in 15-min increments and recorded the total number of spawns, number of streak spawns, size of participating spawners, and fertilization rate. The occurrence of small-sized individuals in the local population is associated with higher frequencies of streaking behavior; these small fish are most often first-year individuals reaching sexual maturity late in the spawning season (August/September). Spawns that included one or more streak spawners had a significantly lower average fertilization rate (89?%) than pair spawns without a streak spawner (97?%). This pattern was confirmed with a field manipulation experiment in which spawning events that included streakers again showed lower fertilization rates (93?%) than spawning events that did not include streakers (98?%). These lower fertilization rates occurred despite the fact that spawns that included multiple males produced, on average, 20?% more sperm than produced in spawns with only a single male. These results indicate that females incur a significant fitness cost when streakers invade a spawning event, a cost not attributable to sperm limitation or any direct effects on the female.     

Revealing the mechanisms of sexual isolation in a case of sympatric and parallel ecological divergence

Two ecotypes of a marine intertidal snail (Littorina saxatilis), living at different microhabitats and shore levels, have evolved in sympatry and in parallel across the Galician rocky shore. These ecotypes differ in many traits (including size) due to differential adaptation. They meet, mate assortatively, and partially hybridize at the mid shore where the two microhabitats overlap. The partial sexual isolation observed is claimed to be a side‐effect of the size differences between ecotypes combined with a size assortative mating found in most populations of this species. We investigated this hypothesis using three complementary experimental approaches. First, we investigated which of the different shell variables contributed most to the variation in individual sexual isolation in the field by using two new statistics developed for that purpose: (1) pair sexual isolation and (2) r_i, which is based on the Pearson correlation coefficient. We found that size is the most important trait explaining the sexual isolation and, in particular, the males appear to be the key sex contributing to sexual isolation. Second, we compared the size assortative mating between regions: exposed rocky shore populations from north‐westwern Spain (showing incomplete reproductive isolation due to size assortative mating) and protected Spanish and Swedish populations (showing size assortative mating but not reproductive isolation between ecomorphs). Most of the variation in size assortative mating between localities was significantly explained by the within‐population level of variation on size. Third, we performed a laboratory male choice experiment, which further suggested that the choice is made predominantly on the basis of size. These results confirm the mechanism proposed to explain the sexual isolation in the Galician hybrid zone and thus support this case as a putative example of parallel incipient speciation. © 2008 The Linnean Society of London, Biological Journal of the Linnean Society, 2008, 94 , 513–526.