Heredity (2008) 101, 83–91 & 2008 Nature Publishing Group All rights reserved 0018-067X/08 $30.00 www.nature.com/hdy ORIGINAL ARTICLE Adaptive responses of independent traits to the same environmental gradient in the intertidal snail Bembicium vittatum MS Johnson and R Black School of Animal Biology (M092), University of Western Australia, Crawley, WA, Australia The snail Bembicium vittatum occupies a wide range of flatter shells than those on gently sloping shores. These intertidal habitats in the Houtman Abrolhos Islands, Western associations are repeatable between three separate groups Australia. Allozyme variation reflects patterns of connectivity, of islands. Spotting shows a weaker, but significant associa- which are independent of local habitat. In contrast, heritable tion with the same gradient. Although shape does not differ differences in shell shape among 83 shore sites vary with between colour morphs within populations, the proportion of habitat, indicating local adaptation. Here we examine dark shells is strongly associated with shape. Thus, the dimorphisms of colour and spotting of the shell in the same independent shell traits are apparently adapted to a populations, as a test of consistency and complexity of common, biologically significant gradient, even though the patterns of local adaptation. Within populations, the fre- adaptive mechanisms probably differ for colour and shape. quency of spotted shells is higher in dark shells. Despite this The parallel variations of independent traits highlight both the association, spatial variations of colour and spotting are only complexity of local adaptation and the potential to reveal weakly correlated. As predicted for traits associated with evolutionarily significant environmental contrasts by examin- local adaptation, subdivision is greater for colour, spotting ing adaptively relevant traits. and shape than for allozymes. Colour and shape are Heredity (2008) 101, 83–91; doi:10.1038/hdy.2008.33; associated with local habitat, such that populations on published online 7 May 2008 vertical shores have higher frequencies of dark and relatively Keywords: adaptation; shell morph; genetic subdivision; gastropod; FST; QST Introduction phenotypic traits are more likely to reflect adaptive variation, and such traits often show different patterns of Understanding genetic structure of populations is divergence to those of molecular markers (Lynch, 1996; fundamental to the study of evolutionary change and McKay and Latta, 2002; van Tienderen et al., 2002). to basic issues in conservation genetics. Genetic structure Although not directly relevant to adaptive divergence, is usually examined using molecular markers, such as neutral molecular markers provide a valuable frame- allozymes, mtDNA sequences, microsatellite DNA and work for detecting adaptively significant variation. For amplified fragment length polymorphisms (AFLPs; example, associations of traits with habitat are potential Avise, 2004). Such markers typically behave as selec- evidence of local adaptation, but confounding effects of tively neutral or at least quasi-neutral, and hence are history or gene flow must be taken into account. Neutral useful for inferring patterns of gene flow and historical markers can reveal whether associations of phenotypic connections among populations, as well as for quantify- traits with habitat are independent of patterns of gene ing genetic variation within populations. Such molecular flow or historical connections, and therefore whether the markers provide a structural and historical context in associations provide strong evidence for local adaptation which adaptive evolution might occur, but their relation- (Schneider et al., 1999; Quesada et al., 2007). A related use ship to adaptive change is unclear, because these markers of neutral markers for recognition of significant traits is are generally poor indicators of functionally significant based on the prediction that adaptively relevant traits traits (for example, Lynch, 1996; Pearman, 2001; Reed should show greater subdivision than neutral markers. and Frankham, 2001). A high degree of genetic subdivi- This notion was clearly expressed by Lewontin and sion indicates the opportunity for local adaptation, but Krakauer (1973) as a test for natural selection, in the isolation also increases the potential for random drift to context of identifying disparities among allozyme loci. dominate. In contrast to neutral molecular markers, As expected from this perspective, levels of genetic subdivision for phenotypic traits are often higher than those for molecular markers (Spitze, 1993; Gockel et al., Correspondence: Professor MS Johnson, School of Animal Biology (M092), 2001; Storz, 2002; Steinger et al., 2002; Luttikhuizen et al., University of Western Australia, 35 Stirling Highway, Crawley, WA 2003; Volis et al., 2005; Schiffer et al., 2007). This approach 6009, Australia. E-mail: [email protected] has also been the basis for detecting subsets of Received 15 November 2007; revised 18 March 2008; accepted 24 potentially important loci, when using AFLP markers March 2008; published online 7 May 2008 as genomic scans (Beaumont, 2005; Storz, 2005), with Complex local adaptation in an intertidal snail MS Johnson and R Black 84 important implications for using genetics to set priorities size and shape (Johnson and Black, 2000), and all but two in conservation (Bonin et al., 2007). of them were used in the allozyme study of genetic Intertidal snails provide useful opportunities for divergence (Johnson and Black, 1996). The average testing contrasting patterns of genetic divergence for sample size was 52, and only four sites had fewer than molecular markers and potentially adaptive morpho- 33 snails (N ¼ 17–26). The samples represent the species’ logical variation, because they are readily accessible and distribution throughout the Abrolhos Islands, spanning often occupy highly contrasting habitats (reviews in nearly 80 km, and include the major habitats in all three McQuaid, 1996; Reid, 1996). An example is the littorine of the separate island groups. Bembicium vittatum in the Houtman Abrolhos Islands, an Although contrasts between high and low on the shore archipelago of more than 100 small islands, 70 km off the are important for local adaptation of some intertidal Western Australian coast. This species lacks planktonic species (Grahame et al., 2006; Quesada et al., 2007), larvae, and consequently is genetically subdivided over B. vittatum in the Abrolhos Islands occupies a tidal range distances of tens to hundreds of metres (Johnson and of o0.5 m, and environmental heterogeneity is modest Black, 1995, 1996, 1998a). The species is abundant in the within sites. In contrast, there are substantial differences Abrolhos Islands, where it occupies a wide range of of habitat among sites occupied by B. vittatum. Quanti- habitats. The advantages of B. vittatum in the Abrolhos tative descriptions of the habitats at 71 of the sites were Islands for studying adaptive genetic divergence are made previously, based on 31 variables that include threefold: (1) analysis of many, well-characterized sites indicators of exposure, water flow, substrate, shore provides statistical power and well-defined environmen- profile and biological communities (Johnson and Black, tal context (Johnson and Black, 1996, 1997); (2) analysis of 1996, 1997). These variables provide a clear description of molecular variation has shown clear patterns of genetic the major variations in habitat, and are good predictors connectivity, but these are independent of any measured of the distributions of high intertidal gastropods in the features of the local intertidal habitat, indicating that Abrolhos Islands (Johnson and Black, 1997). To provide associations of other traits with habitat are not due to composite indices, the habitat variables were subjected to confounding effects of gene flow (Johnson and Black, principal components analysis, calculated using the 1995, 1996, 2006); (3) replication of habitat characteristics Statview SE þ Graphics program (Abacus Concepts among three (genetically divergent) island groups allows Inc.). The raw principal components were used, provid- rigorous testing of associations with habitat (Johnson and ing orthogonal (statistically independent) axes of varia- Black, 1996, 2000). Habitats occupied by B. vittatum differ tion in habitat. The first three principal components are in substrate, slope of the shore, exposure to currents and readily interpretable descriptors: PC1 represents expo- wind, and associated biological communities. These sure to wave action and water flow; PC2 the gradient contrasting habitats provide a specific context in which between vertical and sloping shores; PC3 aspect in local, adaptive divergence can be studied. relation to prevailing winds (Johnson and Black, 1996). In common with many intertidal snails, B. vittatum is variable for shell shape and pigmentation. Variations in shape and spotting of the shell are inherited (Parsons, Shell polymorphism 1997), and hence are candidates for local adaptation. The Shells were scored for the presence of dark spots on the functional importance of such traits is understood in ventral surface and for colour. Spotting is inherited, with broad terms for intertidal snails generally and for the spotted condition being dominant to unspotted littorines in particular (Vermeij, 1993; McQuaid, 1996; (Parsons, 1997). The number of spots varies, but any Reid, 1996). In B. vittatum, there are close associations