- Top of page
Genetic variation supplies the raw material for adaptation, evolution and survival of populations and has therefore been a key focus of conservation biology ever since its foundation (Soulé 1985). In previous decades, the neutral component of genetic diversity (generated by mutation and shaped by drift) has been the subject of intense scientific research, fuelled by the increasing availability of molecular markers. On the other hand, the adaptive component of genetic diversity, which is shaped by the action of natural selection, has long remained elusive and difficult to assess, especially at small spatial or temporal scales (Ouborg et al. 2010). Fortunately, new technological and methodological developments now make it possible to identify loci in the genome that are influenced by selection, and thus to get a more complete view of genetic diversity. One article featured in this issue of Molecular Ecology is a good example of this recent breakthrough. Richter-Boix et al. (2011) examined a network of moor frog populations breeding in contrasting habitats in order to understand how landscape features influence patterns of genetic variation. They combined information from both neutral markers and loci putatively under selection to quantify the relative roles of selection and isolation in the evolution of fine-scale local adaptations in these populations. This study nicely illustrates how data on polymorphisms of neutral and adaptive loci can now be judiciously synthesized to help identify the best strategies for preserving adaptive variation, and more generally to enlighten conservation and population-management plans.
Swedish populations of the moor frog (Rana arvalis) can breed in two very different types of habitat: shallow, shaded swamps within woodlands and deep, permanent ponds in open habitats (Fig. 1). These two environments are contrasted for key parameters such as water temperature, hydroperiod and predator abundance, so tadpoles developing in these habitats are probably subjected to very different selective forces (Skelly 2004; Richter-Boix et al. 2010). In order to understand how geographical and environmental features of the landscape shape neutral and adaptive genetic variation in natural populations of this species, Richter-Boix et al. (2011) sampled a network of 17 populations in a 40 km × 40 km area in Sweden and genotyped individuals at 15 microsatellite markers. Their study system was remarkable because habitat features were not spatially autocorrelated across the landscape, with some forest marshes situated nearby open ponds but further away from similar habitats. As a result, substantial gene flow could be expected between different habitats. Using a combination of three methods, the authors identified one locus for which genetic differentiation between habitats exceeded neutral expectations, and for which significant associations between alleles and environmental variables were demonstrated. These two lines of evidence suggest that this locus is locally under directional selection imposed by environmental conditions. This selection scenario is also consistent with the fact that phenotypic differences between habitats have been observed at the larval stage (Richter-Boix et al. unpublished data). Indeed, this particular locus happens to be partially located inside a thyroid hormone receptor gene cloned from a related species and potentially involved in the metamorphosis pathway. In addition, Richter-Boix et al. (2011) analysed gene-flow patterns for neutral markers and for the locus putatively under selection. For this latter, genetic distances were not proportional to geographical distances, and gene flow was particularly high between populations breeding in similar habitats. Conversely, the relationship between habitat differences and gene flow was weaker for neutral markers. These results suggest that selection maintains higher levels of differentiation between habitats at the locus under selection despite the presence of gene flow.
Figure 1. (a) Two moor frogs: this species can breed in a variety of habitats, ranging from (b) open ponds to (c) forest swamps. Photograph credits: (a) Simon Kärvemo; (b–c) Alex Richter-Boix.
Download figure to PowerPoint
Admittedly, this study most likely gives a very rough picture of the overall adaptive diversity found in this network of populations, because it is reduced to variation at a single adaptive locus. This is because of the limited number of genetic markers analysed here, and thus to the non-exhaustive sampling of the moor frog genome. Ideally, identification of loci harbouring selection signatures requires examining at least several hundreds of markers distributed randomly in the genome, in order to (i) reach the (presumably few) genomic regions under selection; (ii) get an accurate estimation of the background genetic variation caused by global neutral forces like drift or mutation; and (iii) identify loci that significantly diverge from this background level of diversity (Storz 2005). However, this article has the merit to present an original approach that will certainly guide the efforts of future conservation genomics studies. With the development of next-generation sequencing technologies, non-model species are currently being ‘genomicised’ at unprecedented rates, and acquiring a wide range of genomic resources such as genomic or transcriptomic sequences and genome-wide marker panels is not as daunting a task as it used to be (Ouborg et al. 2010; Stapley et al. 2010). Meanwhile, new genomic tools are being developed to reveal signatures of selection in genomes, in particular at the sequence level. As a result, our ability to characterize loci underpinning adaption is being greatly enhanced, as is our potential to precisely estimate adaptive genetic variability for conservation purposes (Allendorf et al. 2010).
Adaptive divergence among nearby populations can result from different factors: selective migration between patches with similar environmental conditions (i.e., preferential dispersal of some genotypes towards particular habitats) and very strong selection at a local scale (Richter-Boix et al. 2010); furthermore, high philopatry can promote global divergence by reducing gene flow. Using neutral markers only, identifying the process(es) actually at play is challenging, but comparing gene flow for neutral and adaptive loci can help discriminate between these factors. In moor frog, selection seems to play the primary role for the locus under consideration. Nevertheless, one analysis showed positive correlation between neutral genetic distance and habitat differences as well. This raises the possibility that selective migration can also have a role in the maintenance of local adaptations. In the last decades, several studies have observed adaptive divergence among amphibian populations breeding in different environments. In some cases, the distance between divergent populations was very low, well below the species’ dispersal capability (Skelly 2004). It might be difficult to explain such adaptive variation among nearby populations because traditional models of population genetics, assuming neutrality of markers, predict that migration would overwhelm selection and homogenize variation. However, habitat choice, strong selection and reduced recombination can maintain divergence in the parts of the genome underlying the adaptive traits, while gene flow homogenizes other genomic regions, and therefore allows differentiation even in presence of substantial gene flow (Via 2009).
Adaptive genetic variation is pivotal for long-term persistence of populations and their adaptation to a changing environment, yet conservation of adaptive variation may be a complex task. Likewise, management of gene flow among populations is a key issue that can be tackled using very different approaches (Fig. 2). On the one hand, connecting populations and facilitating gene flow can allow the maintenance of more genetic variability, reducing the risk that populations go extinct because of demographic and genetic processes (extinction vortex), and may increase the amount of overall adaptive variation. On the other hand, increasing gene flow may disrupt local adaptation and therefore cause loss of fitness (Ficetola & De Bernardi 2005; Allendorf et al. 2010). Deciding the correct approach is challenging and requires extensive information on ongoing processes and their consequences on population dynamics. The approach of Richter-Boix et al. (2011) may help to identify the more appropriate strategy (Fig. 2). If local adaptations are mostly caused by strong directional selection, such as in the case of moor frog populations, allowing gene flow would not disrupt adaptation. Conversely, if restricted gene flow facilitates local adaptation (such as in some partially allopatric populations), management actions should avoid population admixture. Comparing gene flow measured using neutral markers and loci under selection may allow the identification of the process(es) determining local adaptations, and therefore the appropriate management strategy. Furthermore, the integration of geographical, environmental and/or ecological data into analyses is invaluable for refining the picture about the selective forces that are at play in each case.
Figure 2. Strategies for the conservation of adaptive genetic diversity among populations, depending on processes influencing variation.
Download figure to PowerPoint