Genetic structure across the GCE domain

Overview Wares (UGA) and collaborators have conducted a survey of the genetic structure of organisms in the GCE domain. The results illustrate the importance of examining processes of site diversity and diversification at multiple scales and levels of biological organization. As might be predicted based on the larval life history of most of the invertebrate species, there is little to no significant differentiation of marine invertebrates across GCE sites, and results from the marshgrass Spartina alterniflora also exhibit no significant pattern. Figure 1 shows the distribution of pairwise estimates of Wright's Fst from mitochondrial data in the invertebrate species; only a few pairs of populations have statistically significant results, but these results are neither consistent across taxa nor reliable when you consider the effect of multiple comparisons (i.e. after Bonferroni correction). Contrasts of Fst > 0.10 are summarized below with the initials of each species and the pair of populations exhibiting that value (Li = Littorina irrorata; Up = Uca pugnax; Mb = Melampus bidentatus). Although there was little genetic differentiation among sites, detailed examination of the data found two interesting patterns. First , examination of nucleotide diversity - a measure of the amount of polymorphism harbored in a DNA sequence data set - indicated that populations closer to the ocean side of the domain were more than twice as diverse on average as populations in the interior or mainland regions of the domain. This preliminary result suggested that even if sites in the GCE domain were not evolutionarily differentiated from one another - and thus adaptation was not driving ecological differences - that some element of larval recruitment or other effect of the environment on diversity at a site was important. Second, we assessed the pattern of nucleotide diversity across all sites and eight species, including the invasive porcelain crab Petrolisthes armatus, and showed that these spatial patterns were strongly correlated with site species diversity (using previous GCE data for molluscan and plant diversity at each site) (Figure 2). Although the pattern was only marginally significant for Geukensia demissa and a couple of other species, the combined analysis suggested a highly significant relationship between genetic and taxonomic diversity at GCE sites (as well as between appropriate measures of genetic and species richness when only the molluscan species data were included). To consider the possibility that other environmental effects were driving the observed diversity pattern, we developed a linear model to test the fit of GCE salinity and temperature data to the nucleotide diversity data, as well as including data from a preliminary analysis of barnacle (Chthamalus fragilis) recruitment and abundance performed by T. D. Bishop and colleagues. The resultant model indicated that salinity and temperature differences among GCE sites were not strongly associated with diversity patterns, but that recruitment (e.g. larval availability) generated a much stronger fit to the genetic diversity data. Overall, we feel that this suggests that the freshwater-saltwater gradient itself may not influence genetic diversity in the GCE ecosystem, but that larval transport does influence the success and abundance of recruits in the domain. The pattern is significant but not without its wrinkles; one site that harbored very high genetic diversity in several species was GCE 10 on the Duplin River. Although geographically more proximal to the ocean than mainland, this site is upriver and unlikely to maintain diversity in the same way as other estuarine/marsh sites. Considering anecdotal evidence from other researchers at GCE, it may be that the populations of invertebrates in the Duplin River exhibit high self-recruitment (retention) of larvae, leading to the maintenance of higher diversity despite being upriver. We complemented this work on intertidal invertebrate species by evaluating genetic diversity in two grasshopper species, Hesperotettix floridensis and Orchelimum fidicinium. When the genetic diversity of these two species was compared with orthopteran species richness data collected over five years in the GCE-LTER, disparate patterns are recovered (Figure 3). For H. floridensis, the expectation of a positive relationship between genetic and species diversity is upheld in all five comparisons; however, in O. fidicinium, negative correlations between species and genetic diversity are seen. While there are fewer data available in this study to date, our results already suggest strong demographic distinctions between the two species. Both species appear to have experienced a population bottleneck; however, preliminary data suggest that the timing of these inferred demographic events differ greatly, with O. fidicinium having undergone a much more recent reduction in population size. Sampling of marine invertebrates was expanded to other LTER sites on the Atlantic coast to gain contextual data for gene flow and recruitment studies (see Study Area). The results showed that gene flow is high at regional scales as well. We performed a large-scale comparative phylogeographic analysis of 6 intertidal marsh-associated invertebrate species from LTER sites in Massachusetts, Virginia, and the GCE, as well as comparable habitat from central Florida. This analysis explored differentiation across the Upper Virginian (UV), Lower Virginian (LV), Georgian (GCE), and Floridian (FL) provinces as defined by current biogeographic analyses, and indicated no significant differentiation when populations from Chesapeake Bay to Florida were compared (figure; histograms indicate Hudson's nearest neighbor statistic Snn among regional pairwise comparisons). This result shows that ecological and interaction differences across latitudinal gradients in marsh habitats are more likely driven by environmental gradients and/or increased predation stress, rather than the populations being assembled from distinct gene pools. This work did identify 2 (out of 6) species (Geukensia demissa and Uca pugilator) that exhibited significant differentiation across Long Island Sound, consistent with several previous phylogeographic syntheses.