Rising sea levels, and other (direct) anthropogenic pressure, coupled to recurrent flooding during the Atlantic hurricane seasons, puts coastal aquatic ecosystems on and around the Atlantic coast of North America at greater risks. In August of 2011, the Hudson River, NY, was hit by two major tropical storms - Lee and Irene – at the peak growing season of the foundation submersed aquatic plant species Vallinseria americana. Resulting losses of >90% of V. americana cover could translate into declines in genetic diversity and increased genetic differentiation in recovering populations thereby reducing the resilience of these populations to future environmental perturbations. To understand the genetic effects of the storms and patterns of recolonization, we compared 108 shoots, fortuitously sampled from 5 sites in the lower Hudson River just before the storms hit in 2011, with 239 shoots sampled from 8 sites collected during surveys of the same area four years later. These samples gave us the unique opportunity to compare pre-storm and post-storm populations. Based on 10 microsatellite loci, we calculated the number of unique multilocus genotypes. Using one instance of each multilocus genotype at each site, we measured of genetic diversity (allelic richness and heterozygosity) within sites, partitioning among sites (Fst, Principal Coordinates Analysis (PCoA), and Structure), and effective population sizes.
Results/Conclusions
Populations sampled before and after the storms show very similar patterns of genotypic diversity and genetic variability. We found no significant difference in pre-storm vs post-storm populations in terms of genotypic diversity (GD = 0.536 vs 0.496), allelic richness (Ar = 2.023 vs 2.005), mean, gene diversity (Hs = 0.493 vs 0.475), observed heterozygosities (Ho = 0.557 vs 0.560). Genetic differentiation was not elevated in these post-storm population as hypothesized (pre-storm Fst =; 0.137, Dest = 0.144; post-storm Fst = 0.120, Dest = 0.121). Bayesian clustering analyses in Structure and PCoA indicated that individuals sampled across both time intervals formed groups that reflected spatial proximity rather than year. These results suggest patterns of genetic connectivity were not impacted by the storms and that pockets of genetically diverse microrefugia must have persisted and served as the sources of recolonization after the storm. Very low effective population sizes in both 2011 and 2015 indicate populations were chronically or episodically small long before these storms.