Yes, Gallagher et al. (1988) performed a common garden experiment that lasted nine years which suggested some degree of morphotype plasticity. However even after 9 years plant biomass, height, density, diameter, flowering frequency, underground regrowth reserves and flowering frequency remained distinct, which suggests that growth form differentiation may be due to mostly genetic differences.

In general, plants show remarkable phenotypic plasticity, from inflorescence color and size to the photosynthetic pathway employed to meristematic growth. This plasticity is often influenced by environmental variables (in this case salinity, nitrogen, tidal height). The plasticity in S. alterniflora is interesting in that there are only two observed morphotypes occupying distinct yet adjacent habitats. To account for intra-specific phenotypic plasticity among tall- and short-form morphotypes we sampled 20 individuals at each of the seven sites – 280 samples in total. Individuals sampled had to fall within certain environmental and physiological parameters to be included in the study. Since plasticity is so often observed we hypothesized that the morphotype variation observed is a phenotypic response rather than a genetic difference. We believe that the breadth and rigidity of our sampling parameters, along with our sequencing approach should adequately resolve questions surrounding morphological variation in S. alterniflora.

Yes, we are definitely thinking about the differences between various marshes as both an extra factor of which to be wary, and a source of interesting information. We recorded data such as salinity, distance from open water, and tide height for all of the collection sites, and will begin integrating this information with the results of our genetic analyses once the sequencing data is in our hands. However, as I am training in biogeography, I agree that these differences are intriguing—and in fact, part of what motivated our question in the first place. There are clear differences in marsh structure between New England marshes and Southern marshes, including the distribution of tall vs. short forms of Spartina. Thus, our analysis is more of a 2D. We are looking for genetic divergence along the gradient, divergence between tall and short forms, and most interestingly, we are curious if there is a geographic difference in the degree to which genetics vs. phenotypic plasticity determine the height forms.

Of course. There are a lot of potential ways that species and populations could respond to climate change, and the problem in conservation biology right now is that we don’t know enough about most species to be able to predict which response they will have. A good picture of the population genetics of a species yields information about population structure, genetic diversity, migration patterns, and maybe even evolutionary history, thus giving us many more tools to answer these questions. I’ll give you a couple of examples. Many species may shift their geographic ranges to track changing conditions, e.g. towards the poles or higher in altitude (if temperature is their main concern). Population genetics can help us infer past migration patterns as well as barriers to gene flow (which might look like a divergence between populations sharper than expected based on their geographic distance); those characteristics of a species or metapopulation are very important for determining if the species will be able to shift its range. Another possible climate change response is adaptation to the new conditions, which requires enough standing genetic variation on which natural selection can act. It will still be hard to predict the exact nature of adaptation, but that variation needs to exist in order for adaptation to occur; population genetic analyses tell us the amount of standing variation in a population or species. Those are just a couple of many examples, but I hope you get the idea – the genome is full of information, if only we know how to look!
As to how exactly our analyses will answer these questions: with our sequence data for 280 individuals in 7 populations of Spartina alterniflora, we will search for SNPs, which are single nucleotide sites which carry variation between individuals in the species (you can see my response to a below question for more information on SNPs). The frequency of SNPs in the genome suggests the amount of neutral genetic variation in the species. Differences in SNPs between populations (or ecotypes, like our tall and short forms of Spartina) indicate a divergence between those populations, either due to drift or selection. Distributions of SNPs and the frequency of their alleles is what can inform us about population structure and, with more detailed analysis, a variety of other traits such as those I outlined above. Hope this is helpful!

Thanks for your comments! Sorry for the jargon. A SNP, or single nucleotide polymorphism, is a single base pair site in the genome that contains variation within the population or metapopulation. Some percent of the population will have a particular nucleotide at that site, and the rest of the individuals will have a different nucleotide (sometimes more than two). If you assume that individual populations are in equilibrium, then differences in the frequency of those two nucleotides, or alleles, between different populations indicates that there is genetic divergence between those populations, either due to selection or drift. With a single site, or SNP, you cannot make very robust conclusions, but with a large amount of sequence data as we will have, the large number of SNPs we expect to find will be statistically powerful for deducing the structure of this metapopulation. However, sequencing the entire genome for our sample size of 280 individuals would require an enormous, and unnecessary, amount of sequencing and data “space.” Thus, we chose the ddRAD-tag method, which allows us to randomly select a small fraction of the genome (~1%) and still ensure that the same 1% is selected for every individual. That allows us to compare the same SNP sites for every individual across all populations. With a genome size on the order of 4×10^9, and a SNP frequency on the order of 4×10^2, we expect approximately 10^5 SNPs in just that 1% of the genome!