Introduction
A hallmark of angiosperm diversification has been the ability of plant species to adapt to an extremely wide range of environments characterized by different biotic and abiotic stresses. Among abiotic stresses, the physiological stress of growing in high salt environments is particularly challenging for plants. High levels of salt change the osmotic environment, leading to dehydration; in addition, both Na+ and Cl- ions are toxic if they accumulate in cells. Nonetheless, a small number of terrestrial plant species (ca. 2%) have managed to colonize and even thrive in saline ecosystems such as beaches, salt marshes, and estuaries (Glenn, Brown, & Blumwald, 1999). These plants, collectively called halophytes, have become important models for the study of salt tolerance mechanisms and evolution (Cheeseman, 2013).
The term “salt tolerance” is often used as though it is a binary characteristic; a plant is either tolerant or not. However, salt tolerance in halophytes is likely to be quantitative, with tolerance tuned to match the local environment (Cheeseman, 2013). Saline environments are often characterized by broad gradients of environmental salt concentration, from relatively low concentrations up to (and occasionally even beyond) the concentration of seawater. Such gradients are common where fresh and sea water mix, with areas of lower salinity extending along the coastline and far offshore from the mouths of major rivers (Fournier, Reager, Dzwonkowski, & Vazquez‐Cuervo, 2019). This variation offers extensive opportunities for local salinity adaptation within widely distributed halophytic species; however, whether such local adaptation occurs and the forces that influence it are largely unknown. As described below, three common characteristics of halophytic species that could shape patterns of local salinity adaptation are clonal propagation, interspecific hybridization and ploidy variation.
Clonal propagation occurs widely in halophytes and has important implications for geographic patterns of diversity and local salinity adaptation (Bricker, Calladine, Virnstein, & Waycott, 2018; Jefferies & Rudmik, 1991; Róis et al., 2015). Depending on the dispersal capacity of clonal propagules and the frequency at which novel genotypes arise (e.g., through sexual reproduction or somatic mutation), one of two reproductive strategies might be expected. If new genotypes arise often but have limited clonal dispersal, a “frozen niche” strategy might be favored, whereby multiple genetically distinct clonal populations occur, each restricted to the small geographical area to which it is locally adapted (Róis et al., 2015; Vrijenhoek, 1979). Alternatively, if the production of new genotypes is infrequent and/or dispersal potential is high, a “general-purpose genotype” strategy might instead be favored, whereby a few generalist genotypes are present in multiple environments across a wide geographic range (Baker, 1965; Bricker, Calladine, Virnstein, & Waycott, 2018; Coughlan, Han, Stefanović, & Dickinson, 2017). The underlying conditions that lead genotypes to follow one pattern or another remain poorly understood (Vrijenhoek & Parker, 2009).
In contrast to clonal propagation, interspecific hybridization can provide access to new ecological niches by introducing novel genetic variation into a population and/or increasing heterosis (Rieseberg, Archer, & Wayne, 1999; Soltis & Soltis, 2009). Adaptation to saline environments has been shown to be influenced by natural hybridization in several wild systems. Specifically, hybrids derived from two halophytic species may show a relative increase or decrease in tolerance relative to their parents (Gallego-Tévar, Curado, Grewell, Figueroa, & Castillo, 2018; Lee, Ayres, Pakenham-Walsh, & Strong, 2016). It can also produce a novel halophyte from two non-halophytic parental species (Edelist et al., 2009; Welch & Rieseberg, 2002). The relative roles of the genomic contributions from different parental species for phenotypic and niche differentiation in hybrids remains an open question (Bar-Zvi, Lupo, Levy, & Barkai, 2017). High-throughput genomic techniques can help in identifying hybrid genotypes, their genomic composition, and the relative importance of their parental genomes for salinity adaptation.
Polyploidy (i.e. whole-genome duplication) may also play a role in salinity adaptation. Studies in non-halophytes have suggested that inducing autopolyploidy in a diploid species (i.e. doubling its genome) can increase salt tolerance (Chao et al., 2013; Wu, Lin, Jiao, & Li, 2019). However, in the wild most halophytic polyploids are not autopolyploids but rather allopolyploids — having arisen through interspecific hybridization where each parental species contributes its entire genome (Ainouche, Baumel, Salmon, & Yannic, 2004). In such cases the effect of the genome duplication is confounded with the potential effects of interspecific hybridization described above (Fort et al. 2016). Systems where repeated natural hybridization events have resulted in a mix of diploid and polyploid genotypes could potentially be used to untangle these effects. If such a group also varied with respect to the relative numbers of genome copies received from each parental species, whole-genome dosage effects on fitness and other phenotypes could also be examined (Betto-Colliard, Hofmann, Sermier, Perrin, & Stöck, 2018; Harvey, Fjelldal, Solberg, Hansen, & Glover, 2017; Tan et al. 2016).
The halophytic grass seashore paspalum (Paspalum vaginatumSwartz) provides an advantageous system for addressing how clonality, hybridization and polyploidy interact to shape patterns of genotypic variation across salinity gradients. First, it has a worldwide distribution across habitats of varying salinity (Duncan & Carrow, 2000). Second, it shows quantitative variation in salt tolerance, up to and including sea water concentrations (Lee, Duncan, & Carrow, 2004; Lee, Carrow, & Duncan, 2004, 2005; Lee, Carrow, Duncan, Eiteman, & Rieger, 2008), although the genetic basis and geographical distribution of this variation is almost completely unknown. Third, it spreads vegetatively by stolons and is capable of clonal propagation under cultivation, suggesting that dispersal of clonal propagules may affect geographical patterns of genotypic variation in the wild. Fourth, it has been proposed to undergo hybridization with a closely related non-salt tolerant species (P. distichum L.) (Eudy, Bahri, Harrison, Raymer, & Devos, 2017), which could contribute to natural phenotypic variation in salt tolerance. Finally, although it is believed to be predominantly diploid (2n = 2x =20) (Duncan and Carrow, 2000; Eudy, Bahri, Harrison, Raymer, & Devos, 2017), both diploid and triploid cytotypes have been reported (Eudy, Bahri, Harrison, Raymer, & Devos, 2017), which suggests that ploidy level and subgenome composition might also contribute to population structure and local salinity adaptation.
In addition to these life history features, Paspalum vaginatumalso provides the benefits of a genome-enabled model system with experimental tractability and economic value that facilitate applied studies. It has a manageably-sized ~600 Mb diploid genome which has recently been assembled, allowing for genomic analyses (Eudy, Bahri, Harrison, Raymer, & Devos, 2017; Qi et al., 2019). It is closely related to several economically important grass species in the subfamily Panicoideae, including maize, sorghum and sugarcane (Morrone et al., 2012), which could facilitate the transfer of salt tolerance mechanisms to major cereal crops. It is also economically important in its own right as a turfgrass for salt-affected soils. As such, it has resources and germplasm available through breeding programs (Duncan & Carrow, 2000).
One current limitation of publicly available (USDA) P. vaginatumgermplasm is its narrow genetic variability due to genotypic redundancy among accessions (Eudy, Bahri, Harrison, Raymer & Devos, 2017) and collection biases towards plants with favorable turfgrass qualities (Duncan & Carrow, 2000). Additionally, due to the decades-long gap between the collection and genotyping of these USDA samples, it is unclear whether the high degree of genotypic redundancy in the germplasm collection represents natural clonal propagation in the wild or post-collection cross-contamination of accessions, which can easily occur as the aggressive stolons typical of the species invade neighboring pots (Eudy, Bahri, Harrison, Raymer, & Devos, 2017). Genotyping of newly collected material is thus required to quantify the actual levels of clonal propagation in the wild and geographical patterns of genotypic diversity.
In this study, we undertook the largest collection of living wild specimens of P. vaginatum and P. distichum to date and analyzed them together with available USDA germplasm to investigate the nature of genotypic distributions across salinity gradients in the North American species range. Specifically, we asked the following questions: 1) Do wild P. vaginatum populations represent many distinct genotypes or replicates of a few geographically widespread clones? 2) To the extent that clonal propagation occurs in the wild, do geographical and environmental distributions suggest a localized “frozen niche” reproductive strategy, a widespread “general purpose genotype” strategy, or no discernible pattern? 3) To what extent does interspecific hybridization and/or ploidy variation in this system influence patterns of genetic differentiation and diversity in wild populations? 4) How do clonal structure, ploidy variation and subgenome composition vary across geographical gradients in salinity? Our findings provide insight into the roles of genome composition and clonal propagation in local salinity adaptation in P. vaginatum , and lay the necessary groundwork for future investigations into the genetic basis of salt tolerance in this genomic model halophytic grass species.