The dominant paradigm for explaining vegetation pattern in salt marshes is that there is a tradeoff between competitive ability and stress tolerance, leading to vegetation zones along a single gradient of increasing flooding and salinity from high to low marsh. This single-gradient paradigm breaks down, however, when the salinity gradient is decoupled from the flooding gradient, as happens following disturbance or at low latitudes. We propose instead a “centrifugal” model of community organization with two separate stress axes of increasing salinity and increasing flooding, and evaluated this model using four types of data from coastal Georgia: 1) edaphic variation, 2) transplantion and 3) watering experiments, and 4) variation in zonation patterns among marshes.
All four types of data were consistent with the model. First, bivariate plots of soil water content and salinity revealed two distinct gradients, with plants and habitats distributed along these orthogonal gradients. Second, plants transplanted from the “core” habitat (low waterlogging; low salinity) to either peripheral habitat performed poorly, but plants transplanted from either peripheral habitat to the core performed well if competition was reduced. Third, reducing salinity by watering led to the invasion of “core” species into what previously was a “peripheral” habitat. Fourth, comparing across 55 sites, the range of different plant zones was correlated with soil water content, soil salinity content, and abundance of the top competitor. In turn, soil water content and salinity were correlated with landscape position and shape, suggesting important landscape-level drivers of zonation patterns. These results suggest that a two-gradient centrifugal model is required to understand vegetation pattern in low-latitude salt marshes, and that more attention should be paid to landscape-scale drivers of marsh vegetation patterns.