An elevation of saltmarsh vegetation zonation as a predictor of elevation relative to mean sea level

Patterns of biological zonation across salt marshes are commonly used as proxies for mean seal level to reconstruct past environments, and predict future change with sea level rise. Reconstructions employ the transfer function to provide robust R2 values to determine accuracy of the calibration of modern assemblages. However, rarely are such parameters provided when salt marsh vegetation patterns are used to predict future change in a GIS. This study models the predictive capability of salt marsh vegetation at a local and regional level by asking: "Is there a general rule between salt marshes with different tidal ranges that causes zonation." Elevation, tidal inundation period, soil salinity, pH, % sand, moisture and organic content were used to predict the presence of vegetation types using multiple logistic regression. Results show that at a site level, % sand, probably acting as a surrogate for moisture content, was the best predictor of vegetation type before elevation and inundation period. However, at a regional level, between marshes of different tidal range, only elevation and inundation period were capable of predicting vegetation type. Most significant for predicting change in a GIS is the finding that at both meso and micro tidal sites, the seaward edges establish at around 30% inundation period and the marsh upland border extends to less than 1% inundation period.