Soil chemical features as key predictors of plant community occurrence in a Mediterranean coastal ecosystem

Modelling the relationships between soil chemical factors and vegetation in coastal areas provides a significant contribution to understanding the ecology and conservation of these ecosystems in the Mediterranean basin. The present study relates coastal plant communities and soil chemistry using multivariate analysis and generalized additive models (GAMs) in order (a) to determine the relationships between soil factors and plant community occurrence along a coastline-lagoon gradient; (b) to build best models of soil chemical predictors for coastal plant community occurrence. Distance from coastline was the primary driver of changes in plant community composition and soil factors on the Mediterranean coast. Field capacity, pH, CaCO3, NaCl and soil organic matter were the major parameters related to the pattern of vegetation distribution. Along the costaline-lagoon gradient, however, plant community and soil changes agreed for two main sequences of habitats. The first was in the sand dune environment, involving plant communities, the species composition of which varied in response to increasing soil development. The second was near lagoons, where vegetation responded to soil salinity, acting independently of other soil factors. Coastal plant community occurrence was explained by GAMs, using a few soil chemical predictors: pH and CaCO3 were the main predictors for dune communities but only partly explained their occurrence, whereas NaCl was the key factor for salt marsh community occurrence and largely contributed to the model. The results suggest that soil chemical changes can be detected through vegetation distribution, particularly in saline habitats. Thus, habitat-based models could be a useful tool to identify the key factors for different types of vegetation occurrence in coastal habitats.