Predicting interactions between wetland vegetation and the soil-water and surface-water environment using diversity, abundance and attribute values

Abstract

This study investigated the response of freshwater wetland vegetation to hydrological driving factors by assessing collective vegetation variables, traits of dominant plant populations and hydrological and hydrochemical variables, repeat-sampled within wetland sites across Scotland and northern England. Sampling was conducted at 55 permanent sample stations located along 11 independent transects. Eco-hydrological interactions were investigated using a regression-based modelling approach. Facets of the water-table dynamic (e.g., level of drawdown, level of fluctuation), along with vegetation abundance (e.g., biomass, stem density) and diversity (e.g., species richness) values, were used to build predictive models. Of the models predicting vegetation characteristics, the greatest predictive power was R 2=0.67 (p < 0.001) for a model predicting stem density (m−2). Conversely, vegetation variables proved useful for predicting characteristics of the water-table environment. In this instance, the greatest predictive power was R 2=0.79 (p < 0.001) for a model predicting minimum water table level (i.e. maximum level of drawdown). The models were tested using data collected during 2000 from repeat sites and independent sites. This approach might be successfully applied for the purposes of integrated eco-hydrological management and monitoring of freshwater wetland vegetation.