The relationship between woody vegetation and environmental factors: a spatial discriminant analysis

A detailed analysis of the relationship between woody types and environmental variables (pedological and topographical) was carried out inside the city of Rome. Twenty-three sample sites 100 m(2)each were selected according to the principle that they were inside woody vegetation patches greater than 2.3 ha. Presence-absence data were analysed through hierarchical classification and principal coordinates analysis in order to detect woody vegetation types. The six groups identified were then analysed according to thirty-four variables: a spatial discriminant analysis was performed using soil physical and chemical variables measured in the A1 and A2 horizons, topographical variables (altitude, slope and aspect), and annual potential irradiation. This procedure was able to quantify the contribution of the spatial distribution of the samples with respect to that of the environmental variables, thus improving the discriminant model. The combination of three variables: aspect, organic matter A2 and exchangeable cations A2 is the most effective in discriminating the woody types allowing a hypothesis for the planning and management of these communities.