A MODEL OF SPECIES DENSITY IN SHORELINE VEGETATION

Predicting the species density (number of species per unit area) of communities is a major goal of ecology. We present a regression model of species density on a local (0.25 m2) scale for the vegetation of freshwater shorelines in southwestern Quebec, Canada. Two attributes of the vegetation, the amount of aboveground biomass (in grams) and the proportion of the vegetation composed of obligate perennial species, predicted 76% of the variation in species density. The success of the predictor variables suggests that competitive intensity, as reflected in biomass levels, and the time elapsed since the last disturbance event, as reflected in the proportion of the vegetation composed of obligate perennials, are important determinants of local variation in species density. The model was then tested against independent data from shoreline vegetation in southeastern Ontario, Canada. There were no significant differences in the two data sets in their response to the two independent variables in the full model. However, only 42% of the variance in species density was explained in the combined data set.