Species richness and beta diversity of aquatic macrophytes in a large subtropical reservoir (Itaipu Reservoir, Brazil): the influence of limnology and morphometry

Assessment of the environmental factors that control species richness (S) is a central issue in ecology. In this study, aquatic macrophyte S was estimated in 235 sampling sites distributed in 8 arms of a large (1350 km²) subtropical reservoir (Itaipu Reservoir, Brazil). Morphometric variables (area, shoreline development and length of shoreline, all measured for each arm; n= 8) and environmental variables measured at each sampling site (extinction coefficient of light (k), electrical conductivity, fetch, distance from the main reservoir body; n = 235) were used to predict aquatic macrophyte S at two spatial scales. At arm scale, linear regression analysis indicated that length of shoreline was a better predictor of S than area. At sampling site scale, multiple regression analysis indicated that S was significantly predicted by electrical conductivity, fetch and distance from the main body. However, other relationships with predictive interest was demonstrated by using non-traditional regression approaches. This analysis started by the visual inspection of scatter plots. The bivariate relationship between S and fetch, for example, showed an envelope or a `left triangle' pattern. The relationship between the number of submerged species and k showed an asymmetrical left triangle pattern. Using randomization procedures, it was demonstrated that these patterns were not generated by chance alone. Beta diversity (estimated within the arms) was significantly and positively correlated with spatial environmental variability. Overall, these results indicate that the prediction of aquatic macrophytes assemblage variables in large waterbodies, specially S, is more complex than previous studies have suggested.