Predicting aquatic macrophyte occurrence in soft-water oligotrophic lakes (Pyrenees mountain range)

Abstract

Distribution of aquatic macrophytes in lakes is related to geographical, morphological, catchment and water chemistry variables as well as human impacts, which modify the original environment. Here, we aim at building statistical models to establish the ecological niches of 11 aquatic macrophytes (10 different phanerogams and the genus Nitella ) from oligotrophic soft-water lakes and infer their ecological requirements and environmental constraints at the southernmost limit of their distribution. Macrophyte occurrence and environmental variables were obtained from 86 non-exploited oligotrophic soft-water lakes from the Pyrenees (Southern Europe; 42o50´N, 1o00´E); macrophytes inhabited 55 of these lakes. Optimum ranges and macrophyte occurrence were predicted in relation to 18 geographical, morphological, catchment and water chemistry variables using univariate and multivariate logistic models. Lakes at low altitude, in vegetated catchments and with low water concentration of NO 3 - and SO 4 -2 , were the most suitable to host macrophytes. In general, individual species of aquatic macrophytes showed clear patterns of segregation along conductivity and pH gradients, although the specific combination of variables selected in the best models explaining their occurrence differed among species. Based on the species response to pH and conductivity, we found Isoetes lacustris have its optimum in waters with low conductivity and pH (i.e. negative monotonic response). In contrast, Callitriche palustris , Ranunculus aquatilis , Subularia aquatica , Nitella spp., and Myriophyllum alterniflorum showed an optimum at intermediate values ( i.e. unimodal response), whereas Potamogeton berchtoldii , Potamogeton alpinus , and Ranunculus trichophyllus as species had their optimum at relatively high water pH and conductivity (i.e. positive monotonic response). This pattern has been observed in other regions for the same species, although with different optima and tolerance ranges. The parsimonious models developed here allowed us to explore niche requirements for each species and to predict the occurrence of macrophytes and of individual species in any particular lake by using only few, easily measured environmental variables. This is of interest to forecast possible changes related to global change and to assist managers in making conservation, management and restoration decisions.