Lake biology as a function of catchment characteristics and water quality parameters – focus on phytoplankton

Abstract: According to the EU Water Framework Directive, phytoplankton should be included for assessment of the ecological status of lakes. Phytoplankton respond rapidly to environmental changes and are a particularly good indicator of nutrient loads. In this large-scale study, it was examined whether the properties in the catchment area can be used to explain the variation of phytoplankton and total phosphorus (TP) in lakes. A large number of variables were investigated through statistical analysis, in particular if the already established linear relationship between phytoplankton and TP in lakes can be improved. The study used measured values for total biovolume phytoplankton (tot bio), Plankton Trophic Index (PTI) and TP for 487 lakes (represented by 523 waterbodies) in south of Sweden. The lakes associated catchment properties were calculated and analysed through various Geographical Information System (GIS) tools. Each catchment was described regarding land use, soil properties (texture and chemistry), soil distribution, climate and lake properties. In total 59 variables produced with GIS were evaluated from available map data and national soil surveys together with 34 lake variables. All variables were used in Principal Components Analysis (PCA) whereas the 59 catchment variables together with some lake variables were used for other analysis. For each dependent variable (tot bio, PTI and TP) several statistical models were created, and important catchment variables were identified using Partial Least Squares (PLS) analysis. Important variables identified in PLS were then included in multiple regressions. Result shows that the share of agricultural land in the catchment area is positively correlated with phytoplankton (tot bio and PTI) and TP. For phytoplankton models without TP as explanatory variable, a few catchment variables could explain variation of PTI up to 48 % while tot bio could be explained to a lower extent (33 %). The degree of explanation and variables included differed depending on selected statistical model. While TP alone was the strongest explaining variable for both tot bio (66 %) and PTI (56 %). However, TP together with the share of agricultural land significantly improved the explanation of PTI to 65 %. For the lake TP, catchment properties could statistically explain 55 % of the variation in the TP concentration. In summary, TP was shown to correlate positively with specific soil properties of both non-agricultural and agricultural areas of studied catchments. Higher TP concentration could also be expected in lakes with larger share of agricultural land and urban area and smaller water body area. The results also show that catchment properties derived from continuous map data had a higher explanation of the studied lakes’ tot bio, PTI and TP compared to result from field sample point data collected in national soil surveys. The relationship between catchment properties and water quality is important to understand and catchment properties can help to describe the lake phytoplankton and phosphorus levels, which then should be taken into account when developing assessment criteria for lakes.