Reconstruction of trophic state in Danish lakes using subfossil chydorid (Cladocera) assemblages

Surficial sediments from 32 Danish lakes were screened for subfossil chydorid (Cladocera) assemblages. Two data sets from 1965 and 1992 were examined by cluster analysis and nonmetric multidimensional scaling (NMDS) for lake classification. The best environmental variable to explain the 1992 species data was total phosphorus (TP) (r = 0.58, Spearman rank correlation), and NMDS defined three lake classes that were significantly different in communities (p < 0.01, ANOSIM) and in TP values. Canonical correspondence analysis found a strong relationship between the species data and log(TP) (axis 1 eigenvalue = 0.30, total inertia = 0.85, log(TP) intraset correlation = 0.92). Environmental variables were tested by forward selection and six variables were found to be statistically significant by 99 unrestricted Monte Carlo permutations. Weighted averaging regression and calibration models were established from the 1992 chydorid data and corresponding contemporary log-transformed TP data. The model's predictive ability was assessed by the bootstrapped coefficient of determination and the bootstrapped root mean squared error of prediction (r(boot)(2) = 0.79, RMSEboot = 0.24). The patterns of changes in chemical and biological variables (diversity, Bray-Curtis dissimilarity) over the 27-year period placed the 32 lakes in six different response groups.