A diatom-total phosphorus transfer function for lakes in the middle and lower reaches of Yangtze River

Diatom-based transfer function is established to provide an effective approach for the quantitative reconstruction of lake environmental indicators. Based on the investigation of water quality and surface diatoms from 45 lakes along the middle and lower reaches of Yangtze River, the relationships between diatoms and trophic variables were revealed by Canonical Correspondence Analysis (CCA). The result shown that total phosphorus (TP) can explain the most variation of diatom data, suggesting that it is the most important and significant variable which influenced on the diatom distributions. The inverse deshrinking weighted average model (WA) was chosen to establish the diatom-TP transfer function comparing with a series of WA models. According to the jackknifing statistical test the model gave a relative lower root-mean-square of error (RMSEPjack =0. 157). After deleted outliers, the new model improved the prediction ability greatly, with a higher coefficient oi prediction ( R2jack =0. 82) and a 21% decrease of predictive error ( RMSEPjack=0. 12). It performed much better comparing to other regional TP models. The establishment of our diatom-TP transfer functions will lay a scientific foundation for the trophic status reconstruction in typical lakes in Yangtze floodplain, which will likely to be an effective technique to set realistic targets for lake environment harnessing.