Diatom-based dissolved inorganic nitrogen reconstruction in the Changjiang River estuary and its adjacent areas

A five-component weighted average partial least squares (WA-PLS) calibration model was developed by analysing diatom assemblages in 34 surface sediment samples (collected in 2015) from the Changjiang River estuary (CRE) and its adjacent areas to infer dissolved inorganic nitrogen (DIN) concentrations. Eighteen additional sets of surface sediment diatoms and corresponding upper water DIN data (collected in 2012) were used to evaluate the accuracy of the model, and the relationship between observed and diatom-inferred DIN (DI-DIN) values (R-2=0.85) illustrated the strong performance of the transfer function, indicating that precise reconstructions of former DIN are possible. The diatom-DIN transfer function was applied to the diatom record from a sediment core DH8-2 (1962-2012) collected in the Fujian-Zhejiang area south of the CRE. The reconstruction based on the DI-DIN model showed a significant DIN increase from 1962-2012, reflecting the influence of human activities on the very large increase in eutrophication. Three distinct periods can be seen from the changes in DIN and diatom taxa. In the 1962-1972 period, the DIN content was relatively low, with an average of 5.94 mu mol/L, and more than 80% of the diatom species identified were benthic taxa. In the 1972-2004 period, as the impact of human activities intensified, large nutrient inputs caused the DIN content to increase, with an average of 8.25 mu mol/L. The nutrient inputs also caused a significant change in the nutrient components and a distinct increase in small planktonic taxa. In the 2004-2012 period, the DIN content continued to rise, fluctuating at approximately 10 mu mol/L. A continuous increase in the frequency of planktonic taxa (up to 65.48%) indicated that eutrophication was further intensified, which was confirmed by the transformation from diatom-induced red tide to dinoflagellate-induced red tide during this period.

Automated summary

A weighted average partial least squares (WA-PLS) calibration model was developed to infer dissolved inorganic nitrogen (DIN) concentrations in the Changjiang River estuary (CRE) and its adjacent areas. The model showed strong performance in reconstructing former DINs, and was applied to a sediment core from the Fujian-Zhejiang area to reveal a significant increase in DIN and eutrophication due to human activities.