Variations and Spurious Correlations Related to DIN, DIP, TN, TP, DIN/DIP and TN/TP in Lakes, Rivers, and Marine Systems

ABSTRACT

Inherent variations and uncertainties in empirical data from aquatic systems ultimately constrain approaches to predictions and possibilities to identify critical thresholds and points of no return. This work addresses coefficients of variations (CV) for total phosphorus (TP), total nitrogen (TN), dissolved inorganic nitrogen (DIN), dissolved inorganic phosphorus (DIP), DIN/DIP and TN/TP. We have collected data from 28 lakes, 34 rivers, and 38 stations in the Baltic Sea, Kattegat, the Danish Straits and Skagerrak. Key questions are: How do the CV-values vary among lakes, rivers and marine systems? Are there patterns in the CVs related to the trophic conditions? Are there seasonal patterns in the CVs? Selected results: DIN varies most of the studied nutrients. DIP displays a relatively homogenous variation with CVs of about 0.3 to 0.5. The variability in DIN governs the high uncertainty in the DIN/DIP-ratio (CV = 0.8 to 1.4). The CVs for TN/TP are much lower (0.4 to 0.5). The CVs for DIN and TN increase from the Bothnian Bay to the Skagerrak. There is a pattern of lower variabilities in oligotrophic lakes for DIN/DIP as compared to TN/TP. Results from an experiment to illustrate spurious correlations (using randomly generated data) are also given, which clearly show that for low-productive systems, one can find almost any ratio of DIN/DIP or TN/TP. This implies that one must be very cautious in interpretations of what DIN/DIP or TN/TP actually represent in low-productive systems.

Keywords: