A Test of Hypotheses to Explain the Sigmoidal Relationship between Total Phosphorus and Chlorophyll a Concentrations in Canadian Lakes

We tested five hypotheses to explain the sigmoidal relationship between total phosphorus (TP) and chlorophyll a (Chl a), two assuming that the nonlinearity is an artifact of various measurement biases and three assuming that it is based on underlying ecological interactions. Our first hypothesis was rejected; accounting for differences in extraction protocol of Chl a among published studies did not affect the sigmoidality. Our second hypothesis could not be rejected; there was an uncoupling between Chl a and PHYTO in oligotrophic lakes which may explain the initial nonlinearity. Our third hypothesis was upheld; the initial nonlinearity may be attributed to the presence of a disproportionately large fraction of unavailable phosphorus, since Chl a varied linearly with the proportion of total biologically active phosphorus in the TP fraction. The proportion of filamentous cyanophytes varied significantly with TP concentrations, and this was consistent with our fourth hypothesis that the higher Chl a:TP ratio at intermediate TP concentrations is attributable to reduced grazing impact of zooplankton in more productive lakes. Finally, Chl a varied linearly with total nitrogen, and this was consistent with our fifth hypothesis that the departure from linearity at extremely high phosphorus concentrations is indicative of nitrogen limitation.