Eutrophication and Climate Change Revisited

In their ambitious and welcome study, Sinha et al. (2017) concluded that eutrophication will worsen as a result of climate change in the 21st century. This conclusion was based on an analysis of nitrogen (N) loading driven by increased precipitation, alone. Although the overall conclusion of Sinha et al. (2017) may indeed ring true, their analytical approach appears open to question. We encourage the authors to address these questions as outlined below. This may provide the evidence to support the call for large-scale and costly nutrient reduction measures to safeguard freshwater, estuarine and coastal ecosystems.

First, is the model application valid? The statistical model used is extrapolated beyond the calibration period (Sinha et al., 2016) and despite the authors previously reporting a significant effect of temperature on N loading, the impact of warming was not included in their analysis.

Second, will changes in key processes of N-loss alter predictions? Warming is expected to result in less N entering our rivers directly from land and more N being lost from receiving waters through increased emissions of NH3, N2 and N2O. Warming by 5o C is expected to increase global NH3 emissions by 42% by 2100 (Sutton et al., 2013), while N2 and N2O production via denitrification in aquatic ecosystems may double with a 3o C temperature rise (Veraart et al., 2011). Denitrification is a major loss process in estuaries (Brock, 2001), and can decouple ecological effects from increased N loading, a scenario that is becoming increasingly common in warming lakes, regardless of N load (Weyhenmeyer et al., 2007).

Finally, how will freshwater, estuarine and coastal ecosystems respond to multiple and interacting climate stressors? Net ecological responses to multiple climate stressors must be confirmed before we can conclude that eutrophication will increase. For example, an increase in flushing rate and warming can alter the effects of increased nutrient loading on phytoplankton communities in rivers (Bowes et al., 2016), lakes (Carvalho et al., 2011) and estuaries (Harding et al., 2016). Such interactions suggest that future ecological responses may not be a simple function of N load. Predicting the impacts of climate change on eutrophication must be based on predictions of complex ecological responses to changes in multiple and interacting climate stressors, not just precipitation. We hope the authors will extend their model in such a way.