Alex S. Mayer Veronica Griffis David Jr. Watkins Dale M. Robertson Rabi Gyawali Meredith B. LaBeau 2015 <p><span>In this work, we hypothesize that phosphorus (P) concentrations in streams vary seasonally and with streamflow and that it is important to incorporate this variation when predicting changes in P loading associated with climate change. Our study area includes 14 watersheds with a range of land uses throughout the U.S. Great Lakes Basin. We develop annual seasonal load-discharge regression models for each watershed and apply these models with simulated discharges generated for future climate scenarios to simulate future P loading patterns for two periods: 2046&ndash;2065 and 2081&ndash;2100. We utilize output from the Coupled Model Intercomparison Project phase 3 downscaled climate change projections that are input into the Large Basin Runoff Model to generate future discharge scenarios, which are in turn used as inputs to the seasonal P load regression models. In almost all cases, the seasonal load-discharge models match observed loads better than the annual models. Results using the seasonal models show that the concurrence of nonlinearity in the load-discharge model and changes in high discharges in the spring months leads to the most significant changes in P loading for selected tributaries under future climate projections. These results emphasize the importance of using seasonal models to understand the effects of future climate change on nutrient loads.</span></p> application/pdf 10.1007/s10533-015-0149-5 en Springer The importance of considering shifts in seasonal changes in discharges when predicting future phosphorus loads in streams article