Decades of agricultural intensification have led to elevated stream nitrogen (N) concentrations and eutrophication of inland and coastal waters. Despite widespread implementation of a range of strategies to reduce N export, expected improvements in water quality have not been observed. This lack of success has often been attributed to the existence of legacy N stores within the landscape. Here, we use the ELEMeNT-N model to quantify legacy accumulation and depletion dynamics over the last century (1930–2016) across 14 nested basins within the Grand River Watershed, a 6800 km2 agricultural watershed in the Lake Erie Basin. Model results reveal significant legacy N accumulation across the basin, ranging from 705 to 1071 kg ha−1, creating a checkered landscape of N legacies. The largest proportion (82%–96%) of this accumulated N is stored in soil organic N reservoirs, as biogeochemical legacy, and the remaining in groundwater, as hydrologic legacy. The fraction of N surplus accumulated in soil and groundwater is most strongly correlated with the calibrated watershed mean travel time µ, with the accumulation increasing with increases in µ. The mean travel time ranges from 5 to 34 years across the watersheds studied, and increases with increase in tile drainage, highlighting the strong control of anthropogenic management on legacy accumulation. Water quality improvement timescales were found to be heterogeneous across the watersheds, with greater legacies contributing to slower recovery.

Increased use of phosphorus (P) fertilizers and intensive livestock production as well as increasing population densities and use of P-rich detergents have led to widespread eutrophication of both coastal and inland waters. Primary production in freshwater systems is strongly dependent on P availability (Schindler, 1974;Schindler et al., 2016), and P over-enrichment of lakes and reservoirs can lead to a proliferation of green algae, the development of toxic cyanobacterial blooms, and widespread impairment of aquatic ecosystems (Le Moal et al., 2019). Since 1950, global use of commercial P fertilizers has increased approximately four-fold (Van Vuuren et al., 2010). During this time, eutrophication has become a pervasive problem in human-impacted water bodies, from Lake Erie in North America to China's Lake Taihu (Sharpley & Wang, 2014).

Hydrologic model intercomparison studies help to evaluate the agility of models to simulate variables such as streamflow, evaporation, and soil moisture. This study is the third in a sequence of the Great Lakes Runoff Intercomparison Projects. The densely populated Lake Erie watershed studied here is an important international lake that has experienced recent flooding and shoreline erosion alongside excessive nutrient loads that have contributed to lake eutrophication. Understanding the sources and pathways of flows is critical to solve the complex issues facing this watershed. Seventeen hydrologic and land-surface models of different complexity are set up over this domain using the same meteorological forcings, and their simulated streamflows at 46 calibration and seven independent validation stations are compared. Results show that: (1) the good performance of Machine Learning models during calibration decreases significantly in validation due to the limited amount of training data; (2) models calibrated at individual stations perform equally well in validation; and (3) most distributed models calibrated over the entire domain have problems in simulating urban areas but outperform the other models in validation.

<p>Increased use of phosphorus (P) fertilizers and detergents, as well as the growth of animal feeding operations, have more than doubled P inputs to human-impacted watersheds over pre-industrial levels. While P fertilizer use and manure application help to maximize crop yields, excess P is lost to runoff, leading to eutrophication of downstream waters—a phenomenon of great concern in the North American Great Lakes region. Excess P also accumulates across the landscape, leading to legacies that serve as long-term sources of P to surface waters, even after inputs to the watershed are reduced. We developed, for the first time, a process-based model, ELEMeNT-P, designed to capture legacy P accumulation and depletion trajectories along the land-aquatic continuum. To drive the model, we reconstructed a more than 100-year trajectory of P inputs to the Grand River Watershed (GRW), Canada’s largest river basin draining directly to Lake Erie. Our results show that since 1900 the GRW has served as a net P sink, with an estimated accumulation of more than 480 ktons P, of which 89% resides in soils and 6% in reservoirs and riparian areas. Future simulations suggest that while a 40% reduction in P discharge to Lake Erie is possible under aggressive management scenarios, legacy P will continue to elevate P loads to Lake Erie for centuries.</p>

No abstract