Abstract Excess nutrient inputs from agricultural and urban sources have accelerated eutrophication and increased the incidence of algal blooms in the Great Lakes Basin (GLB). Lake basin management to address these threats relies on understanding the key drivers of pollution. Here, we use a random forest machine learning model to leverage information from 202 monitored streams in the GLB to predict seasonal and annual flow‐weighted concentrations of nitrogen and phosphorus, as well as nutrient ratios across the GLB. Land use (agricultural and urban land) and land management (tile drainage and wetland density) emerge as the two most important predictors for dissolved inorganic nitrogen (DIN; NO 3 − + NO 2 − ) and soluble reactive phosphorus (SRP; PO 4 3 ), while soil type and wetland density are more important for particulate P (PP). Partial dependence plots demonstrate increasing nutrient concentrations with increasing tile density and decreasing wetland density. In addition, increasing tile and livestock densities and decreasing forest cover correspond to higher SRP:Total Phosphorus (TP) ratios. Seasonally, the highest proportions of SRP occur in summer and fall. Higher livestock densities are also correlated with increasing N:P (DIN:TP) ratios. Livestock operations can contribute to the buildup of soil nutrients from excess manure application, while increasing subsurface drainage can provide transport pathways for dissolved nutrients. Given that both SRP:TP and the N:P ratios are strong predictors of harmful algal blooms, our study highlights the importance of livestock management, drainage management, and wetland restoration in efforts to address eutrophication in intensively managed landscapes.

Abstract Excess nutrient inputs from agricultural and urban sources have accelerated eutrophication and increased the incidence of algal blooms in the Great Lakes Basin (GLB). Lake basin management to address these threats relies on understanding the key drivers of pollution. Here, we use a random forest machine learning model to leverage information from 202 monitored streams in the GLB to predict seasonal and annual flow‐weighted concentrations of nitrogen and phosphorus, as well as nutrient ratios across the GLB. Land use (agricultural and urban land) and land management (tile drainage and wetland density) emerge as the two most important predictors for dissolved inorganic nitrogen (DIN; NO 3 − + NO 2 − ) and soluble reactive phosphorus (SRP; PO 4 3 ), while soil type and wetland density are more important for particulate P (PP). Partial dependence plots demonstrate increasing nutrient concentrations with increasing tile density and decreasing wetland density. In addition, increasing tile and livestock densities and decreasing forest cover correspond to higher SRP:Total Phosphorus (TP) ratios. Seasonally, the highest proportions of SRP occur in summer and fall. Higher livestock densities are also correlated with increasing N:P (DIN:TP) ratios. Livestock operations can contribute to the buildup of soil nutrients from excess manure application, while increasing subsurface drainage can provide transport pathways for dissolved nutrients. Given that both SRP:TP and the N:P ratios are strong predictors of harmful algal blooms, our study highlights the importance of livestock management, drainage management, and wetland restoration in efforts to address eutrophication in intensively managed landscapes.

Abstract Excess nitrogen from intensive agricultural production, atmospheric N deposition, and urban point sources elevates stream nitrate concentrations, leading to problems of eutrophication and ecosystem degradation in coastal waters. A major emphasis of current US‐scale analysis of water quality is to better our understanding of the relationship between changes in anthropogenic N inputs within watersheds and subsequent changes in riverine N loads. While most water quality modeling assumes a positive linear correlation between watershed N inputs and riverine N, many efforts to reduce riverine N through improved nutrient management practices result in little or no short‐term improvements in water quality. Here, we use nitrate concentration and load data from 478 US watersheds, along with developed N input trajectories for these watersheds, to quantify time‐varying relationships between N inputs and riverine N export. Our results show substantial variations in watershed N import‐export relationships over time, with quantifiable hysteresis effects. Our results show that more population‐dense urban watersheds in the northeastern U.S. more frequently show clockwise hysteresis relationships between N imports and riverine N export, with accelerated improvements in water quality being achieved through the implementation of point‐source controls. In contrast, counterclockwise hysteresis dynamics are more common in agricultural watersheds, where time lags occur between the implementation of nutrient management practices and water‐quality improvements. Finally, we find higher tile‐drainage densities to be associated with more linear relationships between N inputs and riverine N. The empirical analysis in this study is bolstered by modeled simulations to reproduce and further explain drivers behind the hysteretic relationships commonly observed in the monitored watersheds.

Abstract Excess nitrogen from intensive agricultural production, atmospheric N deposition, and urban point sources elevates stream nitrate concentrations, leading to problems of eutrophication and ecosystem degradation in coastal waters. A major emphasis of current US‐scale analysis of water quality is to better our understanding of the relationship between changes in anthropogenic N inputs within watersheds and subsequent changes in riverine N loads. While most water quality modeling assumes a positive linear correlation between watershed N inputs and riverine N, many efforts to reduce riverine N through improved nutrient management practices result in little or no short‐term improvements in water quality. Here, we use nitrate concentration and load data from 478 US watersheds, along with developed N input trajectories for these watersheds, to quantify time‐varying relationships between N inputs and riverine N export. Our results show substantial variations in watershed N import‐export relationships over time, with quantifiable hysteresis effects. Our results show that more population‐dense urban watersheds in the northeastern U.S. more frequently show clockwise hysteresis relationships between N imports and riverine N export, with accelerated improvements in water quality being achieved through the implementation of point‐source controls. In contrast, counterclockwise hysteresis dynamics are more common in agricultural watersheds, where time lags occur between the implementation of nutrient management practices and water‐quality improvements. Finally, we find higher tile‐drainage densities to be associated with more linear relationships between N inputs and riverine N. The empirical analysis in this study is bolstered by modeled simulations to reproduce and further explain drivers behind the hysteretic relationships commonly observed in the monitored watersheds.

Management strategies aimed at reducing nutrient enrichment of surface waters may be hampered by nutrient legacies that have accumulated in the landscape. Here, we apply the Net Anthropogenic Phosphorus Input (NAPI) model to reconstruct the historical phosphorus (P) input trajectories for the province of Ontario, which encompasses the Canadian portion of the drainage basin of the Laurentian Great Lakes (LGL). NAPI considers P inputs from detergent, human and livestock waste, fertilizer inputs, and P outputs by crop uptake. During the entire time period considered, from 1961 to 2016, Ontario experienced positive annual NAPI values. Despite a generally downward NAPI trend since the late 1970s, the lower LGL, especially Lake Erie, continue to be plagued by algal blooms. When comparing NAPI results and river monitoring data for the period 2003 to 2013, P discharged by Canadian rivers into Lake Erie only accounts for 12.5% of the NAPI supplied to the watersheds' agricultural areas. Thus, over 85% of the agricultural NAPI is retained in the watersheds where it contributes to a growing P legacy, primarily as soil P. The slow release of legacy P therefore represents a long-term risk to the recovery of the lake. To help mitigate this risk, we present a methodology to spatially map out the source areas with the greatest potential of erosional export of legacy soil P to surface waters. These areas should be prioritized in soil conservation efforts.

Increasing incidences of eutrophication and groundwater quality impairment from agricultural nitrogen pollution are threatening humans and ecosystem health. Minimal improvements in water quality have been achieved despite billions of dollars invested in conservation measures worldwide. Such apparent failures can be attributed in part to legacy nitrogen that has accumulated over decades of agricultural intensification and that can lead to time lags in water quality improvement. Here, we identify the key knowledge gaps related to landscape nitrogen legacies and propose approaches to manage and improve water quality, given the presence of these legacies.

Increased fluxes of reactive nitrogen (N r ), often associated with N fertilizer use in agriculture, have resulted in negative environmental consequences, including eutrophication, which cost billions of dollars per year globally. To address this, best management practices (BMPs) to reduce N r loading to the environment have been introduced in many locations. However, improvements in water quality associated with BMP implementation have not always been realised over expected timescales. There is a now a significant body of scientific evidence showing that the dynamics of legacy N r storage and associated time lags invalidate the assumptions of many models used by policymakers for decision making regarding N r BMPs. Building on this evidence, we believe that the concepts of legacy N r storage dynamics and time lags need to be included in these models. We believe the biogeochemical research community could play a more proactive role in advocating for this change through both awareness raising and direct collaboration with policymakers to develop improved datasets and models. We anticipate that this will result in more realistic expectations of timescales for water quality improvements associated with BMPs. Given the need for multi-nutrient policy responses to tackle challenges such as eutrophication, integration of N stores will have the further benefit of aligning both researchers and policymakers in the N community with the phosphorus and carbon communities, where estimation of stores is more widespread. Ultimately, we anticipate that integrating legacy N r storage dynamics and time lags into policy frameworks will better meet the needs of human and environmental health. • Nitrogen (N) pollution from agriculture has negative environmental impacts. • Environmental benefits of initiatives to reduce N loads not always detectable. • N storage dynamics and time lag invalidate steady state models often used in policy. • Researchers should advocate for integrating N stores and time lags into policy. • Quantifying N storage aligns with phosphorus and carbon cycling research.

The Upper Mississippi River Basin is the largest source of reactive nitrogen (N) to the Gulf of Mexico. Concentration‐discharge (C‐Q) relationships offer a means to understand both the terrestrial sources that generate this reactive N and the in‐stream processes that transform it. Progress has been made on identifying land use controls on C‐Q dynamics. However, the impact of basin size and river network structure on C‐Q relationships is not well characterized. Here, we show, using high‐resolution nitrate concentration data, that tile drainage is a dominant control on C‐Q dynamics, with increasing drainage density contributing to more chemostatic C‐Q behavior. We further find that concentration variability increases, relative to discharge variability, with increasing basin size across six orders of magnitude, and this pattern is attributed to different spatial correlation structures for C and Q. Our results show how land use and river network structure jointly control riverine N export.

Reactive nitrogen (N) fluxes have increased tenfold over the last century, driven by increases in population, shifting diets, and increased use of commercial N fertilizers. Runoff of excess N from intensively managed landscapes threatens drinking water quality and disrupts aquatic ecosystems. Excess N is also a major source of greenhouse gas emissions from agricultural soils. While N emissions from agricultural landscapes are known to originate from not only current‐year N input but also legacy N accumulation in soils and groundwater, there has been limited access to fine‐scale, long‐term data regarding N inputs and outputs over decades of intensive agricultural land use. In the present work, we synthesize population, agricultural, and atmospheric deposition data to develop a comprehensive, 88‐year (1930–2017) data set of county‐scale components of the N mass balance across the contiguous United States (Trajectories Nutrient Dataset for nitrogen [TREND‐nitrogen]). Using a machine‐learning algorithm, we also develop spatially explicit typologies for components of the N mass balance. Our results indicate a large range of N trajectory behaviors across the United States due to differences in land use and management and particularly due to the very different drivers of N dynamics in densely populated urban areas compared with intensively managed agricultural zones. Our analysis of N trajectories also demonstrates a widespread functional homogenization of agricultural landscapes. This newly developed typology of N trajectories improves our understanding of long‐term N dynamics, and the underlying data set provides a powerful tool for modeling the impacts of legacy N on past, present, and future water quality.

Growing populations and agricultural intensification have led to raised riverine nitrogen (N) loads, widespread oxygen depletion in coastal zones (coastal hypoxia)1 and increases in the incidence of algal blooms.Although recent work has suggested that individual wetlands have the potential to improve water quality2,3,4,5,6,7,8,9, little is known about the current magnitude of wetland N removal at the landscape scale. Here we use National Wetland Inventory data and 5-kilometre grid-scale estimates of N inputs and outputs to demonstrate that current N removal by US wetlands (about 860 ± 160 kilotonnes of nitrogen per year) is limited by a spatial disconnect between high-density wetland areas and N hotspots. Our model simulations suggest that a spatially targeted increase in US wetland area by 10 per cent (5.1 million hectares) would double wetland N removal. This increase would provide an estimated 54 per cent decrease in N loading in nitrate-affected watersheds such as the Mississippi River Basin. The costs of this increase in area would be approximately 3.3 billion US dollars annually across the USA—nearly twice the cost of wetland restoration on non-agricultural, undeveloped land—but would provide approximately 40 times more N removal. These results suggest that water quality improvements, as well as other types of ecosystem services such as flood control and fish and wildlife habitat, should be considered when creating policy regarding wetland restoration and protection.

Land use change and agricultural intensification have increased food production but at the cost of polluting surface and groundwater. Best management practices implemented to improve water quality have met with limited success. Such lack of success is increasingly attributed to legacy nutrient stores in the subsurface that may act as sources after reduction of external inputs. However, current water‐quality models lack a framework to capture these legacy effects. Here we have modified the SWAT (Soil Water Assessment Tool) model to capture the effects of nitrogen (N) legacies on water quality under multiple land‐management scenarios. Our new SWAT‐LAG model includes (1) a modified carbon‐nitrogen cycling module to capture the dynamics of soil N accumulation, and (2) a groundwater travel time distribution module to capture a range of subsurface travel times. Using a 502‐km2 Iowa watershed as a case study, we found that between 1950 and 2016, 25% of the total watershed N surplus (N Deposition + Fertilizer + Manure + N Fixation − Crop N uptake) had accumulated within the root zone, 14% had accumulated in groundwater, while 27% was lost as riverine output, and 34% was denitrified. In future scenarios, a 100% reduction in fertilizer application led to a 79% reduction in stream N load, but the SWAT‐LAG results suggest that it would take 84 years to achieve this reduction, in contrast to the 2 years predicted in the original SWAT model. The framework proposed here constitutes a first step toward modifying a widely used modeling approach to assess the effects of legacy N on the time required to achieve water‐quality goals.

Changes in seasonal nutrient dynamics are occurring across a range of climates and land use types. Although it is known that seasonal patterns in nutrient availability are key drivers of both stream metabolism and eutrophication, there has been little success in developing a comprehensive understanding of seasonal variations in nutrient export across watersheds or of the relationship between nutrient seasonality and watershed characteristics. In the present study, we have used concentration and discharge data from more than 200 stations across U.S. and Canadian watersheds to identify (1) archetypal seasonal concentration regimes for nitrate, soluble reactive phosphorus (SRP), and total phosphorus, and (2) dominant watershed controls on these regimes across a gradient of climate, land use, and topography. Our analysis shows that less impacted watersheds, with more forested and wetland area, most commonly exhibit concentration regimes that are in phase with discharge, with concentration lows occurring during summer low‐flow periods. Agricultural watersheds also commonly exhibit in‐phase behavior, though the seasonality is usually muted compared to that seen in less impacted areas. With increasing urban area, however, nutrient concentrations frequently become essentially aseasonal or even exhibit clearly out‐of‐phase behavior. In addition, our data indicate that seasonal SRP concentration patterns may be strongly influenced by proximal controls such as the presence of dams and reservoirs. In all, these results suggest that human activity is significantly altering nutrient concentration regimes, with large potential consequences for both in‐stream metabolism and eutrophication risk in downstream waterbodies.

Ballard et al . argue that our prediction of a 30-year or longer recovery time for Gulf of Mexico water quality is highly uncertain, and that much shorter time lags are equally likely. We demonstrate that their argument, based on the use of a two-component regression model, does not sufficiently consider fundamental watershed processes or multiple lines of evidence suggesting the existence of decadal-scale lags.

Groundwater discharge can be a major source of nutrients to river systems. Although quantification of groundwater nitrate loading to streams is common, the dependence of surface water silicon (Si) and phosphorus (P) concentrations on groundwater sources has rarely been determined. Additionally, the ability of groundwater discharge to drive surface water Si:P ratios has not been contextualized relative to riverine inputs or in-stream transformations. In this study, we quantify the seasonal dynamics of Si and P cycles in the Grand River (GR) watershed, the largest Canadian watershed draining into Lake Erie, to test our hypothesis that regions of Si-rich groundwater discharge increase surface water Si:P ratios. Historically, both the GR and Lake Erie have been considered stoichiometrically P-limited, where the molar Si:P ratio is greater than the ~16:1 phytoplankton uptake ratio. However, recent trends suggest that eastern Lake Erie may be approaching Si-limitation. We sampled groundwater and surface water for dissolved and reactive particulate Si as well as total dissolved P for 12months within and downstream of a 50-km reach of high groundwater discharge. Our results indicate that groundwater Si:P ratios are lower than the corresponding surface water and that groundwater is a significant source of bioavailable P to surface water. Despite these observations, the watershed remains P-limited for the majority of the year, with localized periods of Si-limitation. We further find that groundwater Si:P ratios are a relatively minor driver of surface water Si:P, but that the magnitude of Si and P loads from groundwater represent a large proportion of the overall fluxes to Lake Erie.

Haunted by the past Reducing the extent of hypoxia in the Gulf of Mexico will not be as easy as reducing agricultural nitrogen use. Van Meter et al. report that so much nitrogen from runoff has accumulated in the Mississippi River basin that, even if future agricultural nitrogen inputs are eliminated, it will still take 30 years to realize the 60% decrease in load needed to reduce eutrophication in the Gulf. This legacy effect means that a dramatic shift in land-use practices, which may not be compatible with current levels of agricultural production, will be needed to control hypoxia in the Gulf of Mexico. Science , this issue p. 427

Unprecedented decreases in atmospheric nitrogen (N) deposition together with increases in agricultural N-use efficiency have led to decreases in net anthropogenic N inputs in many eastern US and Canadian watersheds as well as in Europe. Despite such decreases, N concentrations in streams and rivers continue to increase, and problems of coastal eutrophication remain acute. Such a mismatch between N inputs and outputs can arise due to legacy N accumulation and subsequent lag times between implementation of conservation measures and improvements in water quality. In the present study, we quantified such lag times by pairing long-term N input trajectories with stream nitrate concentration data for 16 nested subwatersheds in a 6800 km2, Southern Ontario watershed. Our results show significant nonlinearity between N inputs and outputs, with a strong hysteresis effect indicative of decadal-scale lag times. The mean annual lag time was found to be 24.5 years, with lags varying seasonally, likely due to differences in N-delivery pathways. Lag times were found to be negatively correlated with both tile drainage and watershed slope, with tile drainage being a dominant control in fall and watershed slope being significant during the spring snowmelt period. Quantification of such lags will be crucial to policy-makers as they struggle to set appropriate goals for water quality improvement in human-impacted watersheds.