Identification of phosphorus emission hotspots in agricultural catchments

Particulate PhozzyLogic Index for policy makers-an index for a more accurate and transparent identification of critical source areas

The identification of critical source areas (CSAs) is a key element in a cost-effective mitigation of diffuse emissions of phosphorus from agricultural soils into surface waters. One of the challenges related to CSAs is how to couple complex, data-intensive fate and transport models with easy-to-use information on field level for management purposes at the scale of large watersheds. To fill such a gap and create a bridge between the two tasks, this study puts forward the new Particulate PhozzyLogic Index (PPLI) based on the innovative combination of the results of a complex watershed model (in this case the PhosFate model) with fuzzy logic. Its main feature is the ability to transform the results of diverse scenarios or even models into a final map showing a catchment-wide ranking of the possibility of high PP emissions reaching surface waters for all agricultural fields. Further, this study enhances the PhosFate model with a new algorithm for the allocation of particulate phosphorus (PP) loads entering surface waters to their sources of origin. This is a basic requirement for the identification of critical PP source areas and in consequence for a cost-effective implementation of mitigation measures. By means of a sensitivity analysis, this study investigates the impacts of storm drains, discharge frequencies and flow directions on the designation of CSAs with the help of present-day scenarios for a case study catchment with an area of several hundred square kilometres. The upfront model calibration exhibits a Nash-Sutcliffe efficiency (NSE) of about 0.95 and a modified Nash-Sutcliffe efficiency (mNSE) of around 0.83. A core result of the sensitivity analysis is that the scenarios at least partially disagree on the identified CSAs and suggest that especially open furrows at field borders have the potential to lead to deviating outcomes. All scenario results nevertheless support the 80:20 rule, which states that about 80% of the phosphorus inputs into the surface waters of a catchment originate from only about 20% of its area.

The changing face of science communication, technology, extension and improved decision-making at the farm-water quality interface

In recent decades, significant advances have been made in understanding the generation, fates and consequences of water quality pollutants in the Great Barrier Reef ecosystem. However, skepticism and lack of trust in water quality science by farming stakeholders has emerged as a significant challenge. The ongoing failures of both compulsory and particularly voluntary practices to improve land management and reduce diffuse agricultural pollution from the Great Barrier Reef catchment underlines the need for more effective communication of water quality issues at appropriate decision-making scales to landholders. Using recent Great Barrier Reef catchment experiences as examples, we highlight several emerging themes and opportunities in using technology to better communicate land use-water quality impacts and delivery of actionable knowledge to farmers, specifically supporting decision-making, behavior change, and the spatial identification of nutrient generation 'hotspots' in intensive agriculture catchments. We also make recommendations for co-designed monitoring-extension platforms involving farmers, governments, researchers, and related agencies, to cut across stakeholder skepticism, and achieve desired water quality and ecosystem outcomes.

Uncertainty in critical source area predictions from watershed-scale hydrologic models

Watershed-scale hydrologic models are frequently used to inform conservation and restoration efforts by identifying critical source areas (CSAs; alternatively 'hotspots'), defined as areas that export relatively greater quantities of nutrients and sediment. The CSAs can then be prioritized or 'targeted' for conservation and restoration to ensure efficient use of limited resources. However, CSA simulations from watershed-scale hydrologic models may be uncertain and it is critical that the extent and implications of this uncertainty be conveyed to stakeholders and decision makers. We used an ensemble of four independently developed Soil and Water Assessment Tool (SWAT) models and a SPAtially Referenced Regression On Watershed attributes (SPARROW) model to simulate CSA locations for flow, phosphorus, nitrogen, and sediment within the ~17,000-km² Maumee River watershed at the HUC-12 scale. We then assessed uncertainty in CSA simulations determined as the variation in CSA locations across the models. Our application of an ensemble of models - differing with respect to inputs, structure, and parameterization - facilitated an improved accounting of CSA prediction uncertainty. We found that the models agreed on the location of a subset of CSAs, and that these locations may be targeted with relative confidence. However, models more often disagreed on CSA locations. On average, only 16%-46% of HUC-12 subwatersheds simulated as a CSA by one model were also simulated as a CSA by a different model. Our work shows that simulated CSA locations are highly uncertain and may vary substantially across models. Hence, while models may be useful in informing conservation and restoration planning, their application to identify CSA locations would benefit from comprehensive uncertainty analyses to avoid inefficient use of limited resources.

Nutrient mitigation under the impact of climate and land-use changes: A hydro-economic approach to participatory catchment management

Excessive nutrient loadings into rivers are a well-known ecological problem. Implemented mitigation measures should ideally be cost-effective, but perfectly ranking alternative nutrient mitigation measures according to cost-effectiveness is a difficult methodological challenge. Furthermore, a particularly practical challenge is that cost-effective measures are not necessarily favoured by local stakeholders, and this may impede their successful implementation in practice. The objective of this study was to evaluate the cost-effectiveness of mitigation measures using a methodology that includes a participatory process and social learning to ensure their successful implementation. By combining cost data, hydrological modelling and a bottom-up approach for three different European catchment areas (the Latvian Berze, the Swedish Helge and the German Selke rivers), the cost-effectiveness of 16 nutrient mitigation measures were analysed under current conditions as well as under selected scenarios for future climate and land-use changes. Fertiliser reduction, wetlands, contour ploughing and municipal wastewater treatment plants are the measures that remove nutrients with the highest cost-effectiveness in the respective case study context. However, the results suggest that the cost-effectiveness of measures not only depends on their design, specific location and the conditions of the surrounding area, but is also affected by the future changes the area may be exposed to. Climate and land-use changes do not only affect the cost-effectiveness of measures, but also shape the overall nutrient loads and potential target levels in a catchment.

Coupled hydrological and biogeochemical modelling of nitrogen transport in the karst critical zone

Transport of nitrogen (N) in karst areas is more complex than in non-karst areas due to marked heterogeneity of hydrodynamic behaviour in the karst critical zone. Here, we present a novel, distributed, coupled hydrological-biogeochemical model that can simulate water and nitrogen transport in the critical zone of karst catchments. This new model was calibrated using integrated hydrometric, water stable isotope, and nitrogen-N concentration data at the outflow of Houzhai catchment in Guizhou province of Southwest China. Hydrological dynamics appears to control N load from the study catchment. Combining flow discharge and water stable isotopes significantly constrained model parameterisation and mitigate the equifinality effects of parameters on the simulated results. Karst geomorphology and land use have functional effects on spatiotemporal variations of hydrological processes and nitrogen transport. In the study catchment, agricultural fertilizer was the largest input source of N, accounting for 86% of the total. Plant uptake consumed about 45% of inputs, primarily in the low-lying valley bottom areas and the plain covered by relatively thick soils. Thus, a large amount of N released from soil reservoirs to the epikarst (via fractures or sinkholes) is then exported to the underground channel in the limestone area to the south. This N draining into groundwater could lead to extensive, potentially long-term contamination of the karst system. Therefore, improving the efficiency of fertilization and agricultural management in valleys/depressions is an urgent need to reduce N losses and contamination risk.

Identifying Flow Pathways for Phosphorus Transport Using Observed Event Forensics and the CRAFT (Catchment Runoff Attenuation Flux Tool)

Identifying key flow pathways is critical in order to understand the transport of Phosphorus (P) from agricultural headwater catchments. High frequency/resolution datasets from two such catchments in Northwest England enabled individual events to be examined to identify the flow (Q) and Total P (TP) and Total Reactive P (TRP) dynamics (forensics). Detailed analysis of multiple flow and water quality parameters is referred to here as the event forensics. Are there more flow pathways than just surface runoff (dominated by overland flow) and baseflow (mainly groundwater) contributing at the outlet of these catchments? If so, hydrograph separation alone will not be sufficient. This forensic analysis gives a classification of four storm event response types. Three classes are based on the balance of old and new water giving enrichment and dilution of TRP pattern in the subsurface flow. A fourth type was observed where a plume of nutrient is lost to the channel when there is no observed flow. Modelling is also essential when used in combination with the event forensics as this additional tool can identify distinct flow pathways in a robust form. A case study will apply the Catchment Runoff Attenuation Flux Tool (CRAFT) to two contrasting small headwater catchments in Northwest England, which formed part of the Demonstration Test Catchments (DTC) Programme. The model will use data collected during a series of events observed in the two catchments between the period 2012 and 2014. It has the ability to simulate fast near surface (that can represent flow in the upper soil horizons and field drains) and event subsurface soil flow, plus slower groundwater discharge. The model can capture P enrichment, dilution and the role that displacement of “old” P rich water has during events by mixing these flows. CRAFT captures the dominant flow and P fluxes as seen in the forensic analysis and can create outputs including smart export coefficients (based on flow pathways) that can be conveyed to policy makers to better underpin decision making.

Logistics Network Management of Livestock Waste for Spatiotemporal Control of Nutrient Pollution in Water Bodies

Nutrient pollution is a widespread water quality problem, which originates from excess nutrient runoff from agricultural land, improperly managed farming operations, and point sources such as wastewater treatment plants. Some nutrient pollution impacts include harmful algal blooms (HABs), hypoxia, and eutrophication. HABs are major environmental events that cause severe health threats and economic losses (e.g., tourism, real estate, commercial fishing). A dimension of the nutrient pollution problem that has not received much attention is that this interacts with organic waste management practices. As a result, it is important to connect the time and location of point and nonpoint nutrient source releases, nutrient soil content, spatial layout, and hydrology of agricultural lands with the transport of nutrients to water bodies and their impacts on aquatic ecosystems. In this work, we show how nutrient concentration in water bodies and other spatiotemporal factors are related to HAB development and how logistics management of livestock waste can be used to conduct space-time management of nutrient pollution. A case study for the Upper Yahara Watershed in the State of Wisconsin (U.S.) is employed to demonstrate the practicability of the modeling framework. Our framework reveals that logistics network management for waste and nutrients can reduce the incidence rates of HABs, but reducing it to nonharmful levels would require long-term efforts such as installing nutrient recovery technologies, coordinating manure storage and application, and deploying management incentive plans.

Assessing the Impact of Storm Drains at Road Embankments on Diffuse Particulate Phosphorus Emissions in Agricultural Catchments

This study presents a simple mapping key suitable for quick and systematic assessments of the types of agricultural and civil engineering structures present in a certain agricultural catchment as well as the impact they may have on the spatial distribution of critical source areas. An application of this mapping key to three small sub-catchments of a case study catchment with an area of several hundred square kilometres (one-stage cluster sampling) in Austria clearly reveals that road embankments with subsurface drainage can exert a major influence on emissions and transport pathways of sediment-bound pollutants like particulate phosphorus (PP). Due to this, the semi-empirical, spatially distributed PhosFate model is extended to separately model PP emissions into surface waters via storm drains along road embankments. Furthermore, the overall share of road embankments with subsurface drainage on all road embankments in the case study catchment is inferred with the help of a Bayesian hierarchical model. The combination of the results of these two models shows that the share of storm drains at road embankments on total PP emissions ranges from about one fifth to one third in the investigated area.

The Role of Attenuation and Land Management in Small Catchments to Remove Sediment and Phosphorus: A Modelling Study of Mitigation Options and Impacts

It is well known that soil, hillslopes, and watercourses in small catchments possess a degree of natural attenuation that affects both the shape of the outlet hydrograph and the transport of nutrients and sediments. The widespread adoption of Natural Based Solutions (NBS) practices in the headwaters of these catchments is expected to add additional attenuation primarily through increasing the amount of new storage available to accommodate flood flows. The actual type of NBS features used to add storage could include swales, ditches, and small ponds (acting as sediment traps). Here, recent data collected from monitored features (from the Demonstration Test Catchments project in the Newby Beck catchment (Eden) in northwest England) were used to provide first estimates of the percentages of the suspended sediment (SS) and total phosphorus (TP) loads that could be trapped by additional features. The Catchment Runoff Attenuation Flux Tool (CRAFT) was then used to model this catchment (Newby Beck) to investigate whether adding additional attenuation, along with the ability to trap and retain SS (and attached P), will have any effect on the flood peak and associated peak concentrations of SS and TP. The modelling tested the hypothesis that increasing the amount of new storage (thus adding attenuation capacity) in the catchment will have a beneficial effect. The model results implied that a small decrease of the order of 5–10% in the peak concentrations of SS and TP was observable after adding 2000 m3 to 8000 m3 of additional storage to the catchment.

Data-driven analysis of nutrient inputs and transfers through nested catchments

A data-driven screening methodology is developed for estimating nutrient input and retention-delivery in catchments with measured water discharges and nutrient concentrations along the river network. The methodology is applied to the Sava River Catchment (SRC), a major transboundary catchment in southeast Europe, with seven monitoring stations along the main river, defining seven nested catchments and seven incremental subcatchments that are analysed and compared in this study. For the relatively large nested catchments (>40,000km²), characteristic regional values emerge for nutrient input per unit area of around 30T/yr/km² for dissolved inorganic nitrogen (DIN) and 2T/yr/km² for total phosphorus (TP). For the smaller nested catchments and incremental subcatchments, corresponding values fluctuate and indicate hotspot areas with total nutrient inputs of 158T/yr/km² for DIN and 13T/yr/km² for TP. The delivered fraction of total nutrient input mass (termed delivery factor) and associated nutrient loads per area are scale-dependent, exhibiting power-law decay with increasing catchment area, with exponents of around 0.2-0.3 for DIN and 0.3-0.5 for TP. For the largest of the nested catchments in the SRC, the delivery factor is around 0.08 for DIN and 0.03 for TP. Overall, the nutrient data for nested catchments within the SRC show consistency with previously reported data for multiple nested catchments within the Baltic Sea Drainage Basin, identifying close nutrient relationships to driving hydro-climatic conditions (runoff for nutrient loads) and socio-economic conditions (population density and farmland share for nutrient concentrations).

Integrative analysis of the Lake Simcoe watershed (Ontario, Canada) as a socio-ecological system

Striving for long-term sustainability in catchments dominated by human activities requires development of interdisciplinary research methods to account for the interplay between environmental concerns and socio-economic pressures. In this study, we present an integrative analysis of the Lake Simcoe watershed, Ontario, Canada, as viewed from the perspective of a socio-ecological system. Key features of our analysis are (i) the equally weighted consideration of environmental attributes with socioeconomic priorities and (ii) the identification of the minimal number of key socio-hydrological variables that should be included in a parsimonious watershed management framework, aiming to establish linkages between urbanization trends and nutrient export. Drawing parallels with the concept of Hydrological Response Units, we used Self-Organizing Mapping to delineate spatial organizations with similar socio-economic and environmental attributes, also referred to as Socio-Environmental Management Units (SEMUs). Our analysis provides evidence of two SEMUs with contrasting features, the "undisturbed" and "anthropogenically-influenced", within the Lake Simcoe watershed. The "undisturbed" cluster occupies approximately half of the Lake Simcoe catchment (45%) and is characterized by low landscape diversity and low average population density <0.4 humans ha^-1. By contrast, the socio-environmental functional properties of the "anthropogenically-influenced" cluster highlight the likelihood of a stability loss in the long-run, as inferred from the distinct signature of urbanization activities on the tributary nutrient export, and the loss of subwatershed sensitivity to natural mechanisms that may ameliorate the degradation patterns. Our study also examines how the SEMU concept can augment the contemporary integrated watershed management practices and provides directions in order to promote environmental programs for lake conservation and to increase public awareness and engagement in stewardship initiatives.

Detecting and analyzing soil phosphorus loss associated with critical source areas using a remote sensing approach

The detection of critical source areas (CSAs) is a key step in managing soil phosphorus (P) loss and preventing the long-term eutrophication of water bodies at regional scale. Most related studies, however, focus on a local scale, which prevents a clear understanding of the spatial distribution of CSAs for soil P loss at regional scale. Moreover, the continual, long-term variation in CSAs was scarcely reported. It is impossible to identify the factors driving the variation in CSAs, or to collect land surface information essential for CSAs detection, by merely using the conventional methodologies at regional scale. This study proposes a new regional-scale approach, based on three satellite sensors (ASTER, TM/ETM and MODIS), that were implemented successfully to detect CSAs at regional scale over 15years (2000-2014). The approach incorporated five factors (precipitation, slope, soil erosion, land use, soil total phosphorus) that drive soil P loss from CSAs. Results show that the average area of critical phosphorus source areas (CPSAs) was 15,056km² over the 15-year period, and it occupied 13.8% of the total area, with a range varying from 1.2% to 23.0%, in a representative, intensive agricultural area of China. In contrast to previous studies, we found that the locations of CSAs with P loss are spatially variable, and are more dispersed in their distribution over the long term. We also found that precipitation acts as a key driving factor in the variation of CSAs at regional scale. The regional-scale method can provide scientific guidance for managing soil phosphorus loss and preventing the long-term eutrophication of water bodies at regional scale, and shows great potential for exploring factors that drive the variation in CSAs at global scale.

Supporting phosphorus management in Austria: Potential, priorities and limitations

Protecting water bodies from eutrophication, ensuring long-term food security and shifting to a circular economy represent compelling objectives to phosphorus management strategies. This study determines how and to which extent the management of phosphorus in Austria can be optimized. A detailed national model, obtained for the year 2013 through Material Flow Analysis, represents the reference situation. Applicability and limitations are discussed for a range of actions aimed at reducing consumption, increasing recycling, and lowering emissions. The potential contribution of each field of action is quantified and compared using three indicators: Import dependency, Consumption of fossil-P fertilizers and Emissions to water bodies. Further, the uncertainty of this assessment is characterized and priorities for the upgrade of data collection are identified. Moreover, all the potential gains discussed in the article are applied to the reference situation to generate an ideal target model. The results show that in Austria a large scope for phosphorus stewardship exists. Strategies based exclusively either on recycling or on the decline of P consumption hold a similar potential to reduce import dependency by 50% each. An enhanced P recycling from meat and bone meal, sewage sludge and compost could replace the current use of fossil-P fertilizers by 70%. The target model, i.e. the maximum that could be achieved taking into account trade-offs between different actions, is characterized by an extremely low import dependency of 0.23kgPcap(-1)y(-1) (2.2kgPcap(-1)y(-1) in 2013), by a 28% decline of emissions to water bodies and by null consumption of fossil-P fertilizers. This case study shows the added value of using Material Flow Analysis as a basis to design sound management strategies. The systemic approach inherent to it allows performing a proper comparative assessment of different actions, identifying priorities, and visualizing a target model.

Integrated cost-effectiveness analysis of agri-environmental measures for water quality

This paper presents an application of integrated methodological approach for identifying cost-effective combinations of agri-environmental measures to achieve water quality targets. The methodological approach involves linking hydro-chemical modelling with economic costs of mitigation measures. The utility of the approach was explored for the River Dee catchment in North East Scotland, examining the cost-effectiveness of mitigation measures for nitrogen (N) and phosphorus (P) pollutants. In-stream nitrate concentration was modelled using the STREAM-N and phosphorus using INCA-P model. Both models were first run for baseline conditions and then their effectiveness for changes in land management was simulated. Costs were based on farm income foregone, capital and operational expenditures. The costs and effects data were integrated using 'Risk Solver Platform' optimization in excel to produce the most cost-effective combination of measures by which target nutrient reductions could be attained at a minimum economic cost. The analysis identified different combination of measures as most cost-effective for the two pollutants. An important aspect of this paper is integration of model-based effectiveness estimates with economic cost of measures for cost-effectiveness analysis of land and water management options. The methodological approach developed is not limited to the two pollutants and the selected agri-environmental measures considered in the paper; the approach can be adapted to the cost-effectiveness analysis of any catchment-scale environmental management options.

Dominant mechanisms for the delivery of fine sediment and phosphorus to fluvial networks draining grassland dominated headwater catchments

Recent advances in monitoring technology have enabled high frequency, in-situ measurements of total phosphorus and total reactive phosphorus to be undertaken with high precision, whilst turbidity can provide an excellent surrogate for suspended sediment. Despite these measurements being fundamental to understanding the mechanisms and flow paths that deliver these constituents to river networks, there is a paucity of such data for headwater agricultural catchments. The aim of this paper is to deduce the dominant mechanisms for the delivery of fine sediment and phosphorus to an upland river network in the UK through characterisation of the temporal variability of hydrological fluxes, and associated soluble and particulate concentrations for the period spanning March 2012-February 2013. An assessment of the factors producing constituent hysteresis is undertaken following factor analysis (FA) on a suite of measured environmental variables representing the fluvial and wider catchment conditions prior to, and during catchment-wide hydrological events. Analysis indicates that suspended sediment is delivered to the fluvial system predominantly via rapidly responding pathways driven by event hydrology. However, evidence of complex, figure-of-eight hysteresis is observed following periods of hydrological quiescence, highlighting the importance of preparatory processes. Sediment delivery via a slow moving, probably sub-surface pathway does occur, albeit infrequently and during low magnitude events at the catchment outlet. Phosphorus is revealed to have a distinct hysteretic response to that of suspended sediment, with sub-surface pathways dominating. However, high magnitude events were observed to exhibit threshold-like behaviour, whereby activation and connection of usually disconnected depositional zones to the fluvial networks results in the movement of vast phosphorus fluxes. Multiple pathways are observed for particulate and soluble constituents, highlighting the challenges faced in mitigating the delivery of contaminant fluxes to headwater river systems.

Soil phosphorus landscape models for precision soil conservation

Phosphorus (P) enrichment in soils has been documented in the Santa Fe River watershed (SFRW, 3585 km) in north-central Florida. Yet the environmental factors that control P distribution in soils across the landscape, with potential contribution to water quality impairment, are not well understood. The main goal of this study was to develop soil-landscape P models to support a "precision soil conservation" approach combining fine-scale (i.e., site-specific) and coarse-scale (i.e., watershed-extent) assessment of soil P. The specific objectives were to: (i) identify those environmental properties that impart the most control on the spatial distribution of soil Mehlich-1 extracted P (MP) in the SFRW; (ii) model the spatial patterns of soil MP using geostatistical methods; and (iii) assess model quality using independent validation samples. Soil MP data at 137 sites were fused with spatially explicit environmental covariates to develop soil MP prediction models using univariate (lognormal kriging, LNK) and multivariate methods (regression kriging, RK, and cokriging, CK). Incorporation of exhaustive environmental data into multivariate models (RK and CK) improved the prediction of soil MP in the SFRW compared with the univariate model (LNK), which relies solely on soil measurements. Among all tested environmental covariates, land use and vegetation related properties (topsoil) and geologic data (subsoil) showed the largest predictive power to build inferential models for soil MP. Findings from this study contribute to a better understanding of spatially explicit interactions between soil P and other environmental variables, facilitating improved land resource management while minimizing adverse risks to the environment.

Phosphorus losses from monitored fields with conservation practices in the Lake Erie Basin, USA

Conservation practices are implemented on farm fields in the USA through Farm Bill programs; however, there is a need for greater verification that these practices provide environmental benefits (e.g., water quality). This study was conducted to assess the impact of Farm Bill eligible conservation practices on soluble P (SP) and total P (TP) losses from four fields that were monitored between 2004 and 2013. No-tillage doubled SP loading compared to rotational tillage (e.g., tilled only before planting corn); however, no-tillage decreased TP loading by 69 % compared to rotational tillage. Similarly, grassed waterways were shown to increase SP loads, but not TP loads. A corn-soybean-wheat-oat rotation reduced SP loads by 85 % and TP loads by 83 % compared to the standard corn-soybean rotation in the region. We can potentially attain TP water quality goals using these Farm Bill practices; however, additional strategies must be employed to meet these goals for SP.

Targeting priority management areas for multiple pollutants from non-point sources

The control of multiple pollutants from non-point sources is very difficult because their loss potentials are not consistent on the same spatial distributions. In this research, an innovative approach was established for multiple-pollutant priority management areas (MP-PMAs). In the new framework, the MP-PMA approach focused on the sensitive areas that contributed a variety of pollutants instead of a specific targeted pollutant by integrating a watershed model and a Pareto-based multi-criteria evaluation approach. Based on the results, multiple levels of MP-PMAs were established with respect to the corresponding requirements of clean water statutes. Compared to traditional separate strategies, the MP-PMA approach would lead to more cost-effective watershed management because those moderate-level PMAs for specific targeted pollutant might be the high-level MP-PMAs. With respect to spatial distribution, the MP-PMA approach provided more accurate target results for the high-level PMAs, especially among the headwater areas. From a scientific view, the MP-PMA approach provides an integrated suggestion for the placement and removal potentials of best management practices at the watershed scale.

Assessment of nitrogen and phosphorus loads and causal factors from different land use and soil types in the Three Gorges Reservoir Area

It is important to identify nitrogen (N) and phosphorus (P) from non-point sources (NPS) to protect watershed water quality. However, few studies have been conducted in a large basin to determine the sources and causal factors of N and P from multiple land use and soil types. In this study, the Soil and Water Assessment Tool (SWAT) was applied in support of the Small-scale Watershed Extended Method (SWEM) in the Three Gorges Reservoir Region (TGRR), China. The first-order error analysis (FOEA) was used to identify the key sub-processes that affected the occurrence of NPS pollution. Based on this study, runoff from the dry land and paddy fields was enriched with nutrients, while runoff from purplish soils, paddy soils and yellow earths made up the largest amount of nutrient loads. The results showed that the contributions of individual parameters to the total uncertainty were different among land use and soil types. This study indicated that management efforts for dry land must explicitly account for conservation practices and proper land covers, while conscientious fertilization practices would result in a greater decrease of NPS yields for paddy fields. Grazing practices should be considered for the purpose of N management for yellow earths, while plant density is important to control soil erosion for purple soils.