Spatial scale effects of landscape metrics on stream water quality and their seasonal changes

Long-term water quality simulation and driving factors identification within the watershed scale using machine learning

Understanding long-term trends and analyzing their driving factors are essential to effectively enhance water quality in watersheds. In China, although the overall quality of surface water continues to improve, significant issues remain in certain regions. The Liao River Basin, a critical industrial hub and key agricultural grain base in northeast China, continues to face unstable water quality conditions, despite over 20 years of management efforts. This study compared several data-driven models (random forest (RF), support vector machine regression (SVR), K-nearest neighbors (KNN), stacking, long short-term memory (LSTM), convolutional-long short-term memory (CNN-LSTM)), to accurately fill the water quality data gaps (i.e., total nitrogen (TN), ammonia nitrogen (NH3-N), total phosphorus (TP), chemical oxygen demand (CODCr), permanganate index (CODMn), electroconductibility (E)) from 1980 to 2022 in Liao River Basin. In addition, the SHapley Additive exPlanations (SHAP) model was employed to quantitatively assess the driving factors of water quality. The results showed that the RF model exhibited robust predictive capabilities. TN showed a steady increase of approximately 20 % from 1980 to 2022, while the other parameters were effectively controlled. Anthropogenic activities, especially in agriculture and urban areas, were found to significantly contribute to water quality deterioration. Additionally, climatic factors such as extreme rainfall, annual average precipitation, and extreme temperatures-along with geographical factors like soil properties and slope, were found to play crucial roles in influencing water quality.

Impact of forest fragmentation on river water quality: an example from a typical subtropical hilly basin

Background

Forest fragmentation, driven by natural and human activities, is increasing. However, the impact of forest fragmentation on river water quality remains ambiguous.

Methods

In this study, water quality data were collected from 15 monitoring sites in the upper Ganjiang River basin in winter and summer, and the forest landscape fragmentation metrics in the sub-basin was calculated to assess its seasonal impact on river water quality.

Results

The results indicated that water quality in the area is generally satisfactory, with total nitrogen (TN) as the main pollutant. Redundancy analysis (RDA) showed that the explanation rate of the six forest landscape fragmentation metrics to the water quality change in summer was 41.21%, and in winter, their explanation rate of water quality change increased by 14.26%. Among them, the effective mesh size (MESH) was negatively correlated with most river water quality indicators, with a contribution rate of 20.9%. While the interspersion and juxtaposition index (IJI) was positively correlated with most water quality indicators in winter, with a contribution rate of 44.9%. It is worth noting that the thresholds for IJI and MESH of forest were the same in winter and summer, 28.1% and 7.89e+0.5ha, respectively, when the probability of an abrupt change in TN concentration reached 100%. This is implied that when the adjacency of forest patches is less than 28.1% and the connectivity of forest patches is more than 7.89e+0.5ha, it may contribute to the reduction of TN concentration in rivers. These findings provide valuable insights into how varying degrees of forest fragmentation can lead to deterioration in river water quality, and allow for further planning of forest structure based on forest fragmentation thresholds to improve regional water quality.

Significant differences in optimal riparian buffer zone on water quality between different segments within the same river

Determining the optimal riparian buffer zone based on the relationship between landscape metrics and water quality is a widely used method for water quality management. However, failing to account for the differences between various segments of the same river can lead to inaccurate identification of riparian buffer zones, thereby affecting the effectiveness of water quality improvement. Here, based on water quality monitoring data from January 2023 to December 2023 in different segments (JuRong segment and XieXi segment) of a typical traditional-intensive agricultural watershed (Qinhuai River watershed), we identified the differences in the optimal riparian buffer zone on water quality between different segments within the same river through redundancy analysis (RDA) and variance partitioning analysis (VPA). Subsequently, utilizing the nonparametric change-point analysis (nCPA), we further identified the critical landscape thresholds causing abrupt changes in water quality within the optimal riparian buffer zone. Results showed that in the JuRong segment, the optimal width for riparian buffer zones was 100 m, with landscape metrics explaining 96.7% of the water quality variation. In the XieXi segment, the optimal riparian buffer zone width was 600 m, with landscape metrics explaining 82.6% of the water quality variation. Interspersion and Juxtaposition index of water (IJI_Water) and Interspersion and Juxtaposition index of Garden (IJI_Garden) were found to be the most influential landscape metrics on water quality in the JuRong and XieXi segments, respectively. Furthermore, the landscape thresholds of IJI_Garden and IJI_Water resulting in abrupt changes in water quality were 68.50 and 39.88 in the JuRong segment, and 76.07 and 56.39 in the XieXi segment, respectively. This study highlights the importance of considering the varying effects in optimal riparian buffer zones and developing distinctive water quality management strategies between different segments within the same river.

A novel water quality risk assessment framework for reservoir water bodies coupling key parameter selection and dynamic warning threshold determination

Water quality early warning is crucial for protecting ecological security and controlling pollution in lakes and reservoirs. However, the traditional warning level may not provide accurate data for a specific area. Therefore, it is necessary to design an adaptive early warning threshold and identification system that conforms to the actual operating environment. This study monitored nine water quality parameters-water temperature (WT), pH, dissolved oxygen (DO), permanganate index (CODMn), chemical oxygen demand (COD), five-day biochemical oxygen demand (BOD5), total nitrogen (TN), total phosphorus (TP), and ammonia nitrogen (NH3-N)-monthly from 11 sampling sites in the Danjiangkou Reservoir, i.e., the largest artificial lake in Asia, from 2017 to 2022. The reservoir was divided into three sub-areas by cluster analysis: Danku, Hanku, and Water intake. The Water Quality Index (WQI) was used for comprehensive spatiotemporal water quality evaluation, and a minimum WQI (WQImin) model was developed using multiple linear regression. Finally, a water quality risk early-warning model was proposed based on frequency analysis, categorizing water quality into six levels. The findings reveal that the water quality in each area maintains at "good" or "excellent" levels during the study period. The average WQI values, from lowest to highest, are Hanku (75.44), Danku (78.78), and Water intake (79.07). This result shows that the water quality of Danjiangkou Reservoir has been maintained at a good level due to the pollution control and management of Chinese government after the operation of the Middle Route of the South-to-North Water Diversion Project of China. The WQImin models for each area have different key parameters: WT, DO, TN, TP, and COD are common in all areas, whereas NH3-N is included in both Hanku and Danku models. BOD5 and pH were unique to the Danku and Water intake models, respectively. TN and TP are identified as the key parameters affecting water quality safety in Danjiangkou Reservoir. The risk thresholds for TN and TP in Hanku are significantly higher than those in Danku and Water intake, indicating that the water quality in Hanku is the worst. These thresholds are dynamically revised through the early warning model as new data became available. The proposed risk assessment framework provides a robust tool for water quality risk early warning and offers a scientific and reliable reference for administrative departments to implement effective water environment risk prevention and management strategies.

Response of river water quality to landscape features in a subtropical hilly region

Landscape features have a profound impact on river water quality. However, its impact in subtropical hilly region is unclear. Here, water quality data from 15 catchments were obtained based on a typical subtropical hilly area, the upper Ganjiang River basin. The landscape features in the catchment and buffer zone were calculated, and its effects on river water quality were investigated using redundancy analysis (RDA) and multiple linear regression (MLR) model. Catchment landscape features were found to better explain overall water quality changes compared to buffer zone, and landscape features were found to explain water quality changes more in winter than in summer. Moreover, within the buffer zone, the percentage of grassland had the greatest impact on winter water quality (72.8%), while at the catchment scale, the aggregation index (AI) of grassland contributed the most to changes in winter water quality (31.6%). Nonparametric change-point analysis (nCPA) was used to identify thresholds of landscape features that lead to abrupt changes in water quality. It was found that river water quality can be improved when the percentage of grassland > 0.193%, the largest patch index (LPI) of forest > 7.48% at the buffer zone or the percentage of impervious surfaces < 2.92%, the AI of forest > 98.6% at the catchment scale. This study demonstrated the pivotal role in enhancing river water quality by implementing informed and effective landscape planning for conservation implementation.

Spatiotemporal variations in land use impacts on river water quality in a mountain-to-plain transitional basin in arid region of northern China

Land use, as an integrated representation of natural conditions and human activities, significantly impacts river water quality. Understanding the spatial and temporal variability of these influences offers valuable insights for improving water quality through the implementation of best management practices. This study examined the impact of land use on river water quality in the Dahei River Basin, a typical mountain-to-plain basin located in the arid region of northern China, which is also the last first-order tributary of Upper Yellow River. Hierarchical clustering analysis was employed to analyze the spatial distribution characteristics of river water quality and redundancy analysis was used to explore the impacts of land use on water quality in upstream buffer zones with radii from 500 m to 14,000 m. The results indicate that river water quality conditions in the mountainous region are much better than in the plain region. In both the dry and wet seasons, land use significantly affects water quality variation, particularly at the 8000 m buffer zone, although the mechanisms differ. In the wet season, the non-point source pollution from storm runoff erosion dominates the positive correlations between water pollution levels and the areas of cropland and urban regions, while for the dry season such positive correlations may come from elevated soil electrolyte levels due to groundwater irrigation and point source pollution from urban activities. For land use types that show a negative correlation with water pollutant levels, the stronger correlation observed in grasslands compared to forests region may be attributed to grasslands' better adaptation to arid conditions. The findings from this study enhance our understanding of the spatiotemporal variations in land use impacts on river water quality and can provide guidance for land use planning at the basin scale in arid regions.

Advancing methodologies for assessing the impact of land use changes on water quality: a comprehensive review and recommendations

With increasing scholarly focus on the ramifications of land use changes on water quality, although substantial research has been undertaken, the findings demonstrate pronounced spatial variability and the heterogeneity of research methodologies. To address this critical gap, this review offers a rigorous evaluation of the strengths and limitations of current research methodologies, providing targeted recommendations for refinement. It systematically assesses the existing body of literature concerning the influence of land use changes on water quality, with particular emphasis on the spatial heterogeneity of research results and the uniformity of employed methodologies. Despite variations in geographical contexts and research subjects, the methodological paradigms remain largely consistent, typically encompassing the acquisition and analysis of water quality and land use data, the delineation of buffer zones, and the application of correlation and regression analyses. However, these approaches encounter limitations in addressing regional disparities, nonlinear interactions, and real-time monitoring complexities. The review advocates for methodological advancements, such as the integration of automated monitoring systems and IoT technologies, alongside the fusion of deep learning algorithms with remote sensing techniques, to enhance both the precision and efficiency of data collection. Furthermore, it recommends the standardization of buffer zone delineation, the reinforcement of foundational water quality assessments, and the utilization of catchment-scale analyses to more accurately capture the influence of land use changes on water quality. Future inquiries should prioritize the development of interdisciplinary ecological models to elucidate the interaction and feedback mechanisms between land use, water quality, and climate change.

Effects of forest structure from graph theory connectivity indicators on river water quality in the Caspian Sea Basin

The structure of a landscape is important as it affects the sources of food for humans and other animals and quality and amount of organic matter in water bodies. The spatial arrangement of landscape features, especially the variations in patch size and the physical space between them, influences the quantity and quality of materials that flow into water bodies and watercourses. In this study located in the Greater Caspian Sea Basin, we investigated how the connectivity of forest patches and the resulting landscape corridors may influence the quality of water. From geographic databases and graph theory we developed 10 landscape metrics and 11 water quality indicators and used these to estimate whether continuous, unbroken landscapes might be influential in improving water quality. We employed Pearson's and Spearman's correlation coefficients to examine the potential relationships between forest-patch connectivity and water quality metrics. Additionally, we employed stepwise, non-linear regression to develop allometry-based power, exponential, and logarithmic models for estimating water quality metrics. The analysis revealed a significant negative correlation between certain forest-patch connectivity indicators (such as landscape coincidence probability and the integral index of connectivity) and certain water quality metrics, suggesting that increased forest connectivity may be associated with improvements in water quality. The modeling results indicated that logarithmic, power and exponential models (or non-simple regression models) with acceptable AIC coefficients (the model selection criteria explained in the materials and methods section) were the most suitable for describing these relationships. As the connectivity of forest patches decreases and subsequently the fragmentation of forest patches increases, water pollution parameters increase while water quality decreases. The models developed included high R2 values for water quality metrics such as CO3 (0.82), water discharge (0.73), calcium (0.77), and total dissolved solids (0.70).

New modeling framework for describing the effects of landscape pattern changes on nutrient pollution transport

Landscape pattern plays a crucial role in regulating hydrological and pollutant migration processes. However, there is a lack of quantitative tools to describe the nutrient pollution transport process under the influence of different landscape patterns. To fill this gap, this study presents a new modeling framework, namely the Landscape Pattern-Source Flow Sink model (LP-SFS). The model consists of three modules: nutrient pollution emission, land transport, and river transport. Each module is implemented using a separate calculation program. It characterizes the transport path of pollutants on landscape units conceptually and operationally and focuses on quantifying the blocking effect of grid-scale landscape units on nutrients. The framework takes the Luanhe River Basin in the North China Plain as an exemplary case. The simulation results of the new framework indicated that in regions predominantly occupied by forest and urban (FU), the intensity of terrestrial pollutant migration attained the highest level. In the slope zone ranging from 25 to 35°, owing to the relatively strong responsiveness of water flow and gravity to the slope, soil erosion is inclined to occur, thereby causing the intensity of terrestrial nutrient pollution transport in this slope zone to reach the maximum value of 4.55 kg/km2. Furthermore, in the elevation zone <700 m, urban and cultivated land was concentratedly distributed, which leads to more pollutants entering surface water bodies and increases the intensity of terrestrial migration of pollutants. The complex boundary shape of forestland and grassland in the watershed weakened the transport capacity of nutrients, resulting in pollutants remaining in the soil and being difficult to be transported to surface water bodies. This new method is applicable to large-scale watersheds with strong spatial heterogeneity and severe landscape fragmentation and can provide technical support for nutrient pollution control and optimization of land resource allocation in large-scale watersheds.

Tracking the optimal watershed landscape pattern for driving pollutant transport: Insights from the integration of mechanistic models and data-driven approaches

Identifying landscape patterns conducive to pollutant transport control is of vitally importance for water quality protection. However, it remains unclear which landscape patterns can weaken the transport capacity of pollutants entering water bodies. To fill this gap, this study proposes a new framework. This framework quantifies the contribution of landscape patterns to pollutant migration; it also identifies the optimal landscape patterns capable of reducing pollutants entering rivers. Furthermore, it analyzes the impact pathways of landscape patterns on pollutant migration by integrating mechanism models, machine learning techniques, and structural equation models (SEM). The results showed that on cultivated land and urban land, when the slope reached 35%, the terrestrial transport intensity of NH₃-N peaked at 34 kg/km2 and 45 kg/km2 respectively, with more pollutants entering the receiving water bodies. Meanwhile, in the forest with a DEM of 900 m, the terrestrial transport intensity of NH₃-N was the highest (50 kg/km2). The complexity of the landscape boundary shape in areas dominated by cultivated land and forest was verified to have a significant impact on the terrestrial migration intensity of NH₃-N, with a contribution rate of over 65%. From the comparison results of multiple land use combinations, it can be seen that the combination of woodland and grassland indirectly weakens the transport capacity of pollutants entering water bodies by directly influencing the connectivity among landscape units. In particular, when the proportion of woodland and grassland reaches 75%, it has a positive effect on improving river pollution and is the optimal landscape combination pattern for reducing the pollution load of the river. The outcomes can be used to develop more efficacious optimization and regulation tactics for landscape patterns and offer a decision - making foundation for the control of pollutant transport in large basins.

Integrating river transport processes and seasonal dynamics to assess watershed nitrogen export risk

Excessive nitrogen exported to water bodies affects the balance of ecosystem and poses a threat to human health. Although the concept of water purification service helps quantify nitrogen export, the impact of river transport remains unclear. This study focused on nitrogen as a pollutant by utilizing the Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) model to assess nitrogen export in the Dongting Lake Basin, taking into account both the processes of sub-basin nitrogen export and river transport. Additionally, the monthly variations within the year were further explored, together with the identification of the priority area of returning cropland to forest land. The results showed that in 2021, the total nitrogen load outside the Dongting Lake was 11.34 × 108 kg·year-1, with the sub-basins retaining a total of 9.02 × 108 kg·year-1, accounting for 79.57 % of the total nitrogen load. Notably, the total nitrogen retention by the river made up 16.87 % of the total nitrogen retention, up to 1.83 × 108 kg·year-1. In view of monthly variation, water purification service based on the entire process was higher in winter and late autumn, and lower in summer and early autumn. The priority areas for ecological project were mainly distributed around the Dongting Lake, achieving an improvement of 58.15 % in water purification service with approximately 7.02 % of the total returnable cropland. This study proposed a new approach of water purification service assessment through integrating river transport processes and seasonal dynamics, aiming to assess watershed nitrogen export risk more accurately.

Enhancing river water quality in different seasons through management of landscape patterns at various spatial scales

Landscape patterns have a great effect on river water quality. However, the strategies for enhancing water quality through landscape pattern management remain unclear. In this study, we aimed to provide effective guidance for water quality management by quantifying the key spatial scales and landscape metrics that influence the seasonal variations in water quality and establishing threshold relationships between these metrics and abrupt variations in water quality in the Chaohu Lake basin, China. Results discovered that water quality was poorer in summer and better in spring, with degraded water conditions primarily concentrated in the middle and lower reaches of the watershed. The 100 m riparian zone buffer scale landscape pattern was identified as the key scale affecting water quality in the summer, which accounted for 51.3% of the overall water quality variation. Furthermore, abrupt threshold analysis indicated that summer water quality could be effectively improved by maintaining the proportion and largest patch index of construction land within the 100 m riparian buffer below 22.0%. At the sub-basin scale, landscape pattern-based water quality management was most effective in spring, explaining 43.6% of the variation in water quality. Setting the largest patch index of construction land at the sub-basin scale below 43.0% and increasing the proportion of forest cover above 36.0% can also alleviate water pollution issues. These findings emphasize the importance of incorporating landscape patterns across scales into environment management decisions, providing a scientific basis for effective watershed water quality management.

Increasing human activity shifts the key spatial scale of landscape patterns on water quality from sub-basins to riparian zones

The relationship between landscape patterns and water quality has been extensively studied, yet the understanding of how human activity modulates the spatial scale effects of landscape patterns on water quality remains limited. Here, we investigated the water quality and landscape patterns of three rivers in the Poyang Lake Basin, China, subjected to different intensities of human activity, and analyzed the extent to which water quality parameters were influenced by human activity to unravel the spatial scale effects and identify critical landscape metrics that significantly influence water quality. The results showed that the influence of riparian zone landscape patterns on water quality progressively exceeded that of sub-basin landscape patterns as the intensity of human activity increased. For water quality parameters that were minimally affected by human activity, the influence of sub-basin landscape patterns slightly exceeded that of riparian zone landscape patterns at different intensities of human activity (differences were 0.63 %, 4.25 % and 7.65 %, respectively). Conversely, for water quality parameters significantly affected by human activity, the landscape patterns of the riparian zone had a significantly greater influence than the sub-basin landscape patterns (differences were 5.90 %, 13.00 % and 17.86 %, respectively). Furthermore, the discrepancy between the influence of riparian zone and sub-basin landscape patterns on water quality increased with increasing intensity of human activity, while the overall influence of landscape patterns on water quality showed a decreasing trend (decreasing from 60.35 % to 39.10 %). In addition, the proportions of construction land, farmland, and forestland, and the fragmentation of grassland, were identified as critical landscape metrics that significantly influenced water quality at different intensities of human activity. This study revealed that different intensities of human activity were key factors influencing the spatial scale effects of landscape patterns on water quality.

The phosphorus export coefficients variability of specific land-use and the threshold response relationship with watershed characteristics in a subtropical hilly region

The magnitude of land-use phosphorus (P) export in subtropical hilly watersheds is subject to the collective influence of various watershed characteristics. However, the spatiotemporal dynamics of key watershed characteristics and their impacts on P exports remain unclear. Total phosphorus export coefficients (TPECs) can quantify the contribution of different land-use types to P exports. Here, we employed continuous observation data of typical subtropical hilly areas in the Laodao River watershed, China, from 2012 to 2019, as well as an enhanced export coefficient model combined with Bayesian statistical methods, to quantify the spatiotemporal variability and uncertainty of TPECs under various hydrological conditions in different land-use types (forest, cropland, and tea plantation). We then determined the key watershed characteristic factors influencing TPECs and threshold responses of TPECs to these factors. The three land-use types exhibited significant spatiotemporal differences in TPECs, i.e., under different hydrological regimes and within each catchment. Watershed characteristic factors explained 68.5-74.1 % and 33.3-50.4 % of TPEC variations under medium- and low-flow regimes, respectively, but only 1.9-3.8 % under the high-flow regime. Compared to topography and soil factors, landscape patterns had a higher individual impact on TPECs. Threshold response relationships existed between TPECs and key influencing factors under both medium- and low-flow regimes. Moreover, the thresholds corresponding to abrupt TPEC change points were comparatively consistent under medium- and low-flow regimes. These findings have practical applications for rapidly characterising critical P pollution source areas and formulating basin-scale land-use plans for water quality protection.

Using surface runoff to reveal the mechanisms of landscape patterns driving on various forms of nitrogen in non-point source pollution

Non-point source (NPS) pollution directly threatens river water quality, constrains sustainable economic development, and poses hazards to human health. Comprehension of the impact factors on NPS pollution is essential for scientific river water quality management. Despite the landscape pattern being considered to have a significant impact on NPS pollution, the driving mechanism of landscape patterns on NPS pollution remains unclear. Therefore, this study coupled multi-models including the Soil and Water Assessment Tool (SWAT), Random Forest, and Partial Least Squares Structural Equation Modeling (PLS-SEM) to construct the connection between landscape patterns, NPS pollution, and surface runoff. The results suggested that increased runoff during the wet season enhances the link between landscape patterns and NPS pollution, and the explained NPS pollution variation by landscape pattern increased from 59.6 % (dry season) to 84.9 % (wet season). Furthermore, from the impact pathways, we find that the sink landscape pattern can significantly and indirectly influence NPS pollution by regulating surface runoff during the wet season (0.301*). Meanwhile, the sink and source landscape patterns significantly and directly impact NPS pollution during different seasons. Moreover, we further find that the percentage of paddy land use (Pad_PLAND) and grassland patch density (Gra_PD) metrics can significantly predict the dissolved total nitrogen (DTN) and nitrate nitrogen (NO3--N) variation. Thus, controlling the runoff migration process by guiding the rational evolution of watershed landscape patterns is an important development direction for watershed NPS pollution management.

Landscape fragmentation of built-up land significantly impact on water quality in the Yellow River Basin

Urbanization development often leads to significant changes in the extent in area and fragmentation of built-up land landscape pattern (BLLP) in river basins, which greatly impact the processes of rainfall runoff and pollutant migration. Understanding the spatial scale effects and driving mechanisms of BLLP changes on water quality in large river basins is a challenging research topic and an international frontier in the interdisciplinary fields of geography and environment. This study analyzes the spatial variations of BLLP and water quality throughout the Yellow River Basin (YRB) during the rainy seasons from 2019 to 2021 (4 h scale). Utilized the random forest model to quantitatively separates the contributions of rainfall processes to surface runoff and water pollution, revealing the scale effects and non-linear driving mechanisms of BLLP impacts on water environment changes. The results indicate that: 1) The YRB exhibits great spatial heterogeneity in terms of both BLLP and water quality, with places with lower water quality displaying bigger areas and higher degrees of BLLP fragmentation. 2) The patch density and built-up land area (PD.B and CA.B) have a major impact on changes in water quality in the YRB, with notable impacts noted in circular buffer zones with radii of 20 km and 5 km, respectively. 3) PD.B is sensitive to water quality in the YRB, explaining 39.1%-49.5% of the variance under different rainfall conditions, and exhibits a significant non-linear relationship, with an impact threshold of 0.38 (n/100 ha). The study suggests that for large-scale regions like the YRB, the degree of BLLP fragmentation is more likely to lead to degradation of water environmental quality compared to its area. BLLP fragmentation due to higher PD.B and CA.B disrupts the original ecosystem and hydrological connectivity, resulting in poorer retention and filtration of pollutants carried by rainfall runoff, while increasing the export of other pollutants. However, once urbanization surpasses a certain threshold, the BLLP fragmentation can enhance water quality by reducing the impermeable surface connectivity, as they are no longer impacted by expanding areas. To achieve ecologically sustainable development, it is necessary to apply rational landscape management and water resource management policies that consider the dual process of how BLLP fragmentation affects the water environment.

Riverine bacterial communities are more shaped by species sorting in intensive urban and agricultural watersheds

Bacterial communities play a crucial role in maintaining the stability of river ecosystems and driving biogeochemical cycling, exhibiting high sensitivity to environmental change. However, understanding the spatial scale effects and assembly mechanisms of riverine bacterial communities under distinct anthropogenic disturbances remains a challenge. Here, we investigated bacterial communities across three distinct watersheds [i.e., intensive urban (UW), intensive agricultural (AW), and natural (NW)] in both dry and wet seasons. We explored biogeographic patterns of bacterial communities and the influence of landscape patterns at multi-spatial scales and water chemistry on bacterial communities. Results showed that α diversity was significantly lower in UW and AW compared to NW, particularly in the dry season. A gradient of β diversity with NW > UW > AW was observed across both seasons (p < 0.05). Pseudomonadota, Bacteroidota, and Actinobacteriota were the most abundant phyla across all watersheds, with specific taxa enriched in each watershed (i.e., the class Actinobacteria was significant enrichment in UW and AW, and Clostridia in NW). The influence of landscape patterns on bacterial communities was significantly lower in human-disturbed watersheds, particularly in UW, where this influence also varied slightly from near riparian buffers to sub-watershed. Homogeneous selection and drift jointly dominated the bacterial community assembly across all watersheds, with homogeneous selection exhibiting a greater influence in UW and AW. Landscape patterns explained less variance in bacterial communities in UW and AW than in NW, and more variance was explained by water chemistry (particularly in UW). These suggest that the stronger influence of species sorting in UW and AW was driven by more allochthonous inputs of water chemistry (greater environmental stress). These findings provide a theoretical foundation for a deeper understanding of riverine bacterial community structure, spatial scale effects, and ecological management under different anthropogenic activities.

The spatial scale and threshold effects of the relationship between landscape metrics and water quality in the Hutuo River Basin

The impact of landscape patterns on river water quality has been widely studied; however, it remains unclear which spatial scale has the greatest impact on water quality. Here, we analyzed the spatial scale and threshold impacts of the link between landscape metrics and water quality in a large-scale basin using the random forest (RF) model and nonparametric change point analysis (nCPA) method. The concentrations of nitrate nitrogen (NO3--N) and total nitrogen (TN) were comparatively high in winter and relatively low during spring and summer, whereas the total phosphorus (TP) concentrations were comparatively low during winter and summer and relatively high during spring. The R2 values of the RF models at the sub-basin scale were generally higher than those at the riparian zone scale. Moreover, the R2 of water quality modelling at the riparian zone scale demonstrated a declining tendency from a riparian zone 30 m-210 m wide in the majority of seasons. This shows that landscape metrics at the subbasin scale provide a better explanation for the variability in water quality than those at the riparian zone scale in the Hutuo River Basin. The results of the RF model indicated that landscape metrics of landscape configuration were more important in determining water quality during winter, whereas landscape metrics of landscape composition or physiography were more important in determining water quality during summer. Furthermore, several abrupt thresholds were estimated by nCPA; for example, the summertime slope abrupt threshold was 10.79° in the relationship between the slope and NO3--N. This study contributes to the understanding of the debate regarding the scale effects of landscape patterns on water quality, emphasizing the significance of the basin area and offering managers valuable insights into the control of non-point source pollution.

Interactive impact of landscape composition and configuration on river water quality under different spatial and seasonal scales

Currently, the comprehensive effect of the landscape pattern on river water quality has been widely studied. However, the interactive influences of landscape type, namely composition (COM) and configuration (CON) on water quality variations, as well as the specific landscape driving types affecting water quality variations under different spatial and seasonal scales remain unclear. To further improve the effectiveness of landscape planning and water quality protection, this study collected monthly water samples from the Fengyu River Watershed in southwestern China from 2018 to 2021, the Biota-Environment Matching Analysis (Bioenv) was used to identify key metrics representing landscape COM and CON, respectively. Then, the multiple regression (MLR) and redundancy analysis (RDA) were used to explore the relationship between these landscape metrics and water quality. In addition, this study used a variation partitioning analysis (VPA) to quantify the interactive and independent influence of landscape COM and CON on water quality. Results revealed that construction land and the Shannon's diversity index (SHDI) were the key metrics of landscape COM and CON, respectively, for predicting water pollution concentrations. The interactive contribution was particularly sensitive to seasonal changes in riparian buffer areas (27.66 % to 48.73 %), while it remained relatively stable at the sub-watershed scale (38.22 % to 40.51 %). Moreover, landscape CON had a higher independent contribution to variations on water quality across most spatio-temporal scales. Overall, identifying and managing key landscape type and consequential metrics, matching with the spatio-temporal scale, holds promise for enhancing water quality conservation. Furthermore, this study provides valuable insights into the identification and selection of core landscape metrics.

A novel nonlinear direct-mapping approach for multiple time scale driving force analysis of surface water quality variations under intense human interference

Identifying the driving forces of surface water quality variations is crucial for urban environmental management, especially in densely populated regions. Statistic mapping is an approach that allows researchers to directly explore the response of surface water quality to potential drivers. Conventionally, these methods encounter a mixture of issues, including nonlinear relationships and information on multiple time-scale, caused by disparities in the influencing frequencies and degrees of driving factors. In this research, a nonlinear direct-mapping approach was developed to quantitatively analyze the driving force of surface water quality under multiple time scales. This approach separated the fluctuation and trend information from water quality data and then established a direct-mapping relationship, thereby achieving the visible multilayer structure containing both linear and nonlinear information from the time scale. Typical water pollutants including total nitrogen (TN) and total phosphorus (TP) in the Pearl River Delta (PRD), were used to verify the methodology and compare its ability to analyze driving forces with traditional statistic approaches. The results demonstrated that this approach could establish a visual multilayer mapping structure with strong interpretability, which effectively captured the contained nonlinear information, thus improving the fitting degree by 12.43% compared with traditional methods. Moreover, it successfully identified the dominant driving forces of TN and TP in the PRD as human activities related to NO2 and PM and natural factors. Its application in the changing environment demonstrated a potentially increased risk of TP in the PRD under multiple scenarios. Overall, this approach could serve as a reliable reference for pollution early warning in the short term and for industrial structure adjustment planning in the long term.

Impacts of land use on phosphorus and identification of phosphate sources in groundwater and surface water of karst watersheds

The thin soil layer with uneven distribution in karst areas facilitates the migration of phosphorus (P) to groundwater, threatening the safety of water sources seriously. To offer a scientific guidance for water pollution control and land use planning in karst areas, this study examined the relationships between land use and P in groundwater and surface water, and quantified the phosphate sources in Gaoping river basin, a small typical watershed in karst areas. Spatial distribution analysis revealed that the highest mean P concentrations in groundwater and surface water were in farmland and construction-farmland zones, respectively. Land use impact analysis showed that the concentration of P in groundwater was influenced positively by farmland but negatively by forest land. In contrast, the concentration of P in surface water was influenced positively by both farmland and construction land. The mixed end-element and Bayesian-based Stable Isotope Analysis in R (SIAR) model results showed that agricultural fertilizers were the main phosphate source for groundwater in farmland and forest-farmland zones, while urban sewage was the main source in the construction-farmland zone. For surface water, the main phosphate source was agricultural fertilizers in both farmland and construction-farmland zones. This study indicates that controlling P pollution in local water bodies should pay close attention to the management of land use related to human activities, including regulating sewage discharge from construction land and agricultural fertilizer usage.

Response of water quality in major tributaries to the difference of multi-scale landscape indicators in Dongting Lake basin, China

River water quality has been increasingly deteriorated because of the influence of natural process and anthropogenic activities. Quantifying the influence of landscape metrics, namely topography and land use pattern, which encompass land use composition and landscape configuration, across different spatial and seasonal scales that reflect natural process and anthropogenic activities, is highly beneficial for water quality protection. In this study, we focused on investigating the effects of topography, landscape configuration and land use composition on water quality at different spatial scales, including 1-km buffer and sub-watershed, and seasonal scales, including wet and dry season, based on the monthly water quality data in 2016 of Dongting Lake in China. Multivariate statistical analysis of redundancy analysis and partial redundancy analysis was used to quantify the contributions of these factors under different scales. Our results showed that among the three environmental groups, topography made the greatest pure contribution to water quality, accounting for 11.4 to 30.9% of the variation. This was followed by landscape configuration, which accounted for 9.4 to 23.0%, and land use composition, which accounted for 5.9 to 15.7%. More specifically, water body made the greatest contribution to the water quality variation during dry season at both spatial scales, contributing 16.6 to 17.2% of the variation. In contrast, edge density was the primary interpreter of the variability in water quality during wet season at both spatial scales, accounting for 9.9 to 11.1% of the variation. The spatial variability in the influence of landscape metrics on water quality was not markedly distinct. However, these metrics have a minimal impact difference on water quality at the buffer scale and the sub-watershed scale. Moreover, the contribution of landscape configuration varied the most from the buffer to sub-watershed scales, indicating its importance for the spatial scale difference in water quality. The findings of this study offer useful insights into enhancing water quality through improved handling of landscape metrics.

Modelling relationships between land use and water quality using statistical methods: A critical and applied review

Land use/land cover (LULC) can have significant impacts on water quality and the health of aquatic ecosystems. Consequently, understanding and quantifying the nature of these impacts is essential for the development of effective catchment management strategies. This article provides a critical review of the literature in which the use of statistical methods to model the impacts of LULC on water quality is demonstrated. A survey of these publications, which included hundreds of original research and review articles, revealed several common themes and findings. However, there are also several persistent knowledge gaps, areas of methodological uncertainty, and questions of application that require further study and clarification. These relate primarily to appropriate analytical scales, the significance of landscape configuration, the estimation and application of thresholds, as well as the potentially confounding influence of extraneous variables. Moreover, geographical bias in the published literature means that there is a need for further research in ecologically and climatically disparate regions, including in less developed countries of the Global South. The focus of this article is not to provide a technical review of statistical techniques themselves, but to examine important practical and methodological considerations in their application in modelling the impacts of LULC on water quality.

Space-time cube uncovers spatiotemporal patterns of basin ecological quality and their relationship with water eutrophication

Maintaining an optimal eco-environment is important for sustainable regional development. However, existing methods are inadequate for examining both spatial and temporal dimensions. Here, we propose a systematic procedure for spatiotemporal examination of the eco-environment using the space-time cube (STC) model and describe a preliminary investigation of the coupling relationships between basin ecological quality and water eutrophication in upstream of the Han River basin between 2000 and 2020. The STC model considers the temporal dimension as the third dimension in calculations. We first categorized the basin into three sub-watershed types: forest, cultivated land, and artificial surface. Subsequently, the ecological quality and driving factors were assessed and identified using the remote sensing ecological index (RSEI) and Geodetector method, respectively. The findings indicated that the forest basin and artificial surface basin had the highest and lowest ecological quality, respectively. The spatiotemporal cold spots of ecological quality during the past 20 years were mostly located in the vicinity of reservoirs, rivers, and artificial surface areas. Human activity, precipitation, and the percentage of cultivated land were other important driving factors in the artificial surface, forest, and cultivated land sub-watersheds, respectively, in addition to the dominant factors of elevation and temperature. The results also indicated that when the ecological quality degraded to a certain extent, water eutrophication was significantly coupled with the ecological quality of the catchments. The findings of this study are useful for ecological restoration and sustainable river basin development.

Influence of watershed characteristics and human activities on the occurrence of organophosphate esters related to dissolved organic matter in estuarine surface water

Organophosphate esters (OPEs) are widespread in aquatic environments and pose potential threats to ecosystem and human health. Here, we profiled OPEs in surface water samples of heavily urbanized estuaries in eastern China and investigated the influence of watershed characteristics and human activities on the spatial distribution of OPEs related to dissolved organic matter (DOM). The total OPE concentration ranged from 22.3 to 1201 ng/L, with a mean of 162.6 ± 179.8 ng/L. Chlorinated OPEs were the predominant contaminant group, accounting for 27.4-99.6 % of the total OPE concentration. Tris(2-chloroisopropyl) phosphate, tris(1,3-dichloro-2-propyl) phosphate, and tributyl phosphate were the dominant compounds, with mean concentrations of 111.2 ± 176.0 ng/L, 22.6 ± 21.5 ng/L, and 14.8 ± 14.9 ng/L, respectively. Variable OPE levels were observed in various functional areas, with significantly higher concentrations in industrial areas than in other areas. Potential source analysis revealed that sewage treatment plant effluents and industrial activities were the primary OPE sources. The total OPE concentrations were negatively correlated to the mean slope, plan curvature, and elevation, indicating that watershed characteristics play a role in the occurrence of OPEs. Individual OPEs (triisobutyl phosphate, tris(2-butoxyethyl) phosphate, tris(2-chloroethyl) phosphate, and tricresyl phosphate) and Σalkyl-OPEs were positively correlated to the night light index or population density, suggesting a significant contribution of human activity to OPE pollution. The co-occurrence of OPEs and DOM was also observed, and the fluorescence indices of DOM were found to be possible indicators for tracing OPEs. These findings can elucidate the potential OPE dynamics in response to DOM in urbanized estuarine water environments with intensive human activities.

Effects of landscape pattern on water quality at multi-spatial scales in Wuding River Basin, China

Urbanization and agricultural land use have led to water quality deterioration. Studies have been conducted on the relationship between landscape patterns and river water quality; however, the Wuding River Basin (WDRB), which is a complex ecosystem structure, is facing resource problems in river basins. Thus, the multi-scale effects of landscape patterns on river water quality in the WDRB must be quantified. This study explored the spatial and seasonal effects of land use distribution on river water quality. Using the data of 22 samples and land use images from the WDRB for 2022, we quantitatively described the correlation between river water quality and land use at spatial and seasonal scales. Stepwise multiple linear regression (SMLR) and redundancy analyses (RDA) were used to quantitatively screen and compare the relationships between land use structure, landscape patterns, and water quality at different spatial scales. The results showed that the sub-watershed scale is the best spatial scale model that explains the relationship between land use and water quality. With the gradual narrowing of the spatial scale range, cultivated land, grassland, and construction land had strong water quality interpretation abilities. The influence of land use type on water quality parameter variables was more distinct in rainy season than in the dry season. Therefore, in the layout of watershed management, reasonably adjusting the proportion relationship of vegetation and artificial building land in the sub-basin scale and basin scope can realize the effective control of water quality optimization.

Fertilization intensities at the buffer zones of ponds regulate nitrogen and phosphorus pollution in an agricultural watershed

The sudden increase in water nutrients caused by environmental factors have always been a focus of attention for ecologists. Fertilizer inputs with spatio-temporal characteristics are the main contributors to water pollution in agricultural watersheds. However, there are few studies on the thresholds of nitrogen (N) and phosphorus (P) fertilization rates that affect the abrupt deterioration of water quality. This study aims to investigate 28 ponds in Central China in 2019 to reveal the relationships of basal and topdressing fertilization intensities in surrounding agricultural land with pond water N and P concentrations, including total N (TN), nitrate (NO3--N), ammonium (NH4+-N), total P (TP), and dissolved P (DP). Abrupt change analysis was used to determine the thresholds of fertilization intensities causing sharp increases in the pond water N and P concentrations. Generally, the observed pond water N and P concentrations during the high-runoff period were higher than those during the low-runoff period. The TN, NO3--N, TP, DP concentrations showed stronger positive correlations with topdressing intensities, while the NH4+-N concentrations exhibited a higher positive correlation with basal intensities. On the other hand, the NO3--N concentrations had a significant positive correlation with the topdressing N, basal N, and catchment slope interactions. Significant negative correlations were observed between all water quality parameters and pond area. Spatial scale analysis indicated that fertilization practices at the 50 m and 100 m buffer zone scales exhibited greater independent effects on the variations in the N and P concentrations than those at the catchment scale. The thresholds analysis results of fertilization intensities indicated that pond water N concentrations increased sharply when topdressing and basal N intensities exceeded 163 and 115 kg/ha at the 100 and 50 m buffer zone scales, respectively. Similarly, pond water P concentrations rose significantly when topdressing and basal P intensities exceeded 117 and 78 kg/ha at the 50 m buffer zone scale, respectively. These findings suggest that fertilization management should incorporate thresholds and spatio-temporal scales to effectively mitigate pond water pollution.

Response of stream water quality to the vegetation patterns on arid slope: a case study of Huangshui River basin

Vegetation patterns on slopes strongly affect the water cycle processes in a basin, especially the water yield and confluence in arid areas. Quantifying and evaluating the effects of hydrological change on the migration and transformation of pollutants are challenging. Based on 4-year stream water quality data of 13 monitoring sites in the Huangshui River basin, a typical arid watershed of the Chinese Loess Plateau, the redundancy analysis (RDA) and structural equation modeling (SEM) analysis tools were used to quantify its relationship with vegetation patterns. In the study, land use and the enhanced vegetation index (EVI) were used as a metric of vegetation patterns; accordingly, the 13 catchments were divided into three groups via the cluster analysis, including large (over 80%), medium (70 ~ 80%), and small (below 70%) proportion vegetation patterns (LVP, MVP, SVP). The results of the LVP group showed that vegetation patterns negatively affected the contamination of total phosphorus (TP), ammonia nitrogen (NH3-N), permanganate index (CODMn), and biochemical oxygen demand (BOD5) in the stream water, and the contribution rates were - 0.57. While the proportion of urban area positively correlated with stream water quality in the groups of MVP and SVP, the contribution rates were 0.46 and 0.36, respectively. Moreover, the precipitation in the groups of MVP and SVP negatively correlated with pollutants (- 0.24 and - 0.26). Those results revealed the response of stream water quality to vegetation patterns on the slope with the consideration of precipitation, land use, and socio-economic factors for the regional water and land resource allocation. This study has important management implications for vegetation patterns on slope of fragile ecosystems in arid areas.

Evaluation of impact of land use and landscape metrics on surface water quality in the northeastern part along Lake Tanganyika, Africa

As the second deepest lake in Africa, Lake Tanganyika plays an important role in supplying fish protein for the catchment's residents and is irreplaceable in global biodiversity. However, the lake's water environment is threatened by socioeconomic development and rapid population growth along the lake. This study analyzed the spatial scale effects and seasonal dependence of land use types and landscape metrics on water quality in 16 sub-basins along northeastern Lake Tanganyika at different levels of urbanization. The results revealed that land use types had a higher influence on water quality in urban areas than that in rural areas; the explanatory variance in the urban area was 0.78-0.96, while it was 0.21-0.70 in the rural area. The explanatory ability of land use types on water quality was better at the buffer scale than at the sub-watershed scale, and the 500 m buffer scale had the highest explanatory ability in the urban area and rural area both in the rainy season and dry season, and artificial surface and arable land were the main contributing factors. And this phenomenon was more obvious in dry season than in rainy season. We identified that CONTAG was the key landscape metric in urban area and was positively correlated with nutrient variables, indicating that water quality degraded in less fragmented landscapes. The sub-watershed scale had the highest explained ability, while in rural area, the 1500 m buffer scale had the highest explained ability and IJI had the highest explanatory variance, which had a negative effect on water quality. Research on the relationship between land use and water quality would help assess the water quality in the unmonitored watershed as monitoring is expensive and time-consuming in low-income area. This knowledge would provide guideline to watershed managers and policymakers to prioritize the future land use development within Lake Tanganyika basin.

The relationship between mountain wetland health and water quality: A case study of the upper Hanjiang River Basin, China

This study investigates the degradation process of mountain wetlands in the upper Hanjiang River Basin (HRB) over a 30-year span from 1990 to 2020. In particular, the landscape development intensity (LDI) index was employed to conduct a comprehensive assessment of the wetland health. This was subsequently combined with the spatio-temporal changes of water quality in the basin to explore the potential correlations between the health status of mountain wetlands and the associated watershed water quality. The results show that over the past three decades, wetland ecosystems have shrunk by 18% due to conversion into farmland, grass, construction land and forest land. This was significant between 2010 and 2020, as shown by a land use dynamic index of -1.121% during 2010-2020, which was significantly higher than that in the preceding two decades (0.003%, 0.367%) (p < 0.05). LDI values for individual sub-watersheds across different years ranged from 2.39 to 4.93, demonstrating an increasing trend since 2010. This indicates a heightened level of human interference in mountain wetlands. Although the water quality within the basin generally adhered to the Class II surface water quality standard, total nitrogen (TN) (primarily from farming) was a concern. Areas with relatively more human activity were observed to exhibit increased pollution levels, as demonstrated by a positive correlation between LDI and the concentrations of total phosphorus (TP), ammonium nitrogen (NH4+-N), and chemical oxygen demand (COD) in the basin. The LDI of the mountain wetland exhibited a consistent positive correlation with the water quality comprehensive function, both during the flood (r = 0.77-0.81) and non-flood (r = 0.61-0.70) seasons (p < 0.05). This indicates the significant impact of the wetland landscape structure on the water quality within a 1000 m radius on either side of the river. Special attention should be paid to the management and allocation of wetland landscapes within this 1000 m buffer zone. Furthermore, efforts to control upstream pollutant emission should be strengthened.

Effects of environmental factors on the river water quality on the Tibetan Plateau: a case study of the Xoirong River, China

Climate, topography, and landscape patterns affect river water quality through processes that influence non-point source pollution. However, little is known about the response of the water quality of rivers on China's Tibetan Plateau to these environmental factors. Based on the water quality parameters data of the Xoirong River on the Tibetan Plateau in western China, the redundancy analysis and variation partitioning analysis were adopted to determine the main influencing factors affecting river water quality and their spatial scale effects. The major water pollutants were further analyzed using the partial least square-structural equation modeling (PLS-SEM). Another mountainous river with a similar latitude, the same stream order, and low anthropogenic disturbance in central China, the Jinshui River, was also selected for comparative discussion. The results indicated that the overall river water quality on the Tibetan Plateau was superior to that of the Jinshui River. At the catchment scale, the cumulative explanatory powers of the influencing factors of both rivers were greatest. Landscape composition and configuration were the determinant factors for the overall water quality of the two rivers, while the river on the Tibetan Plateau was also significantly affected by climatic and topographical factors. Regarding the main water quality issue, i.e., total nitrogen, agricultural production activities might be the main cause of the river on the Tibetan Plateau. This study unveiled that the river water quality on the Tibetan Plateau is sensitive to climate and topography through comparative studies.

Explainable AI approach with original vegetation data classifies spatio-temporal nitrogen in flows from ungauged catchments to the Great Barrier Reef

Transfer of processed data and parameters to ungauged catchments from the most similar gauged counterpart is a common technique in water quality modelling. But catchment similarities for Dissolved Inorganic Nitrogen (DIN) are ill posed, which affects the predictive capability of models reliant on such methods for simulating DIN. Spatial data proxies to classify catchments for most similar DIN responses are a demonstrated solution, yet their applicability to ungauged catchments is unexplored. We adopted a neural network pattern recognition model (ANN-PR) and explainable artificial intelligence approach (SHAP-XAI) to match all ungauged catchments that flow to the Great Barrier Reef to gauged ones based on proxy spatial data. Catchment match suitability was verified using a neural network water quality (ANN-WQ) simulator trained on gauged catchment datasets, tested by simulating DIN for matched catchments in unsupervised learning scenarios. We show that discriminating training data to DIN regime benefits ANN-WQ simulation performance in unsupervised scenarios ( p< 0.05). This phenomenon demonstrates that proxy spatial data is a useful tool to classify catchments with similar DIN regimes. Catchments lacking similarity with gauged ones are identified as priority monitoring areas to gain observed data for all DIN regimes in catchments that flow to the Great Barrier Reef, Australia.

Impact of land use on water quality in buffer zones at different scales in the Poyang Lake, middle reaches of the Yangtze River basin

The water quality of Poyang Lake (PYL) is significantly influenced by land use which is a crucial factor that can exhibit complex changes in the environment and can serve as an indicator of the intensity of human activity. Therefore, this study analyzed the spatial and temporal distribution characteristics of nutrients and investigated the effects of land-use factors on water quality in the PYL during 2016-2019. The main conclusions are as follows: (1) Although there was some variation in the accuracy of the water quality inversion models (random forest (RF), support vector machine (SVM), and multiple statistical regression models), they were homogeneous. In particular, ammonia nitrogen (NH3-N) concentration from band (B) 2 and B2-B10 regression model was more consistent with each other. In contrast, the overall concentration levels from the combined B9/(B2-B4) triple-band regression model were relatively low, with approximately 0.03 mg/L in most areas of PYL. (2) The optimal inversion method varied for different water quality parameters. For instance, RF obtained better inversion of total phosphorus (TP) and total nitrogen (TN), with the fitting coefficient (r2) of 0.78 and 0.81, respectively; SVM had higher accuracy in the inversion of permanganate index (CODMn), with r2 of approximately 0.61; the accuracy of multi-band combined regression model in the inversion of each water quality parameter was at a higher level. (3) The influence of land use on water quality at different scales of buffer zone was different. In general, the correlation between water quality parameters and land use was higher at large spatial scales (1000-5000 m) than at small spatial scales (100 m, 500 m). A common feature of all hydrological stations was the significant negative correlation between crops, buildings, and water quality at all buffer scales. This study is of great practical significance for promoting water environment management and water quality health in the PYL.

Identifying the influence of natural and human factors on seasonal water quality in China: current situation of China's water environment and policy impact

Agricultural production, urbanization, and other anthropogenic activities, the major causes of surface water pollution in China, have dramatically altered hydrological processes and nutrient cycles. Identifying and quantifying the key factors affecting water quality are essential for the better prevention and management of water pollution. However, due to the limitations of traditional statistical analysis methods, it is difficult to evaluate the spatial changes and interactions of influencing factors on water quality. In addition, research on a national scale is difficult, as it involves large-scale and long-term water quality monitoring work. In this study, we collected and collated the monthly average concentrations of four water quality parameters, dissolved oxygen, ammonia nitrogen, chemical oxygen demand, and total phosphorous, based on data from 1547 water quality monitoring stations in China. The combined pollution level of the water quality was assessed using the water quality index. Based on the water quality characteristics, water quality monitoring sites in the dry and wet seasons were grouped using k-means clustering. Eleven environmental factors were evaluated using geodetector software, including six human factors and five natural factors. The results showed that there are high-risk areas for water quality pollution in the eastern and southeastern coastal regions of China in both the dry and wet seasons and that surface water pollution in China is highly spatial heterogenous in both the dry and wet seasons. Among the anthropogenic factors, urban land area is the main factor of water quality pollution in the dry season, and the explanation rate of spatial heterogeneity of integrated water quality pollution index is 20.3%. The number of poultry farms and the area of farmland explained 12.4% and 12.1% of the integrated water quality pollution index in the wet season. The nonlinear relationship between these three anthropogenic and natural factors and their interaction exacerbated water quality pollution. Based on this analysis, we identified the key factors affecting surface water quality in China during the dry and wet seasons, evaluated the achievements of the water environmental protection policies in China in recent years, and proposed future management measures for the effective prevention and control of water quality pollution in high-risk areas.

Effects of comprehensive landscape patterns on water quality and identification of key metrics thresholds causing its abrupt changes

Comprehensive landscape patterns influence water quality with multiple factors, complex processes, and scale dependence. However, studies identifying landscape thresholds causing abrupt water quality changes and characterizing the contribution of topography to water quality are still limited. Exploring the impact mechanisms of natural geographical and landscape characteristics on spatial and seasonal water quality variations is conducive to watershed water resource protection and ecosystem restoration. Based on water quality monitoring data of Minjiahe River in the typical headwater area of the upstream Dan River in China from 2019 to 2021, we employed redundancy analysis, partial redundancy analysis, and nonparametric change-point analysis to analyze the relationship between stream water quality and multi-spatial scale comprehensive landscape patterns, to obtain the interactive and independent contributions of different landscape categories at multi-spatial scales on water quality, and to find the key landscape threshold leading to abrupt changes in water quality. Results showed that landscape configuration, landscape composition, and topographic factors collectively explain over 89.1% of water quality variation. Most seasonal variations in water quality were primarily caused by landscape configuration. The landscape composition was mainly responsible for the differences in water quality variations among spatial scales. The topographic factors made the least independent contribution and had a potential impact on overall water quality variation. In order to protect the water quality of streams, it is more reasonable to regulate the landscape at different scales. At the sub-catchment scale, interspersion and juxtaposition index (IJI) and landscape shape index (LSI) should be controlled below 82% and 22. At the 100 m riparian scale, farmland, urban land, IJI, and LSI should be controlled below 29%, 6.5%, 92%, and 26, respectively. Our results provide important guidance for optimizing landscape patterns and water conservation in the watershed.

Urbanization and agriculture intensification jointly enlarge the spatial inequality of river water quality

Rivers are severely polluted by multiple anthropogenic stressors. An unevenly distributed landscape pattern can aggravate the deterioration of water quality in rivers. Identifying the impacts of landscape patterns on the spatial characteristics of water quality is helpful for river management and water sustainability. Herein we quantified the nationwide water quality degradation in China's rivers and analyzed its responses to spatial patterns of anthropogenic landscapes. The results showed that the spatial patterns of river water quality degradation had a strong spatial inequality and worsened severely in eastern and northern China. The spatial aggregation of agricultural/urban landscape and the water quality degradation exhibits high consistency. Our findings suggested that river water quality would further deteriorate from high spatial aggregation of cities and agricultures, which reminded us that the dispersion of anthropogenic landscape patterns might effectively alleviate water quality pressures.

Output characteristics and driving factors of non-point source nitrogen (N) and phosphorus (P) in the Three Gorges reservoir area (TGRA) based on migration process: 1995-2020

Although physical models at present have made important achievements in the assessment of non-point source pollution (NPSP), the requirement for large volumes of data and their accuracy limit their application. Therefore, constructing a scientific evaluation model of NPS nitrogen (N) and phosphorus (P) output is of great significance for the identification of N and P sources as well as pollution prevention and control in the basin. We considered runoff, leaching and landscape interception conditions, and constructed an input-migration-output (IMO) model based on the classic export coefficient model (ECM), and identified the main driving factors of NPSP using geographical detector (GD) in Three Gorges Reservoir area (TGRA). The results showed that, compared with the traditional export coefficient model, the prediction accuracy of the improved model for total nitrogen (TN) and total phosphorus (TP) increased by 15.46 % and 20.17 % respectively, and the error rates with the measured data were 9.43 % and 10.62 %. It was found that the total input volume of TN in the TGRA had declined from 58.16 × 10⁴ t to 48.37 × 10⁴ t, while the TP input volume increased from 2.76 × 10⁴ t to 4.11 × 10⁴ t, and then decreased to 4.01 × 10⁴ t. In addition Pengxi River, Huangjin River and the northern part of Qi River were high value areas of NPSP input and output, but the range of high value areas of migration factors has narrowed. Pig breeding, rural population and dry land area were the main driving factors of N and P export. The IMO model can effectively improve prediction accuracy, and has significant implications for the prevention and control of NPSP.

Effects of changes in land use structure on nitrogen input in the Pingzhai Reservoir watershed, a karst mountain region

Optimizing land use composition to control nitrogen input into water bodies is one way to address surface source pollution in karst mountain regions. In this study, changes in land use, N sources, and spatial and temporal changes of N migration in the Pingzhai Reservoir watershed were evaluated from 2015 to 2021, and the relationship between land use composition and N input was elucidated. N was the main pollution in the water of the watershed; NO₃^- was the dominant form of N, and it did not react during migration. N came from soil, livestock manure or domestic sewage, and atmospheric deposition. Isolating the fractionation effects of source nitrogen is crucial to improve the accuracy of nitrogen and oxygen isotope traceability in the Pingzhai Reservoir. From 2015 to 2021, the grassland area in the Pingzhai Reservoir increased by 5.52%, the woodland area increased by 2.01%, the water area increased by 1.44%, the cropland decreased by 5.8%, unused land decreased by 3.18%, and construction land remained unchanged. Policies and reservoir construction were the main drivers of changes in land-use type in the catchment. Changes in land use structure affected nitrogen input patterns, with unused land having a highly significant positive correlation with inputs of NH₃-N, NO₂^-, and TN, and construction land having a significant positive correlation with the input of NO₂^-. The inhibitory effect of forest and grassland on nitrogen input in the basin was offset by the promoting effect of cropland and construction land on nitrogen input, with unused land becoming a new focus area for nitrogen emissions due to a lack of environmental management. Modifying the area of different land use types in the watershed can effectively control nitrogen input to the watershed.

Spatial and temporal dynamics of base flow in semi-arid montane watersheds and the effects of landscape patterns and topography

Base flow (BF) is harder to predict than other hydrological signatures. The lack of hydrologically relevant information or adequately broad spectrum of typically selected catchment attributes (particularly landscape and topography) hinders the explanatory power. Our goals were to identify the most influential controls on base flow spatially and temporally and to elucidate the response relationships. Base flow in 19 semi-arid sub-watersheds was separated by digital filtering. One hundred and fourteen sub-watershed attributes were related to base flow using random forest regression. The main results were as follows: (1) Annual BF significantly declined since 1999 due to decreased precipitation, increased air temperature, afforestation, urban expansion, and increasing water consumption. Annual base flow index (BFI), varying between 0.319 and 0.695, showed less noticeable temporal trends. (2) Precipitation (P) and underlying carbonate rocks primarily controlled the spatial variation of annual BF and total flow (TF), with the impacts being positive. Landscape was less influential. After the abrupt runoff decline, landscape composition rather than configuration exerted greater impacts on spatial BF and TF, and the importance of forest increased, whereas landscape configuration was decisive for BFI during the whole observation period. The absence of significant links between landscape configuration and water quantity may result from a scale issue. Concave profile curvatures were found to be topographic variables more important than slopes. The impact of soil was the least. This study would benefit the selection of catchment attributes and spatial extents to quantify these attributes in building BF predicting models in future studies.

Effects of land use and slope on water quality at multi-spatial scales: a case study of the Weihe River Basin

Exploring the impact of land use and slope on basin water quality can effectively contribute to the protection of the latter at the landscape level. This research concentrates on the Weihe River Basin (WRB). Water samples were collected from 40 sites within the WRB in April and October 2021. A quantitative analysis of the relationship between integrated landscape pattern (land use type, landscape configuration, slope) and basin water quality at the sub-basin, riparian zone, and river scales was conducted based on multiple linear regression analysis (MLR) and redundancy analysis (RDA). The correlation between water quality variables and land use was higher in the dry season than in the wet season. The riparian scale was the best spatial scale model to explain the relationship between land use and water quality. Agricultural and urban lands had a strong correlation with water quality, which was most affected by land use area and morphological indicators. In addition, the greater the area and aggregation of forest land and grassland, the better the water quality, while urban land presented larger areas with poorer water quality. The influence of steeper slopes on water quality was more remarkable than that of plains at the sub-basin scale, while the impact of flatter areas was greater at the riparian zone scale. The results indicated the importance of multiple time-space scales to reveal the complex relationship between land use and water quality. We suggest that watershed water quality management should focus on multi-scale landscape planning measures.

Assessing the impact of land use and land cover on river water quality using water quality index and remote sensing techniques

The impact of land use on water quality is becoming a global concern due to the increasing demand for freshwater. This study aimed to assess the effects of land use and land cover (LULC) on the surface water quality of the Buriganga, Dhaleshwari, Meghna, and Padma river system in Bangladesh. To determine the state of water, water samples were collected from twelve locations in the Buriganga, Dhaleshwari, Meghna, and Padma rivers during the winter season of 2015 and collected samples were analysed for seven water quality indicators: pH, temperature (Temp.), conductivity (Cond.), dissolved oxygen (DO), biological oxygen demand (BOD), nitrate nitrogen (NO₃-N), and soluble reactive phosphorus (SRP) for assessing water quality (WQ). Additionally, same-period satellite imagery (Landsat-8) was utilised to classify the LULC using the object-based image analysis (OBIA) technique. The overall accuracy assessment and kappa co-efficient value of post-classified images were 92% and 0.89, respectively. In this research, the root mean squared water quality index (RMS-WQI) model was used to determine the WQ status, and satellite imagery was utilised to classify LULC types. Most of the WQs were found within the ECR guideline level for surface water. The RMS-WQI result showed that the "fair" status of water quality found in all sampling sites ranges from 66.50 to 79.08, and the water quality is satisfactory. Four types of LULC were categorised in the study area mainly comprised of agricultural land (37.33%), followed by built-up area (24.76%), vegetation (9.5%), and water bodies (28.41%). Finally, the Principal component analysis (PCA) techniques were used to find out significant WQ indicators and the correlation matrix revealed that WQ had a substantial positive correlation with agricultural land (r = 0.68, P < 0.01) and a significant negative association with the built-up area (r =  - 0.94, P < 0.01). To the best of the authors' knowledge, this is the first attempt in Bangladesh to assess the impact of LULC on the water quality along the longitudinal gradient of a vast river system. Hence, we believe that the findings of this study can support planners and environmentalists to plan and design landscapes and protect the river environment.

Influences of landscape pattern on water quality at multiple scales in an agricultural basin of western China

Determining the associations between landscape pattern and river water quality and quantifying the abrupt change points of landscape metrics are vital to optimize landscape planning and improve basin water quality. This study took an agricultural basin in western China as a case study. River water quality of 61 sub-basin outlets were monitored during wet and dry seasons from 2020 to 2021. Landscape metrics were extracted at 100 m, 300 m, 500 m riparian buffer and sub-basin scales, respectively. Relationships between water quality and landscape pattern at multiple scales were explored by using redundancy analysis (RDA). Results showed that urban-related landscape metrics served as the primary contributor to degrade water quality during both seasons, followed by cropland-related metrics, which might be attributed to the increase of urban land and reduction of agricultural chemical fertilizer use. Landscape metrics could better explain the water quality variations during wet season than dry season. The explanatory abilities of landscape metrics to overall water quality appeared little difference among spatial scales during wet season, whereas landscape metrics within 100 m riparian buffer had much larger explanatory rate than other spatial scales during dry season. Results of abrupt change point analysis revealed that the abrupt change interval values (ACIVs) of percentage of urban land (PLAND_urban) and the largest patch index of urban land (LPI_urban) differed among COD, TN, and TP. The recommended threshold values of PLAND_urban and LPI_urban for COD, TN, and TP management were smaller than 11.0%, 2.5%, and 1.0%, respectively. When the PLAND_urban or LPI_urban exceeded 19.0%, the TN, TP, and COD pollution would all significantly accelerate. Therefore, a limit value of 19% of PLAND_urban and LPI_urban, respectively is put forward. From dry season to wet season, the ACIVs of PLAND_urban and LPI_urban for COD concentration increased, whereas they decreased for TN and TP concentrations. Our results can provide scientific insights into sustainable landscape planning and effective water quality protection in agricultural basins.

Comparison between graph theory connectivity indices and landscape connectivity metrics for modeling river water quality in the southern Caspian sea basin

The maintenance of connectivity is critical to the proper functioning of an ecosystem. The present study was conducted with the aim of comparing graph theory connectivity indices and landscape connectivity metrics for the purpose of modeling river water quality. To conduct this study, a forest layer was extracted from land cover map and 25 large watersheds were selected. River water quality was then assessed from the perspective of 8 landscape connectivity metrics and 12 graph theory indices. We developed predictive models using stepwise linear regression, power, exponential, and logarithmic models to locate the best model form for each water quality parameter (dependent variable) we examined. The results indicated that models developed using graph theory connectivity indices resulted in higher coefficients of determination (R²) than models developed using landscape metrics. Only 5 independent variables from a potential set of 13 were significant in explaining the variation in water quality parameters. Also, the models with the highest R² attempted to explain variations in CO₃ (0.818), water discharge (0.733), and Ca levels (0.702). Therefore, the results of this study showed that graph theory connectivity indices had more significant correlation with water quality parameters compared to landscape connectivity metrics. This work also indicates that there exist nonlinear relationships among connectivity indices and water quality parameters.

Managing landscape patterns at the riparian zone and sub-basin scale is equally important for water quality protection

Widespread attention has been given to understanding the effect of the landscape pattern on river water quality. However, which spatial scale (riparian zone versus sub-basin) has the greater impact on water quality has long been controversial, since the key metrics that affect water quality varied with spatial scale. Thus, quantifying the spatial scale effects of key landscape metrics on water quality is critical to clarifying which scale of landscape pattern is more conducive to water quality conservation. Here, we adopted variation partitioning analysis (VPA) and random forest models to quantify the landscape pattern impact on water quality at northern Erhai Lake during the 2019 rainy season (early, mid, and late), and comprehensively analyze the key landscape metrics on different scales. The results revealed that the riparian zone and sub-basin scale landscape patterns explained similar water quality variations (difference only 0.9%) in the mid (August) and late rainy season (October), but exhibited a large difference (24.1%) during the early rainy season (June). Furthermore, rivers were primarily stressed by nitrogen pollution. Maintaining the Grassland_ED > 27.99 m/ha, Grassland_LPI > 4.19%, Farmland_LSI < 3.2 in the riparian zone, and Construction_ED < 1.69 m/ha, Construction_LSI < 2.46, Farmland_PLADJ < 89.0% at the sub-basin scale could significantly reduce the TN concentration in the stream. Meanwhile, managing of these metrics can effectively prevent rapid increases of TN in rivers. Moreover, due to the low phosphorus concentration in the rivers, none of the landscape metrics significantly explained the variation in TP. This study explored the spatial scale effect of landscape patterns on water quality and revealed the driving factors of nutrient variation. This study will provide a scientific basis for aquatic environmental management in plateau watersheds.

Stream water quality optimized prediction based on human activity intensity and landscape metrics with regional heterogeneity in Taihu Basin, China

The driving effects of landscape metrics on water quality have been acknowledged widely, however, the guiding significance of human activity intensity and landscape metrics based on reference conditions for water environment management remains to be explored. Thus, we used the self-organized map, long- and short-term memory (LSTM) algorithm, and geographic detectors to simulate the response of human activity intensity and landscape metrics to water quality in Taihu Lake Basin, China. Fitting results of LSTM displayed that the accuracy was acceptable, and scenario 2 (regional heterogeneity) was more efficient than scenario 1 (regional consistent) in the improvement of water quality. In the driving analysis for the reference conditions, clusters I and II (urban agglomeration areas) were mainly affected by the amount of production wastewater per unit of developed land and the amount of livelihood wastewater per unit of developed land, respectively. Their optimal values were 0.09 × 10³ t/km² (reduction of 35.71%) and 0.2 × 10³ t/km² (reduction of 4.76%). Cluster III (agricultural production areas) was mainly affected by interference intensity, and the optimal value was 2.17 (increased 38.22%), and cluster IV (ecological forest areas) was mainly affected by the fragmentation of cropland, and the optimal value was 1.14 (reduction of 1.72%). The research provides a reference for the prediction of water quality response and presents an ecological and economic sustainability way for watershed governance.

New method for scaling nonpoint source pollution by integrating the SWAT model and IHA-based indicators

Nonpoint source (NPS) pollution shows spatial scaling effects because it is affected by topography, river networks, and many other factors. Currently, the lack of an integrated methodology for quantifying the scaling effect has become a crucial barrier in evaluating NPS pollution. In this study, a new method was proposed for scaling NPS pollution by integrating hydrological model and hydrological alteration indicators. Nested catchments were delineated by eight-direction algorithm, and a semidistributed hydrological model was used to simulate the interannual process within the drainage area and to obtain data series of runoff, sediment, and total phosphorus (TP) at different spatial scales. In addition, the average, the extrema, the change rate and feature variables of each type of indicators were proposed to quantitatively describe the pattern of NPS pollution at different spatial scales. The results show the coefficients of variation (CVs) of most runoff and TP indicators are 0.6-0.8, while those of sediment vary greatly from 0.4 to 1.6 with the threshold of those indicators being 0.33. With the increase in drainage area, the NPS load-related indicators show an increasing trend, while load intensity indicators show a decreasing trend and their changing patterns are affected by the heterogeneity of topographic or hydrological information included. Based on logarithmic variance of the change rate, 825 km² was identified as the turning point for scaling transformation where the slope changes dramatically. The proposed methodology comprehensively describes features of the NPS scaling effect that could be utilized for targeted monitoring and control of NPS pollution in other watersheds.

Sources and migration similarly determine nitrate concentrations: Integrating isotopic, landscape, and biological approaches

Rapid land use change has significantly increased nitrate (NO₃^-) loading to rivers, leading to eutrophication, and posing water security problems. Determining the sources of NO₃^- to waters and the underlying influential factors is critical for effectively reducing pollution and better managing water resources. Here, we identified the sources and influencing mechanisms of NO₃^- in a mixed land-use watershed by integrating stable isotopes (δ¹⁵N-NO₃^- and δ¹⁸O-NO₃^-), molecular biology, water chemistry, and landscape metrics measurements. Weak transformation processes of NO₃^- were identified in the river, as evinced by water chemistry, isotopes, species compositions, and predicted microbial genes related to nitrogen metabolism. NO₃^- concentrations were primarily influenced by exogenous inputs (i.e., from soil nitrogen (NS), nitrogen fertilizer (NF), and manure & sewage (MS)). The proportions of NO₃^- sources seasonally varied. In the wet season, the source contributions followed the order of NS (38.6 %) > NF (31.4 %) > atmospheric deposition (ND, 16.2 %) > MS (13.8 %). In the dry season, the contributions were in the order of MS (39.2 %) > NS (29.2 %) > NF (29 %) > ND (2.6 %). Farmland and construction land were the original factors influencing the spatial distribution of NO₃^- in the wet and dry seasons, respectively, while slope, basin relief (HD), hypsometric integral (HI), and COHESION, HD were the primary indicators associated with NO₃^- transport in the wet and dry seasons, respectively. Additionally, spatial scale differences were observed for the effects of landscape structure on NO₃^- concentrations, with the greatest effect at the 1000-m buffer zone scale in the wet season and at the sub-basin scale in the dry season. This study overcomes the limitation of isotopes in identifying nitrate sources by combining multiple approaches and provides new research perspectives for the determination of nitrate sources and migration in other watersheds.

A holistic assessment of spatiotemporal variation, driving factors, and risks influencing river water quality in the northeastern Qinghai-Tibet Plateau

The Qinghai-Tibet Plateau (QTP) is the source for many of the most important rivers in Asia. It is also an essential ecological barrier in China and has the characteristic of regional water conservation. Given this importance, we analyzed the spatiotemporal distribution patterns and trends of 10 water quality parameters. These measurements were taken monthly from 67 monitoring stations in the northeastern QTP from 2015 to 2019. To evaluate water quality trends, major factors influencing water quality, and water quality risks, we used a series of analytical approaches including Mann-Kendall test, Boruta algorithm, and interval fuzzy number-based set-pair analysis (IFN-SPA). The results revealed that almost all water monitoring stations in the northeastern QTP were alkaline. From 2015 to 2019, the water temperature and dissolved oxygen of most monitoring stations were significantly reduced. Chemical oxygen demand, permanganate index, five-day biochemical oxygen demand, total phosphorus, and fluoride all showed a downward trend across this same time frame. The annual average total nitrogen (TN) concentration fluctuation did not significantly decrease across the measured time frame. Water quality index (WQI-DET) indicated bad or poor water quality in the study area; however, water quality index without TN (WQI-DET') reversed the water quality value. The difference between the two indexes suggested that TN was a significant parameter affecting river water quality in the northeastern QTP. Both Spearman correlation and Boruta algorithm show that elevation, urban land, cropland, temperature, and precipitation influence the overall water quality status in the northeastern QTP. The results showed that between 2015 and 2019, most rivers monitored had a relatively low risk of degradation in water quality. This study provides a new perspective on river water quality management, pollutant control, and risk assessment in an area like the QTP that has sensitive and fragile ecology.

A source-sink landscape approach to mitigation of agricultural non-point source pollution: Validation and application

Optimizing landscape pattern to reduce the risk of non-point source (NPS) pollution is an effective measure to improve river water quality. The "source-sink" landscape theory is a recent research tool for landscape pattern analysis that can effectively integrate landscape type, area, spatial location, and topographic features to depict the spatial heterogeneity of NPS pollution. Based on this theory, we quantitatively analyzed the influence of "source-sink" landscape pattern on the river water quality in one of the most intensive agricultural watersheds in Southeastern China. The results indicated that the proportion of "sink" landscape (68.59%) was greater than that of "source" landscape (31.41%) in the study area. In addition, when elevation and slope increased, the "source" landscape proportion decreased, and the "sink" landscape proportion increased. Nitrogen (N) and phosphorus (P) pollutants in rivers showed significant seasonal and spatial variations. Farmland was the primary source of nitrate nitrogen (NO₃^--N) and total nitrogen (TN) pollution, whereas residential land was the primary source of ammonium nitrogen (NH₄⁺-N) and total phosphorus (TP) pollution. Intensively cultivated areas and densely inhabited areas degraded water quality despite high proportions of forest land. The four "source-sink" landscape indices (LWLI, LWLI'e, LWLI's, LWLI'd) had significant positive correlations with NO₃^--N and TN and weak correlations with NH₄⁺-N and TP. The capacity of LWLI to quantify the NPS pollution was greater in agricultural areas than in residential areas. The "source-sink" landscape thresholds resulted in abrupt changes in water quality. When LWLI was ∼0.35, the probability of river water quality degradation increased sharply. The results suggest the importance of optimizing the "source-sink" landscape pattern for mitigating agricultural NPS pollution and provide policy makers with adequate new information on the agroecosystem-environmental interface in highly developed agricultural watersheds.