Control Catchment Research Articles

The effect of veld fires on the hydrological response of streamflow

Veld fires are natural occurrences with the potential to impact thousands of hectares of vegetation, and in doing so, changes soil characteristics, for both urban and rural areas. It is therefore reasonable to assume that the hydrological response of a catchment could be affected by fire. The main aim of this research was to investigate the hydrological changes caused by fire on a catchment scale using a case study. On 9 March 2015, a wildfire which started in Jonkershoek nature reserve destroyed indigenous fynbos vegetation and afforested areas. Within the nature reserve, there are multiple rainfall and runoff stations, which provided a means of measuring any possible hydrological changes due to these fire events. Four catchments were used for this research, one main catchment (fynbos area) and three sub-catchments (afforested areas). Fifty-six percent of the main catchment burned, while two sub-catchments were completely burned and the other primarily unaffected by the fire. The main catchment’s hydrological response was analysed by comparing the hydrographs of comparable pre- and post-fire runoff events. Eighteen comparable events were used for the analysis. The mean runoff volume increased by approximately 7% after the fire and mean peak flow by 50%. The change was even more noticeable when comparing the two sub-catchments affected by the veld fire and the unburned sub-catchment with each other. All the sub-catchments were similar in size and were located close enough to each other to be represented by one rainfall station. Before the fire, the average daily streamflows between the unburned (control) and burned catchments were similar; however after the fire the average daily streamflow of the two burned catchments in comparison to the control catchment increased by 45% and 50%, respectively. The mean runoff volume from the two affected/burned catchments, after the fire, for individual events increased by approx. 72% and 52% in comparison to the control catchment. The mean peak flows increased by approximately 173% and 110% in comparison to the control catchment.

Do awareness-focussed approaches to mitigating diffuse pollution work? A case study using behavioural and water quality evidence

Efforts to tackle diffuse water pollution from agriculture are increasingly focusing on improving farmers' awareness under the expectation that this would contribute to adoption of best management practices (BMPs) and, in turn, result in water quality improvements. To date, however, no study has explored the full awareness-behaviour-water quality pathway; with previous studies having mostly addressed the awareness-behaviour link relying on disciplinary approaches. Using an interdisciplinary approach, we investigate whether awareness-focussed approaches to mitigating diffuse water pollution from agriculture indeed result in water quality improvement, addressing the pathway in full. We worked with Dŵr Cymru Welsh Water (a water and waste utility company in the UK) on a pesticide pollution intervention programme, referred to as “weed wiper trial”. The main goal of the trial was to raise farmers' awareness regarding pesticide management practices and to promote uptake of BMPs to tackle the rising concentrations of the pesticide MCPA (2-methyl-4-chlorophenoxyacetic acid) in raw water in three catchments in Wales. Using factorial analysis of variance, we analysed MCPA concentrations from 2006 to 2019 in the three targeted catchments and in three control catchments. This was followed by semi-structured in-depth interviews with institutional stakeholders and farmers with varying degrees of exposure to the weed wiper trial. Results show that MCPA concentration for both targeted and control catchments had reduced after the implementation of the weed wiper trial. However, the decline was significantly larger (F(1) = 6.551, p < 0.05, n = 3077, Partial eta-squared (ηp2) = 0.002) for the targeted catchments (mean = 45.2%) compared to the control catchments (mean = 10.9%). Results from the stakeholder interviews indicate that improved awareness contributed to changes in farmers' behaviour and that these can be related to the water quality improvements reflected by the decline in MCPA concentration. Alongside awareness, other psychosocial, economic, agronomic factors, catchment and weather conditions also influenced farmer's ability to implement BMPs and thus overall water quality improvements.

Influence of the anti-erosion complex on erosion-hydrological indicators and productivity of cultivated crops

An assessment of the impact of the anti-erosion complex on the productivity of cultivated crops has presented on the example of many years of experience in contour-ameliorative agriculture of the Kursk FARC located in the northern part of the Medvensky district of the Kursk region. Two subtypes of chernozem represent the soil of the experiment: typical and leached. As anti-erosion measures, narrow-row stock-regulating poplar forest belts, reinforced by a simple hydraulic structure (ditch and rampart along the lower edge), have designed and planted along the contour at a distance of 216 m. There are no anti-erosion measures in the control catchment. Analysis of long-term data showed a decrease in runoff and the absence of soil washout in the catchment area with a system of narrow-row forest belts 38 years after the experiment has laid in comparison with the control. For the control catchment, a maximum soil washing of 22 t/ha recorded which significantly exceeds the permissible standards of flush. A significant increase in crop productivity averaged 5.2 feed units. The profitability of growing crops in fields with forest belts exceeds the control by almost 50%. The anti-erosion complex performs the task of strengthening the soil-protective role of the farming system.

Quantifying the temporal variation of the contribution of fine sediment sources to sediment yields from Chilean forested catchments during harvesting operations

Fingerprinting techniques were incorporated into a paired catchment investigation in southern Chile to quantify the contribution of three fine sediment sources (catchment surfaces, forest roads and stream channels) to catchment suspended sediment yields during forest harvesting and replanting operations. Optimum composite fingerprints for use in sediment source discrimination and apportionment comprised 137Cs and 210Pbex for the control catchment (LUC) throughout the study and for the treatment catchment (LUT) during the pre-harvest period, and 137Cs and soil organic matter during harvest and post-harvest periods for LUT. Prior to harvesting, the dominant sediment source to the sediment load in both catchments was the stream channel and remained relatively constant throughout the study for LUC. For the entire study period the total suspended sediment yield from LUT (3,160 kg ha-1) approximately doubled that from LUC (1,650 kg ha-1). Most of this difference is accounted for by the increase in sediment output during the rainy months following clearcutting. The disturbance associated with forest operations in LUT caused the contributions to the load from the catchment slopes and forest roads to increase markedly (total contributions 835 and 795 kg ha-1, respectively). However, the total contribution from the stream channel for LUT during the study period (1,530 kg ha-1) remained similar to that from LUC. The results of the investigation demonstrated that any attempt to reduce sediment loading from forest harvesting would require adopting best management practices to reduce sediment mobilization from catchment surfaces and forest roads.

Outflow hydrology of tailwater recovery systems with comparisons to a control catchment*

AbstractConservation practices termed tailwater recovery (TWR) systems are being installed throughout the Lower Mississippi River basin to mitigate damage from agricultural production. Many of the costs and benefits of these systems have yet to be documented; therefore, the objective of this study is to assess whether TWR systems have the potential to impact on the hydrology of downstream systems. In this study, we compare hydrologic parameters of two TWR systems—one with an on‐farm storage reservoir (TWR‐OFS) and one without (TWR‐D)—and a control ditch (CTRL). Systems were monitored at one outflow location each using flow meters. Estimated percentage differences between TWR‐D and CTRL sites were determined, in which the CTRL site was found to have greater values reflected in the parameters, including length of event median 94% and mean 60%; time to peak median 80% and mean 70%; time to base median 88% and mean 60%; peak flow median 86% and mean 67% and discharge median 134% and mean 101%. The TWR‐D and TWR‐OFS sites differed in peak flow, with TWR‐OFS being 40% greater. Evidence from this investigation shows that TWR systems have the potential to influence downstream hydrology by retaining water.

District-level health system strengthening for universal health coverage: evidence from a longitudinal cohort study in rural Madagascar, 2014-2018

IntroductionDespite renewed commitment to universal health coverage and health system strengthening (HSS) to improve access to primary care, there is insufficient evidence to guide their design and implementation. To address...

Restoration of afforested peatland: Immediate effects on aquatic carbon loss

Peatland restoration is undertaken to bring back key peatland ecosystem services, including carbon storage. In the case of drained, afforested blanket peatlands, restoration through drain blocking and tree removal may impact upon aquatic carbon concentrations and export, which needs to be accounted for when considering the carbon benefits of restoration.This study investigated concentrations and export of aquatic carbon from a drained, afforested blanket bog catchment, where 12% of the catchment underwent drain blocking and conifer removal (termed ‘forest-to-bog’ restoration), and from two control catchments: one in open bog and one that remained afforested. Using a before-after-control-impact (BACI) design, we found no significant increases in concentrations or export of aquatic carbon (DOC, POC or DIC) in the first year following forest-to-bog restoration (i.e. across the whole post-restoration period). However, increased DOC concentrations were observed in the first summer (2015) post-restoration, and seasonally increased DOC export was noted during storm events in the autumn of the same year. The lack of significant effects of forest-to-bog restoration on aquatic carbon export may be a consequence of the small proportion of the catchment (12%) undergoing management. In terms of management, the removal of more of the forestry residues (i.e., brash) may help to mitigate effects on aquatic carbon, by removing a potential DOC and POC source. Restoring small areas at a time (≤12%) should result in minimal aquatic carbon export issues, in contexts similar to the current study.

Short-term runoff prediction with GRU and LSTM networks without requiring time step optimization during sample generation

Runoff forecasting is an important approach for flood mitigation. Many machine learning models have been proposed for runoff forecasting in recent years. To reconstruct the time series of runoff data into a standard machine learning dataset, a sliding window method is usually used to pre-process the data, with the size of the window as a variable parameter which is commonly referred to as the time step. Conventional machine learning methods, such as artificial neural network models (ANN), require optimization of the time step because both too small and too large time steps reduce prediction accuracy. In this work two popular variants of Recurrent Neural Network (RNN) named Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) networks were employed to develop new data-driven flood forecasting models. GRU and LSTM models are in theory able to filter redundant information automatically, and therefore a large time step is expected to not reduce prediction accuracy. The three models (LSTM, GRU, and ANN) were applied to simulate runoff in the Yutan station control catchment, Fujian Province, Southeast China, using hourly discharge measurements of one runoff station and hourly rainfall of four rainfall stations from 2000 to 2014. Results show that the prediction accuracy of LSTM and GRU models increases with increasing time step, and eventually stabilizes. This allows selection of a relatively large time step in practical runoff prediction without first evaluating and optimizing the time step required by conventional machine learning models. We also show that LSTM and GRU models perform better than ANN models when the time step is optimized. GRU models have fewer parameters and less complicated structures compared to LSTM models, and our results show that GRU models perform equally well as LSTM models. GRU may be the preferred method in short term runoff predictions since it requires less time for model training.

Sediment-related water quality of small hill-country streams near Whatawhata, New Zealand. Response to integrated catchment management (ICM)

ABSTRACT Degradation of aquatic ecosystems by fine sediment is a global environmental problem, and damages reflect particle properties as well as mass concentration. We studied sediment-related water quality (SRWQ) of small streams in hill-country near Whatawhata, NZ, that have been monitored since 1995. In 2001, an integrated catchment management (ICM) plan was implemented in the Mangaotama (originally sheep-beef) catchment – including conversion of steep headwaters to pine plantation, cattle exclusion, and native riparian plantings. The long-term SRWQ within the Mangaotama catchment was examined with reference to a nearby native-forested control catchment. Correlations between visual clarity, suspended particulate matter concentration and related variables were moderate to fairly strong. The ICM has not improved SRWQ as expected – plausibly because removal of cattle disturbance has been offset by decline in pasture ground cover under increasing shade of tree plantings. Worsening SRWQ is expected over the next several decades as channels widen towards a ‘forest’ morphology.

Change in low flows due to catchment management dynamics—Application of a comparative modelling approach

AbstractUnderstanding the natural low flow of a catchment is critical for effective water management policy in semi‐arid and arid lands. The Geba catchment in Ethiopia, forming the headwaters of Tekeze‐Atbara basin was known for its severe land degradation before the recent large scale Soil and Water conservation (SWC) programs. Such interventions can modify the hydrological processes by changing the partitioning of the incoming rainfall on the land surface. However, the literature lacks studies to quantify the hydrological impacts of these interventions in the semi‐arid catchments of the Nile basin. Statistical test and Indicators of Hydrological Alteration (IHA) were used to identify the trends of streamflow in two comparatives adjacent (one treated with intensive SWC intervention and control with fewer interventions) catchments. A distributed hydrological model was developed to understand the differences in hydrological processes of the two catchments. The statistical and IHA tools showed that the low flow in the treated catchment has significantly increased while considerably decreased in the control catchment. Comparative analysis confirmed that the low flow in the catchment with intensive SWC works was greater than that of the control by &gt;30% while the direct runoff was lower by &gt;120%. This implies a large proportion of the rainfall in the treated catchment is infiltrated and recharge aquifers which subsequently contribute to streamflow during the dry season. The proportion of soil storage was more than double compared to the control catchment. Moreover, hydrological response comparison from pre‐ and post‐intervention showed that a drastic reduction in direct runoff (&gt;84%) has improved the low flow by &gt;55%. This strongly suggests that the ongoing intensive SWC works have significantly improved the low flows while it contributed to the reduction of total streamflow in the catchment.

Catchment-scale export of antibiotic resistance genes and bacteria from an agricultural watershed in central Iowa.

Antibiotics are administered to livestock in animal feeding operations (AFOs) for the control, prevention, and treatment of disease. Manure from antibiotic treated livestock contains unmetabolized antibiotics that provide selective pressure on bacteria, facilitating the expression of anti-microbial resistance (AMR). Manure application on row crops is an agronomic practice used by growers to meet crop nutrient needs; however, it can be a source of AMR to the soil and water environment. This study in central Iowa aims to directly compare AMR indicators in outlet runoff from two adjacent (221 to 229 ha) manured and non-manured catchments (manure comparison), and among three catchments (600 to 804 ha) with manure influence, no known manure application (control), and urban influences (mixed land use comparison). Monitored AMR indicators included antibiotic resistance genes (ARGs) ermB, ermF (macrolide), tetA, tetM, tetO, tetW (tetracycline), sul1, sul2 (sulfonamide), aadA2 (aminoglycoside), vgaA, and vgaB (pleuromutilin), and tylosin and tetracycline resistant enterococci bacteria. Results of the manure comparison showed significantly higher (p<0.05) tetracycline and tylosin resistant bacteria from the catchment with manure application in 2017, but no differences in 2018, possibly due to changes in antibiotic use resulting from the Veterinary Feed Directive. Moreover, the ARG analysis indicated a larger diversity of ARGs at the manure amended catchment. The mixed land use comparison showed the manure amended catchment had significantly higher (p<0.05) tetracycline resistant bacteria in 2017 and significantly higher tylosin resistant bacteria in 2017 and 2018 than the urban influenced catchment. The urban influenced catchment had significantly higher ermB concentrations in both sampling years, however the manure applied catchment runoff consisted of higher relative abundance of total ARGs. Additionally, both catchments showed higher AMR indicators compared to the control catchment. This study identifies four ARGs that might be specific to AMR as a result of agricultural sources (tetM, tetW, sul1, sul2) and optimal for use in watershed scale monitoring studies for tracking resistance in the environment.

Quickflow response to forest harvesting and recovery in a northern hardwood forest landscape

AbstractForest harvesting often increases catchment quickflow (QF, water delivered rapidly to the stream channel), a metric of high‐flow events controlling a catchment's solute and sediment export. Nevertheless, our understanding of QF responses to various silvicultural strategies (e.g., clearcutting, selection harvest, and shelterwood harvest) is incomplete. We present a 31‐year examination of QF delivery from treatment (clearcut, selection harvest, and shelterwood harvest) and control catchments in a deciduous forest landscape in central Ontario, Canada. Growing season root‐zone storage capacity was estimated using a water balance approach to evaluate temporal changes in QF response to precipitation (P) for pretreatment and posttreatment periods. Threshold relationships between QF and P were assessed for control and treatment catchments for pretreatment and posttreatment periods using piecewise regression. Root‐zone storage capacity demarcated shifts in the hydrologic regime arising from forest harvesting and subsequent regeneration. This was particularly pronounced for clearcutting where postharvest decline in root‐zone storage capacity was followed by a rise to preharvest values. Similar pretreatment threshold relationships between QF and P, and near‐identical P thresholds for producing significant QF, reflected similar soil and overburden depths in the catchments. Harvesting effects were indicated by increases in QF/P ratios for relative small P and the number of P events that generated QF, thus changing treatment QF vs. P threshold relationships. Prior to harvesting there was no significant increase in QF with P below a threshold P of 35–45 mm; however, there was a significant QF vs. P relationship below this threshold for all treatments postharvest. Clearcutting increased the number of QF events for the entire postharvest period and the first 9‐year postharvest compared to the other treatments; nevertheless, evidence for intertreatment differences in total QF depth delivered from the catchments during the growing season was inconclusive. Our work suggests that changes in threshold relationships between QF and P, coupled with knowledge of the physical processes underlying them, are useful when evaluating hydrologic responses to forest harvesting.

The Legacy of a Severe Wildfire on Stream Nitrogen and Carbon in Headwater Catchments

Large, high-severity wildfires alter the physical and biological conditions that determine how catchments retain and release nutrients and regulate streamwater quality. The short-term water quality impacts of severe wildfire are often dramatic, but the longer-term responses may better reflect terrestrial and aquatic ecosystem recovery. We followed streamwater chemistry for 14 years after the largest fire in recorded Colorado history, the 2002 Hayman Fire, to characterize patterns in nitrogen (N) and carbon (C) export. Throughout the post-fire period, stream nitrate and total dissolved N (TDN) remained elevated in 10 burned catchments relative to pre-burn periods and 4 unburned control catchments. Both the extent of fire in a catchment and wildfire severity influenced stream N concentrations. Nitrate was more than an order of magnitude higher in streams draining catchments that burned to a high extent (> 60% of their areas) compared to unburned catchments. Unburned catchments retained more than 95% of atmospheric N inputs, but N retention in burned catchments was less than half of N inputs. Unlike N, stream C was elevated in catchments that burned to a lesser extent (30–60% of their areas burned), compared to either unburned or extensively burned catchments. Remotely sensed estimates of upland and riparian vegetation cover suggest that burned forests could require several more decades before forest cover and nutrient demand return to pre-fire levels. The persistent stream N increases we report are below drinking water thresholds, but exceed ecoregional reference concentrations for healthy stream ecosystems and indicate that extensively burned headwater catchments no longer function as strong sinks for atmospheric N. Combined with increasing trends in wildfire severity and elevated N deposition, our findings demonstrate the potential for substantial post-wildfire changes in ecosystem N retention and have implications for nutrient export to downstream waters.

Water quality improvements from implementation of tailwater recovery systems

Impacts of nutrient loading from agricultural landscapes have gained widespread attention and led to the implementation of conservation practices aimed at mitigating nutrient loss to downstream systems. A relatively new conservation practice, tailwater recovery (TWR) systems, has been considered as a potential mitigation strategy. The objective of this study was to compare concentrations and loads between effluent of TWR systems (TE) and effluent of similar catchments without TWR systems (CE), under the same agronomic management. Tailwater recovery systems, CE, and paired differences between TE and CE were compared over seasons. Nutrient concentrations and hydrologic discharge were monitored at five TWR system outflow and control catchment outflow locations on a flow event basis. Results revealed that of all experimental analytes, including total suspended solids, total phosphorus, total Kjeldahl nitrogen, nitrate–nitrite, and ammonium, only total phosphorus concentrations were greater (F1,97 = 8.58, p < 0.005) at CE locations than TE locations. The difference between loads of TE and CE showed reduced loads per hectare leaving TWR systems. This included loads of total phosphorus, total nitrogen, organic nitrogen, total inorganic nitrogen, and nitrate–nitrite. Analyses revealed no differences across seasons for TE and paired differences between TE and CE. However, CE concentrations and loads were seasonally different, where ammonium was greater in the summer than in the winter and spring, but not different than fall. Additionally, only loads of total nitrogen were different across seasons, where winter was greater than spring, but not different than summer and fall. This study provides evidence that TWR systems are a positive contribution to reducing nutrient loads compared to control locations.

Time-varying parameter models for catchments with land use change: the importance of model structure

Abstract. Rapid population and economic growth in Southeast Asia has been accompanied by extensive land use change with consequent impacts on catchment hydrology. Modeling methodologies capable of handling changing land use conditions are therefore becoming ever more important and are receiving increasing attention from hydrologists. A recently developed data-assimilation-based framework that allows model parameters to vary through time in response to signals of change in observations is considered for a medium-sized catchment (2880 km2) in northern Vietnam experiencing substantial but gradual land cover change. We investigate the efficacy of the method as well as the importance of the chosen model structure in ensuring the success of a time-varying parameter method. The method was used with two lumped daily conceptual models (HBV and HyMOD) that gave good-quality streamflow predictions during pre-change conditions. Although both time-varying parameter models gave improved streamflow predictions under changed conditions compared to the time-invariant parameter model, persistent biases for low flows were apparent in the HyMOD case. It was found that HyMOD was not suited to representing the modified baseflow conditions, resulting in extreme and unrealistic time-varying parameter estimates. This work shows that the chosen model can be critical for ensuring the time-varying parameter framework successfully models streamflow under changing land cover conditions. It can also be used to determine whether land cover changes (and not just meteorological factors) contribute to the observed hydrologic changes in retrospective studies where the lack of a paired control catchment precludes such an assessment.

Adolescent motor vehicle crash prevention through a trauma center-based intervention program.

Therapy, level IV.

Long-Term Downstream Effects of a Dam on a Lowland River Flow Regime: Case Study of the Upper Narew

Most European riverine ecosystems suffer from the negative influence of impoundments on flow regime. Downstream effects of dams lead to a number of environmental and socioeconomic risks and, therefore, should be thoroughly examined in specific contexts. Our study aims to quantify the downstream effects of the Siemianówka Reservoir (Upper Narew, Poland), using statistical analysis of key elements of the river’s flow regime, such as the flow duration and recurrence of floods and droughts. In a comparative study on control catchments not influenced by impoundments (the Supraśl and Narewka Rivers), we revealed the following downstream effects of the analyzed dam: significant shortening of spring floods, reduction of the duration and depth of summer droughts, decrease of the maximum discharge, and homogenization of the discharge hydrographs. Although we determined a significant decrease in the duration of summer floods in the “before” and “after” dam function periods, we showed that this issue is regional, climate-related, and replicated in control catchments, rather than an evident downstream effect of the dam. We conclude that significant hydrological downstream effects of the Siemianówka dam–reservoir system could have been the main driver inducing the deterioration of the anastomosing stretch of the Narew River downstream of the dam.

Nitrate leaching from a sub-alpine coniferous forest subjected to experimentally increased N deposition for 20 years, and effects of tree girdling and felling

Increased nitrogen (N) deposition rates were simulated for 20 years (1995–2015) by sprinkling rain water enriched with NH4NO3 (+22 kg ha−1 year−1 N) in a small headwater catchment within a spruce (Picea abies) forest at Alptal (central Switzerland). The added N was labelled with 15NH 4 15 NO3 during the first year. A control catchment was labelled in years 6 and 8. NO3 − leaching doubled in the N-addition catchment during the first year, resulting almost entirely from loss of 15N-labelled NO3 −. The proportion of added N leached as NO3 −–N reached 1/3 after 5 years, then remained stable for the next 10 years, at 6.6–10 kg ha−1 year−1 compared with 1.4–2.9 kg ha−1 year−1 in the control catchment. In each catchment, half of the large trees were girdled during the 15th year and felled 1 year later to assess their role in soil N retention. Girdling and felling resulted in a strong increase in NO3 −–N leaching from the N-addition catchment (up to 19 kg ha−1 year−1). The increase was significantly smaller in the control (up to 4 kg ha−1 year−1). NO3 − leached after tree girdling contained only 0.18% of the originally applied tracer in the control and none detectable in the N-addition catchment. The increased leaching was thus mainly from new N and due to reduced N immobilisation by trees rather than to increased N mineralisation and nitrification in the soil. Dissolved organic N leaching (average 6.1 kg ha−1 year−1), in contrast, was unaffected by the treatment.

Sediment transport in steep forested catchments – An assessment of scale and disturbance

Stream sediment loads (both bedload and suspended sediment) are problematic to measure due to the time and equipment needed. There is a dearth of such data sets globally let alone for Australia. However, such data are needed to quantify sediment transport type and rates, landscape evolution, effect of human disturbance as well as patterns and temporal response. Here we present the findings from 8 steepslope forested catchments dominated by headwater streams (size range 15–100ha) in south-eastern Australia where both bedload and suspended load have been measured over multiple years. The results demonstrate that suspended load is the dominant component and there is no consistent suspended to bedload ratio for the catchments. The suspended sediment to bedload ratio appears to be catchment specific. There was no relationship between total load (or bedload/suspended load) and average catchment slope, stream length, shape or any geomorphic descriptor. However catchment total load was found to be significantly related to catchment area. Of the 8 catchments examined here, 6 had been harvested for timber in previous decades (with large areas of forest removed) while 2 catchments had minimal disturbance (Control catchments). There was no difference in sediment loads from the harvested and Control catchments. The results demonstrate that although land disturbance had previously occurred the management practices employed in each catchment were effective in the long term. This provides confidence that the forest harvesting and subsequent management do not produce detrimental effects in the medium to long term. An assessment of erosion rates and likely soil production rates suggests that the catchments are eroding soil at the rate it is being produced.

Do agri‐environment schemes result in improved water quality?

Summary Improved water quality, through a reduction in diffuse pollution from agricultural sources, is an expected benefit of agri‐environment schemes, but this has yet to be demonstrated in practice. Here, we evaluate the impact of Welsh agri‐environment schemes on water quality and freshwater ecosystem condition through a combined monitoring and modelling framework. To determine the influence of the agri‐environment schemes on ecosystem condition, spatially independent catchments dominated by a single scheme (&gt;40% of catchment) were compared to control catchments dominated (&gt;70%) by agricultural land that was not part of any scheme. Biological indicators of water quality were monitored at the outfall of each catchment and a spatially explicit modelling framework of diffuse pollutant emissions applied to each of the 80 catchments. Direct comparison (scheme/non‐scheme) was unable to identify any significant effect of agri‐environment scheme participation. However, derived biological indicators that reflected organic pollution, eutrophication and pesticide run‐off were strongly correlated with modelled concentrations of corresponding diffuse pollutants, thus providing a ground‐truth for the models. Scenarios that assessed the correct counterfactuals (i.e. the influence of scheme entry on pollutant output) were developed for the whole of Wales. The models indicated an important effect of scheme entry on water quality, but this effect was not evenly distributed across the landscape. Synthesis and applications. Our results indicate that agri‐environment schemes can deliver improvements in water quality, through a reduction in diffuse pollution from agricultural sources. However, it is not easy to demonstrate scheme effectiveness; the combination of field survey and modelling used here provides a framework for addressing these difficulties. A spatially targeted approach for agri‐environment scheme options to protect water resources from diffuse pollution is likely to be most effective at delivering water quality improvements.