Search

The file attached to this record is the author's final peer reviewed version. The Publisher's final version can be found by following the DOI link. Pesticide losses from agricultural land to water can result in the environmental deterioration of receiving systems. Mathematical models can make important contributions to risk assessments and catchment management. However, some mechanistic models have high parameter requirements which can make them difficult to apply in data-poor areas. In addition, uncertainties in pesticide properties and applications are difficult to account for using models with long run-times. Alternative, simpler, conceptual models are easier to apply and can still be used as a framework for process interpretation. Here, we present a new conceptual model of pesticide behaviour in surface water catchments, based on continuous water balance calculations. Pesticide losses to surface waters are calculated based on the displacement of a limited fraction of the soil pore water during storm events occurring after application. The model was used to describe the behaviour of metaldehyde in a small (2.2 km2) under-drained catchment in Eastern England. Metaldehyde is a molluscicide which has been regularly detected at high concentrations in many drinking water supply catchments. Measured peak concentrations in stream water (to about 9 micro g L-1) occurred in the first few storm events after application in mid-August. In each event, there was a quasi-exponential decrease in concentration during hydrograph recession. Peak concentrations decreased in successive events - responding to rainfall but reflecting an effective exhaustion in soil supply due to degradation and dissipation. Uncertain pesticide applications to the catchment were estimated using land over-analysis of satellite data, combined with a Poisson distribution to describe the timing of application. Model performance for both the hydrograph (after calibration of the water balance) and the chemograph was good and could be improved via some minor adjustments in assumptions which yield general insights into the drivers for pesticide transport. The use of remote sensing offers some promising opportunities for estimating catchment-scale pesticide applications and associated losses.

Greenhouse gas (GHG) fluxes from livestock grazed pasture soils are highly variable in both space and time but the quantitative importance of the factors regulating this variation remain poorly understood. Our aim was to explore this variability on contrasting extensively (low input) and intensively managed sheep-grazed ‘case-study’ pastures. We quantified (through standard and spatially-informed regressions) the statistical relationships between GHG fluxes (nitrous oxide (N2O), carbon dioxide (CO2) and methane (CH4)) and a range of soil, field and management characteristics. Fluxes of these three GHGs at two study sites were highly variable, but spatial structure (i.e. autocorrelation) was only observed in the variability of N2O fluxes across the intensive site and CO2 fluxes across the extensive site. The regression analyses identified significant GHG predictor variables for the extensive site as: NO3- (p < 0.001) and vegetation-type (p < 0.01) for N2O (R2 = 0.57; p = 0.000); NH4+ (p < 0.05), slope (p < 0.05) and elevation (p < 0.01) for CO2 (R2 = 0.34; p = 0.000); and NO3- (p < 0.01), NH4+ (p < 0.05) and soil moisture (p < 0.05) for CH4 (R2 = 0.25; p = 0.005). Significant GHG predictor variables for the intensive site were soil moisture (p < 0.01) and bulk density (p < 0.01) for N2O (R2 = 0.27; p = 0.005); soil moisture (p < 0.001) for CO2 (R2 = 0.31; p = 0.001); while none were found for CH4 (R2 = 0.10; p = 0.655). Key factors driving GHG variation were both site- and GHG-specific, with fluxes controlled by local conditions leading to differences in limiting factors (possibly even at the within-site scale). Our statistical analyses suggest a larger range of driving variables (e.g. air and soil temperature or other soil chemical properties such as total extractable N) may be required to more fully capture the observed variability in the GHG processes considered here, and that it may also be fruitful for future analyses to consider non-linear, non-stationary and interacting relationships across space- and time-scales. Adequacies of each site’s sample design also played a key interpretive role in the GHG processes, requiring further evaluation through additional sampling campaigns.

Multi-media fate and transport models (MFTMs) are invaluable tools in understanding and predicting the likely behaviour of organic pollutants in the environment. However, some parameters describing the properties of both the environmental system and the chemical pollutant under consideration are uncertain and or variable in space and time. Furthermore, model performance is often evaluated using sparse data sets on chemical concentrations in different media. This can result in equifinality - the phenomenon in which several different combinations of model parameters can result in similar predictions of environmental concentrations. We explore this idea for MFTMs for the first time using, as examples, three cyclic volatile methyl siloxanes (cVMS: D4, D5 and D6) and the QWASI lake model applied to Tokyo Bay. Monte Carlo simulation was employed with parameters selected from probability distributions representing estimated uncertainty in a large number of iterations. This generated distributions of predicted chemical concentrations in water (CW) and sediment (CS) which represent the aleatory uncertainty envelope but which also demonstrate significant equifinality. For all three compounds, the uncertainty implied in the CW was lower (coefficient of variation, CV, of the order of 20%) than for CS (CV ca. 45%), reflecting the propensity of cVMS compounds to sorb to sediment and the sensitivity of the model to KOC. Confidence intervals were particularly high for the persistence of D5 and D6 in sediment which both ranged between approximately 1.7 years and approximately 26 years for Tokyo Bay. Predicted concentration distributions matched observations well for D5 and D6 not for D4. Equifinality could be reduced by better constraining acceptable parameter sets using additional measured data from different environmental compartments.

UV-visible spectroscopy has been shown to be a useful technique for determining dissolved organic carbon (DOC) concentrations. However, at present we are unaware of any studies in the literature that have investigated the suitability of this approach for tropical DOC water samples from any tropical peatlands, although some work has been performed in other tropical environments. We used water samples from two oil palm estates in Sarawak, Malaysia to: i) investigate the suitability of both single and two-wavelength proxies for tropical DOC determination; ii) develop a calibration dataset and set of parameters to calculate DOC concentrations indirectly; iii) provide tropical researchers with guidance on the best spectrophotometric approaches to use in future analyses of DOC. Both single and two-wavelength model approaches performed well with no one model significantly outperforming the other. The predictive ability of the models suggests that UV-visible spectroscopy is both a viable and low cost method for rapidly analyzing DOC in water samples immediately post-collection, which can be important when working at remote field sites with access to only basic laboratory facilities.

Ruminant urine patches are potential sites of reactive nitrogen (N) loss to the environment. Quantification of N losses from grazed grasslands requires measurement of the frequency of urine deposition, as well as its volume and chemical composition. However, studies to date are typically restricted to analyses of few replicate animals and urination events, especially for sheep. Here, we present data on urine frequency, volume, chemical composition (n = 193 events from n = 6 sheep) and metabolomic profile (n = 4–5 events from n = 4–5 sheep) from penned sheep. Differences in urine parameters and chemical composition data were compared seasonally and between two sites (improved and semi-improved pasture). Sheep urinated 8–11 times d−1, assuming time within pens represented a 24 h period. The mean urine event volume recorded was 289 ± 14 mL, from which we estimated a daily urine production value of 2.77 ± 0.15 L urine sheep−1 d−1. Daily urine N excretion and individual urine N concentrations were greater from sheep in improved pasture (26.7 ± 2.3 g N sheep−1 d−1; 7.0 ± 0.2 g N L−1) compared to those in semi-improved pasture (16.7 ± 1.2 g N sheep−1 d−1; 5.5 ± 0.4 g N L−1), but this did not equate to greater individual urine patch N loadings due to site differences in the urine-to-soil surface area influenced (17.5 L m-2 at the semi-improved site and 8.9 L m-2 at the improved site). Urine chemical composition varied seasonally and by site. Site- and season-specific urine should, therefore, be used in studies assessing N losses from urine patches. Based on the urine chemical composition data, we provide an updated artificial sheep urine ‘recipe’ which could be utilised to replicate natural sheep urine. The urine metabolomic profile clustered according to pasture quality, while clustering according to season was less evident. Our results provide important information for experimental and modelling studies assessing the scale and nature of N pollution arising from sheep-grazed pastures.

Nitrogen behaviour in domestic wastewater was investigated to develop a mechanistic understanding of ammonia dynamics in Floating Treatment Wetlands (FTWs) and to improve design and operation. A critical FTW design (full coverage of the floating mat and maximum plant density) was evaluated to optimize treatment performance of a pilot-scale system in removing ammonia under different operational volumes. A system dynamics approach as a framework was employed for robust understanding of the treatment mechanisms and system performance in removing ammonia. The results revealed that ammonia removal was enhanced in FTWs and the magnitude of removal was controlled by the design examined. Findings suggest that a design code of full coverage of water surface with mat material, high plant density, and low water volume can be considered as a critical design for FTW system to remove ammonia from domestic wastewater. This design promoted nitrification as principal ammonia removal process even when plants were present. The contribution of nitrification to overall ammonia removal was estimated to be between 81 and 85%. Plant uptake contributed to 14-19% of the total N loss. Ammonia loss via volatilization was determined to be negligible in examined system. Kinetics parameters for ammonia removal were between 0.14-0.22 day-1 in FTWs compared to 0.01-0.02 day-1 in controls. Understanding the contribution of removal processes of N could be useful to improve treatment system design, and enhance treatment performance.

Intact tropical peatlands are dense long-term stores of carbon. However, the future security of these ecosystems is at risk from land conversion and extensive peatland drainage. This can enhance peat oxidation and convert long-term carbon sinks into significant carbon sources. In Southeast Asia, the largest land use on peatland is for oil palm plantation agriculture. Here, we present the first annual estimate of exported fluvial organic carbon in the drainage waters of four peatland oil palm plantation areas in Sarawak, Malaysia. Total organic carbon (TOC) fluxes from the plantation second- and third-order drains were dominated (91 %) by dissolved organic carbon (DOC) and ranged from 34.4 ± 9.7 C m−2 yr−1 to 57.7 %, 16.3 g C m−2 yr−1 (± 95 % confidence interval). These fluxes represent a single-year survey which was strongly influenced by an El Ninõ event and therefore lower discharge than usual was observed. The magnitude of the flux was found to be influenced by water table depth, with higher TOC fluxes observed from more deeply drained sites. Radiocarbon dating on the DOC component indicated the presence of old (pre-1950s) carbon in all samples collected, with DOC at the most deeply drained site having a mean age of 735 years. Overall, our estimates suggest fluvial TOC contributes ∼ 5 % of total carbon losses from oil palm plantations on peat. Maintenance of high and stable water tables in oil palm plantations appears to be key to minimising TOC losses. This reinforces the importance of considering all carbon loss pathways, rather than just CO2 emissions from the peat surface, in studies of tropical peatland land conversion.

Rising nitrate concentrations in rivers and groundwater in regions with intensive agriculture have been of concern for several decades. Long records of nitrate concentration are rare; in the UK, few pre-date 1974. Records from the River Stour from 1937 are analysed using "moving-windows", to explore the effects of window width on trend detection and process inference. Nitrate concentrations rose sharply from the late 1950s in response to widespread ploughing of grassland, but have fallen since the early 1980s, suggesting that widespread nitrogen control measures, which were introduced after the start of the downward trend, may not be the principal driver for recent change. Short windows (5 years) are too noisy; longer windows (10, 15 and 30 years) identify signals (i.e. trends) more reliably. Since system response times can be of the order of 20–30 years, management decisions made on shorter data sets (even up to 15 years) could be misleading.

Summary This paper investigates the use of expert knowledge as a resource for digital soil mapping. To do this, three models of topsoil soil bulk density (Db) were produced: (i) a random forest model formulated and cross-validated with the limited data available (which served as the benchmark), (ii) a naive Bayesian network (BN) where the conditional probabilities that define the relations between Db and explanatory landscape variables were derived from expert knowledge rather than data and (iii) a ‘hierarchical’ BN where model structure was also defined by expert knowledge. These models were used to generate spatial predictions for mapping topsoil Db at a landscape scale. The results show that expert knowledge-based models can identify the same spatial trends in soil properties at a landscape scale as state-of-the-art mapping algorithms. This means that they are a viable option for soil mapping applications in areas that have limited empirical data.

A dynamic multi-media model that includes temperature-dependency for partitioning and degradation was developed to predict the behaviour of petroleum hydrocarbons during biopiling at low temperature. The activation energy (Ea) for degradation was derived by fitting the Arrhenius equation to hydrocarbon concentrations from temperature-controlled soil mesocosms contaminated with crude oil and diesel. The model was then applied to field-scale biopiles containing soil contaminated with diesel and kerosene at Casey Station, Antarctica. Temporal changes of total petroleum hydrocarbons (TPH) concentrations were very well described and predictions for individual hydrocarbon fractions were generally acceptable (disparity between measured and predicted concentrations was less than a factor two for most fractions). Biodegradation was predicted to be the dominant loss mechanism for all but the lightest aliphatic fractions, for which volatilisation was most important. Summertime losses were significant, resulting in TPH concentrations which were about 25% of initial concentrations just 1 year after the start of treatment. This contrasts with the slow rates often reported for hydrocarbons in situ and suggests that relatively simple remediation techniques can be effective even in Antarctica.

Cyclic volatile methylsiloxanes (cVMS) are used in personal care products and emitted to aquatic environments through wastewater effluents, and their bioaccumulation potential is debated. Here, a new bentho-pelagic version of the ACC-HUMAN model was evaluated for polychlorinated biphenyls (PCBs) and applied to cVMS in combination with measurements to explore their bioaccumulation behavior in a subarctic lake. Predictions agreed better with measured PCB concentrations in Arctic char (Salvelinus alpinus) and brown trout (Salmo trutta) when the benthic link was included than in the pelagic-only model. Measured concentrations of decamethylcyclopentasiloxane (D5) were 60 ± 1.2 (Chironomidae larvae), 107 ± 4.5 (pea clams Pisidium sp.), 131 ± 105 (three-spined sticklebacks: Gasterosteus aculeatus), 41 ± 38 (char), and 9.9 ± 5.9 (trout) ng g-1 wet weight. Concentrations were lower for octamethylcyclotetrasiloxane (D4) and dodecamethylcyclohexasiloxane (D6), and none of the cVMS displayed trophic magnification. Predicted cVMS concentrations were lower than measured in benthos, but agreed well with measurements in fish. cVMS removal through ventilation was an important predicted loss mechanism for the benthic-feeding fish. Predictions were highly sensitive to the partition coefficient between organic carbon and water (KOC) and its temperature dependence, as this controlled bioavailability for benthos (the main source of cVMS for fish). This is an open access article published under an ACS AuthorChoice License, which permits copying and redistribution of the article or any adaptations for non-commercial purposes.

Pesticide pollution of surface water is a major concern in many agricultural catchments The development of rapid and accurate methods for determining pesticide concentrations in water samples is, therefore, important. Here we describe a method for the simultaneous analysis of six pesticides (metaldehyde, quinmerac, carbetamide, metazachlor, propyzamide and pendimethalin) in natural waters by direct aqueous injection with liquid chromatography-tandem mass spectrometry. The method validation showed good linearity from 0.2 to 50.0 µg/L with correlation coefficients between 0.995 and 0.999. Method accuracy ranged from 84 to 100% and precision Relative standard deviation (RSD) from 4 to 15%. The limits of detection for the targeted pesticides ranged from 0.03 to 0.36 µg/L. No significant matrix effects on quantification were observed (t-test). The method was tested on water samples from a small arable catchment in eastern England. Peak concentrations for the determinands ranged from 1 to 10 µg/L.

Phosphorus (P) is one of the most limiting macronutrients for crop productivity and P deficiency is a common phenomenon in agricultural soils worldwide. Despite long-term application of phosphate fertilizers to increase crop yields, P availability is often low, due to the high affinity of phosphate for the soil solid phase. It has been suggested that the accumulated (surplus) P in agricultural soils is sufficient to sustain crop yields worldwide for about 100years. In this paper, we try to clear up the potential for making use of legacy P in soils for crop growth potentially alleviating the global P resource shortage. Specifically, we try to clear up the potential of soil "P activators" for releasing fixed P. P activators accelerate and strengthen process which transform P into bio-available forms via a range of chemical reactions and biological interactions. They include phosphate solubilizing microorganisms, phosphatase enzymes and enzyme activators, low molecular weight organic acids, humic acids, lignin, crop residues, biochar and zeolites. Although reported performance is variable, there is growing evidence that P activators can promote the release of phosphate from soil and, hence, have potential for mitigating the impending global P crisis. Further basic and applied research is required to better understand the mechanisms of interaction of P activators with natural soils and to maximize activator efficacy.

Water suppliers in parts of Europe currently face occasional Drinking Water Directive compliance challenges for a number of pesticide active substances including metal- dehyde, clopyralid and propyzamide. Water Framework Directive (WFD) Article 7 promotes a prevention-led (catchment management) approach to such issues. At the same time, European pesticide legislation is driving reduced active substance availability. In this context, embedding agronomic drivers of pesticide use into catchment management and regulatory decision making processes can help to ensure that water quality problems are addressed at source without imposition of disproportionate cost on either agriculture or potable water suppliers. In this study agronomist knowledge, perception and expectations of current and possible future pesticide use was assessed and the significance of this knowledge to other stakeholders involved with pesticide catchment management was evaluated. This was then used to provide insight into the possible impacts of active substance restrictions and associated adaptation options. For many arable crops, further restrictions on the range of pesticides available may cause increased use of alternatives (with potential for "pollution swapping"). However, in many cases alternatives are not available, too costly or lack a proven track record and other adaptation options may be selected which catchment managers need to be able to anticipate.

The influence of biochar (5%) on the loss, partitioning and bioaccessibility of (14)C-isoproturon ((14)C-IPU) was evaluated. Results indicated that biochar had a dramatic effect upon (14)C-IPU partitioning: (14)C-IPU extractability (0.01 M CaCl2) in biochar-amended treatments was reduced to2% while, (14)C-IPU extractability in biochar free treatments decreased with ageing from 90% to 40%. A partitioning model was constructed to derive an effective partition coefficient for biochar:water (KBW of 7.82 × 10(4) L kg(-1)). This was two orders of magnitude greater than the apparent Kfoc value of the soil organic carbon:water (631 L kg(-1)). (14)C-radiorespirometry assays indicated high competence of microorganisms to mineralise (14)C-IPU in the absence of biochar (40.3 ± 0.9%). Where biochar was present (14)C-IPU mineralisation never exceeded 2%. These results indicate reduced herbicide bioaccessibility. Increasing IPU application to ×10 its recommended dose was ineffective at redressing IPU sequestration and its low bioaccessibility.

The marine fate and pelagic food chain transfer of three cyclic volatile methyl siloxanes (cVMS: D4, D5 and D6) was explored in the Inner Oslofjord, Norway, using two dynamic models (the Oslofjord POP Model and the aquatic component of ACC-HUMAN). Predicted concentrations of D4, D5, and D6 in the water column were all less than current analytical detection limits, as was the predicted concentration of D4 in sediment (in agreement with measured data). The concentrations predicted for D5 and D6 in sediment were also in broad agreement with measured concentrations from the Inner Oslofjord. Volatilisation was predicted to be the most important loss mechanism for D5 and D6, whereas hydrolysis was predicted to dominate for D4. Concentrations of all three compounds in sediment are controlled by burial below the active mixed sediment layer. The marine food web model in ACC-HUMAN predicted “trophic dilution” of lipid-normalised cVMS concentrations between zooplankton and herring (Culpea harengus) and between herring and cod (Gadus morhua), principally due to a combination of in-fish metabolism and reduced gut absorption efficiency (as a consequence of high KOW). Predicted D5 concentrations in herring and cod agree well with measured data from the inner fjord, particularly when measured concentrations in zooplankton were used to set the initial dissolved-phase aqueous concentrations. Predicted concentrations of D4 and D6 in fish were over- and under-estimated by the model – possibly due to extrapolation of the metabolism rate constant from D5.

This work was conducted as part of the Irish Soil Information System (ISIS) Project, managed by Teagasc (the Irish Agriculture and Food Development Authority) and co-funded by the Environmental Protection Agency (EPA) of Ireland through their Science, Technology, Research and Innovation for the Environment (STRIVE) Programme, as part of the National Development Plan 2007–2013 peer-reviewed Soil bulk density (Db) is a major contributor to uncertainties in landscape-scale carbon and nutrient stock estimation. However, it is time consuming to measure and is, therefore, frequently predicted using surrogate variables, such as soil texture. Using this approach is of limited value for estimating landscape-scale inventories, as its accuracy beyond the sampling point at which texture is measured becomes highly uncertain. In this paper, we explore the ability of soil landscape models to predict soil Db using a suite of landscape attributes and derivatives for both topsoil and subsoil. The models were constructed using random forests and artificial neural networks. Using these statistical methods, we have produced a spatially distributed prediction of Db on a 100 m × 100 m grid, which was shown to significantly improve topsoil carbon stock estimation. In comparison to using mean values from point measurements, stratified by soil class, we found that the gridded method predicted Db more accurately, especially for higher and lower values within the range. Within our study area of the Midlands, UK, we found that the gridded prediction of Db produced a stock inventory of over 1 million tonnes of carbon greater than the stratified mean method. Furthermore, the 95% confidence interval associated with total C stock prediction was almost halved by using the gridded method. The gridded approach was particularly useful in improving organic carbon (OC) stock estimation for fine-scale landscape units at which many landscape–atmosphere interaction models operate. This work was conducted as part of the Irish Soil Information System (ISIS) Project, managed by Teagasc (the Irish Agriculture and Food Development Authority) and co-funded by the Environmental Protection Agency (EPA) of Ireland through their Science, Technology, Research and Innovation for the Environment (STRIVE) Programme, as part of the National Development Plan 2007–2013

The behaviour of cyclic volatile methyl siloxanes (cVMS) in lakes was explored using a fugacity-based steady-state non-equilibrium multimedia fate and transport model (a modified version of QWASI). Three substances were investigated: Octamethylcyclotetrasiloxane (D4), Decamethylcyclopentasiloxane (D5) and Dodecamethylcyclohexasiloxane (D6) in two contrasting North American lakes: Lake Ontario and Lake Pepin (a natural shallow lake on the Mississippi river between Minnesota and Wisconsin). Values for the principal partition coefficients and the hydrolysis rate constant were adjusted for the mean annual temperatures of each lake. Hydrolysis rate constants were also adjusted for the fraction of chemical calculated to be in the freely dissolved phase. Half-lives in sediment were calculated from partitioning theory, assuming hydrolysis can occur only in the dissolved phase. Best estimates of substance-specific emissions were obtained by combining current per capita approximations of usage and fraction lost to domestic waste water, the population of the lake catchment and cVMS removal in waste water treatment. Concentrations were generally lower and chemical residence times longer in Lake Ontario than in Lake Pepin, owing to greater depth, a higher hydraulic retention time and a higher degree of dilution. Overall persistence in Lake Pepin is significantly influenced by the high rate of sediment burial assumed in the model, as well as by a relatively high rate of water discharge. Despite the many similarities of the compounds considered, the dominant loss mechanisms vary significantly and are not the same in each lake system. This highlights the pitfalls of subjective evaluation of chemical fate and illustrates the important role which models have to play in providing a quantitative framework for assessing chemical behaviour objectively under the influence of a complex and interacting set of factors.

The adsorption isotherm and the mechanism of the buffering effect are important controls on phosphorus (P) behaviors in estuaries and are important for estimating phosphate concentrations in aquatic environments. In this paper, we derive phosphate adsorption isotherms in order to investigate sediment adsorption and buffering capacity for phosphorus discharged from sewage outfalls in the Yangtze Estuary and Hangzhou Bay near Shanghai, China. Experiments were also carried out at different temperatures in order to explore the buffering effects for phosphate. The results show that P sorption in sediments with low fine particle fractions was best described using exponential equations. Some P interactions between water and sediment may be caused by the precipitation of CaHPO4 from Ca2&plus; and HPO42− when the phosphate concentration in the liquid phase is high. Results from the buffering experiments suggest that the Zero Equilibrium Phosphate Concentrations (EPC0) vary from 0.014 mg L−1 to 0.061 mg L−1, which are consistent with measured phosphate concentrations in water samples collected at the same time as sediment sampling. Values of EPC0 and linear sorption coefficients (K) in sediments with high fine particle and organic matter contents are relatively high, which implies that they have high buffering capacity. Both EPC0 and K increase with increasing temperature, indicating a higher P buffering capacity at high temperatures.

An ultra performance liquid chromatography method coupled to a triple quadrupole mass spectrometer was developed to determine nonylphenol and 15 of its possible precursors (nonylphenol ethoxylates and nonylphenol carboxylates) in aqueous and particulate wastewater matrices. Final effluent method detection limits for all compounds ranged from 1.4 to 17.4 ng l(-1) in aqueous phases and from 1.4 to 39.4 ng g(-1) in particulate phases of samples. The method was used to measure the performance of a trickling filter wastewater treatment works, which are not routinely monitored despite their extensive usage. Relatively good removals of nonylphenol were observed over the biological secondary treatment process, accounting for a 53 % reduction. However, only an 8 % reduction in total nonylphenolic compound load was observed. This was explained by a shortening in ethoxylate chain length which initiated production of shorter polyethoxylates ranging from 1 to 4 ethoxylate units in length in final effluents. Modelling the possible impact of trickling filter discharge demonstrated that the nonylphenol environmental quality standard may be exceeded in receiving waters with low dilution ratios. In addition, there is a possibility that the EQS can be exceeded several kilometres downstream of the mixing zone due to the biotransformation of readily degradable short-chained precursors. This accentuates the need to monitor 'non-priority' parent compounds in wastewater treatment works since monitoring nonylphenol alone can give a false indication of process performance. It is thus recommended that future process performance monitoring and optimisation is undertaken using the full suite of nonylphenolic moieties which this method can facilitate.

Pesticide losses to water can present problems for environmental management, particularly in catchments where surface waters are abstracted for drinking water supply. The relative role of different transfer pathways (spray drift, spills, overland flow and leaching from soils) is often uncertain, and there is a need for experimental observation and modelling to ensure that processes are understood under a range of conditions. Here we examine the transport of propyzamide and carbetamide in a small (15.5 ha) headwater sub-catchment dominated by an artificially drained field with strongly undulating topography (topographic gradients >1:10). Specifically, we explore the validity of the "field-scale lysimeter" analogy by applying the one dimensional mathematical model MACRO. Although one dimensional representation has been shown to be reasonable elsewhere, the scale and topography of the monitored system challenge many of the underlying assumptions. MACRO considers two interacting flow domains: micropores and macropores. The effect of subsurface drains can also be included. A component of the outflow from the main drain was identified as originating from an upslope permeable shallow aquifer which was represented using a simple groundwater model. Predicted herbicide losses were sensitive to drain spacing and the organic carbon to water partition coefficient, K(OC). The magnitude of the peak water and herbicide transport and their timing were simulated satisfactorily, although model performance was poor following a period of one month when snow covered the ground and precipitation was underestimated by the rain gauge. Total herbicide loads were simulated adequately by MACRO, suggesting that the field-scale lysimeter analogy is valid at this scale, although baseflow contributions to flow needed to be accounted for separately in order to adequately represent hydrological response.

Article 7 of the European Water Framework Directive (WFD) promotes a prevention-led approach to European Drinking Water Directive (DWD) compliance for those parameters that derive from anthropogenic influences on raw water quality. However, the efficacy of pollution prevention interventions is currently uncertain and likely to be variable, which makes absolute compliance with the drinking water standard a significant challenge. Member State governments, the WFD competent authority, the DWD competent authority, water suppliers and agriculture are all affected by and have a different perspective on the nature of this challenge. This paper presents a discussion of these perspectives applicable to stakeholders in all European Member States; the analysis is supported with examples from England and Wales. Improved understanding of the challenges faced by each group is needed if these groups are to achieve the shared goals of WFD Article 7 compliance and DWD compliance without a disproportionately negative impact on agricultural productivity. In addition, the European Commission needs to be aware of and address a potential incompatibility between WFD Article 7 and the DWD. With this in mind, targeted recommendations for action are presented for each stakeholder group.

Summary This paper compares the fluvial flux of nitrate at the UK’s tidal limit with an estimate of losses from the terrestrial biosphere at the point of entry to the river system to estimate in-stream losses across the UK’s entire river network between 1925 and 2007. These estimates were used to support interpretation of the current trend in the fluvial flux of nitrogen from the UK, and to give an assessment of the fluvial component of the total nitrogen budget of UK. The release of nitrate to the point at which it enters the stream network is estimated, using a Monte-Carlo framework, from a description of annual soil N mineralisation and immobilisation combined with an export coefficient model. Long-term records of land use (including – agricultural, forestry and urban uses); livestock; human population and atmospheric deposition are used. The study shows that: (i) The flux of nitrate from UK soils in the period since 1925 varied between 420 and 1463 ktonnes N yr −1 , with two peaks, one in 1944 and one in 1967. The first of these peaks was caused by mineralisation of soil organic matter following radical land use change in the 1940s, and the second was a multifactorial response to land use change and agricultural intensification in the 1960s. (ii) The current trend in N release from soils appears to be downward but the current fluvial flux at the tidal limit is slowly increasing. It is possible that in the near future nitrate–N losses at the tidal limit will, in some years, exceed N losses from soils implying a net N gain into the fluvial network. This apparent gain can be explained by the breakthrough of high nitrate groundwater into surface waters. (iii) The predicted export from UK soils at source as water enters the stream network implies that the UK will soon become a net source of total reactive N species, as the capacity for in-stream removal becomes exhausted by the ingress of N-rich groundwater to the river system.

Phospholipid fatty acid (PLFA) analysis is an informative method for characterising and quantifying changes in the phenotypic profile of the soil microbial community when soils are exposed to chemical toxicants and other xenobiotics. However, where such materials are lipophilic, a range of non-polar compounds can be co-extracted with PLFAs and can consequently mask PLFA chromatograms. We found this to be the case with the lipophilic anti-microbial compound triclosan, which can enter the soil via the addition of sewage sludge. A simple method of washing soil in solvent prior to extraction was developed in order to remove triclosan without altering the relative abundance of PLFAs. Three contrasting soils were spiked with 500 mg kg −1 of triclosan before being washed with methanol (MeOH), dichloromethane (DCM), hexane or aqueous solutions of these solvents. PLFAs were then extracted and analysed. All treatments were found to remove triclosan effectively, allowing all peaks to be identifiable. Whilst the polar solvents MeOH and DCM significantly altered the relative abundance of extracted fatty acids in most of the soils tested, soil washing with a small quantity of hexane was able to remove triclosan whilst best preserving the fidelity of the PLFA profiles.

Summary A model of catchment-scale nitrate transport is presented for a small, rural headwater basin (Alton Pancras: 2 ) in Dorset, UK, for the period 1930–2007. Estimates of annual nitrogen (N) loading were based on parish land-use data, held in the UK’s National Archives, and previously reported figures of typical UK N loadings from livestock, fertiliser, ploughing of permanent pasture, atmospheric deposition, biological fixation and crop uptake. Loading calculations were performed within an uncertainty framework to allow for the reliance on literature data sources. Loading calculations show that all significant sources must be included not just fertiliser application which, at most, contributes 50% of N input in any given year. A simple algorithm was used to transform estimated catchment N loading (1930–2007) into a river nitrate response (observed data: 1980–2004). This assumed N-loads were delayed by some catchment mean travel time (MTT), t a , attenuated according to a Peclet number, P e , converted into solute concentrations by a factor, α , to increase some initial baseline river concentration, C b . Simple graphical translation of estimated catchment N loading to the river concentration data suggested a MTT of around 37 years. As P e → ∞ , the transport model simplified to a linear relationship between catchment N load and river nitrate concentration response lagged by the MTT. Hence, the model results suggest that, in this catchment, advection is the dominant mechanism for transport of diffuse pollution from land to river: there is little or no dispersion present. The MTT ( t a ) was then reconsidered using an estimated distribution of unsaturated zone depths in the Alton Pancras catchment. Conclusions suggest that, in modelling of long-term nutrient transport, a detailed source term is of much greater importance than a complex hydrogeological model. Implications of epistemic uncertainty, long-term prediction and management of diffuse agricultural pollution are discussed.

Anaerobic digestion (AD) is becoming increasingly popular for treating organic waste. The methane produced can be burned to generate electricity and the digestate, which is high in mineral nitrogen, can be used as a fertiliser. In this paper we evaluate potential losses of ammonia via volatilisation from food waste anaerobic digestate using a closed chamber system equipped with a sulphuric acid trap. Ammonia losses represent a pollution source and, over long periods could reduce the agronomic value of the digestate. Observed ammonia losses from the experimental system were linear with time. A simple non-steady-state partitioning model was developed to represent the process. After calibration, the model was able to describe the behaviour of ammonia in the digestate and in the trap very well. The average rate of volatilisation was approximately 5.2 g N m −2 week −1 . The model was used to extrapolate the findings of the laboratory study to a number of AD storage scenarios. The simulations highlight that open storage of digestate could result in significant losses of ammonia to the atmosphere. Losses are predicted to be relatively minor from covered facilities, particularly if depth to surface area ratio is high.

Higher-tier environmental risk assessments on "down-the-drain" chemicals in river networks can be conducted using models such as GREAT-ER (Geography-referenced Regional Exposure Assessment Tool for European Rivers). It is important these models are evaluated and their sensitivities to input variables understood. This study had two primary objectives: evaluate GREAT-ER model performance, comparing simulated modelled predictions for LAS (linear alkylbenzene sulphonate) with measured concentrations, for four rivers in the UK, and investigate model sensitivity to input variables. We demonstrate that the GREAT-ER model is very sensitive to variability in river discharges. However it is insensitive to the form of distributions used to describe chemical usage and removal rate in sewage treatment plants (STPs). It is concluded that more effort should be directed towards improving empirical estimates of effluent load and reducing uncertainty associated with usage and removal rates in STPs. Simulations could be improved by incorporating the effect of river depth on dissipation rates.

In many parts of the world clothes are washed near to or in rivers and streams. Little information is available on resulting concentrations of detergent ingredients or on any potential effects caused. In this study, the fate of a commonly used anionic surfactant, linear alkylbenzene sulphonate (LAS) was investigated in a reach of the Balfour River (Eastern Cape Province, South Africa) which was regularly used as a site for laundry activity. Samples of river water were collected upstream of the main washing site and at a number of locations downstream on several occasions in winter and summer. Sediment samples were also collected and analysed. In addition, a household survey was conducted to ascertain the amount of detergent used and the distribution of washing practices. The results of the survey suggested that the use of riverside locations for laundry activities was seasonal. Most washing tended to be done at home during the winter with riverside sites used more frequently during the summer months. The monitoring data showed that LAS concentrations in water were very variable. They were occasionally high in the immediate vicinity of the laundry site (up to 342 microg L(-1)) but were generally very low (

It is useful to have available a variety of catchment-scale water quality models that range in complexity, spatial resolution and data requirements. In a previous paper [Warren, C., Mackay, D., Whelan, M., Fox, K., 2005. Mass balance modelling of contaminants in river basins: a flexible matrix approach. Chemosphere 61, 1458-1467] a series of simple to intermediately complex mass balance models was presented which can be used for tiered exposure assessments in river basins. The connectivity of the segments is expressed using a matrix that permits flexibility in application, enabling the model to be re-segmented and applied to different catchments as required. In this paper, the intermediate models, QWASI matrix-rate constant (QMX-R) and QWASI matrix-fugacity (QMX-F) are used to estimate concentrations of linear alkylbenzene sulfonates (LAS) in the rivers Aire and Calder, UK, and of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) in the Fraser River basin, Canada. The results compare satisfactorily with monitoring data, suggesting that these QWASI-based models for exposure and risk assessment may be applicable under data-limited conditions. The use of QWASI-based models for regulatory purposes in an evaluative river system is also discussed with reference to assessments of para-dichlorobenzene (pDCB), trichloroethylene (TCE), bis(2-ethylhexyl) phthalate (DEHP) and toluene. It is shown that multi-media QWASI model predictions can be usefully depicted graphically on chemical space diagrams and used to highlight regions in which advection, partitioning to sediments and volatilization may be important determinants of chemical fate in river systems.

Summary The performance of the SWAT-2000 model was evaluated using stream flow at the outlet of the 142 ha Colworth catchment (Bedfordshire, UK). This catchment has been monitored since October 1999. The soil type consists of clay loam soil over stony calcareous clay and a rotation of wheat, oil seed rape, grass, beans and peas is grown. Much of the catchment is tile drained. Acceptable performance in hydrological modelling, along with correct simulation of the processes driving the water balance were essential first requirements for predicting contaminant transport. Initial results from SWAT-2000 identified some necessary modifications in the model source code for correct simulation of processes driving water balance. After modification of the code, hydrological simulation, crop growth and evapotranspiration (ET) patterns were realistic when compared with empirical data. Acceptable model performance (based on a number of error measures) was obtained in final model runs, with reasonable runoff partitioning into overland flow, tile drainage and base flow.

Summary Distributed models used in hydrological modelling, have many parameters. To get useful results from the model, every parameter is required to have a sensible value. Usually a calibration is undertaken to reduce the uncertainties associated with the estimation of model parameters. To ensure efficient calibration, a sensitivity analysis is conducted to identify the most sensitive parameters. This paper describes simple and efficient approaches for sensitivity analysis, calibration and identification of the best methodology within a modelling framework. For this study, the SWAT-2000 model was used on a small catchment of 141.5 ha in the Unilever Colworth estate, in Bedfordshire, England. Acceptable performance in hydrological modelling, and correct simulation of the processes driving the water balance were essential requirements for subsequent pesticide modelling. SWAT gives various options for both evapotranspiration and runoff modelling. Identification of the best modelling option for these processes is a pre-requisite to achieve these requirements. As a first step, a sensitivity analysis was conducted to identify the sensitive parameters affecting stream flow for subsequent application in stream flow calibration. Hydrological modelling has been carried out for the catchment for the period September 1999 to May 2002 inclusive using both daily and sub-daily rainfall data. The Hargreaves and Penman-Montieth methods of evapotranspiration estimation and the NRCS curve number (CN) and Green and Ampt infiltration methods for runoff estimation techniques were used, in four different combinations, to identify the combination of methodologies that best reproduced the observed data. In addition, as the initial calibration period, starting in September 1999, was substantially wetter than the following corresponding validation period, the calibration and validation periods are interchanged to test the impact of calibration using wet or dry periods.

When river water quality fluctuates over relatively short periods of time with respect to the sampling frequency, the collection of grab samples may be inappropriate for characterising average water quality. This paper presents the results of a water quality monitoring study carried out on a stretch of the river Lambro (northern Italy) dominated by a periodically overloaded sewage treatment works (STW) located near its upstream end. Water quality was strongly influenced by a pronounced diurnal cycle in pollutant loads caused by the regular emission of untreated waste water during periods of high domestic flow (daytime). Two different sampling techniques were employed: grab sampling and 24-h composite sampling using automatic samplers. Samples were collected at the plant overflow and at several sites along the river and analysed for two common ingredients of household detergents, linear alkylbenzene sulphonate (LAS) and boron (B) and for routine water quality variables. The results obtained show that: (1) The diurnal variability of point-source-derived chemical concentrations in the river downstream of the undersized STW increased with increasing removal efficiency in sewage treatment. (2) The shape of the diurnal concentration signal remained relatively intact for a considerable distance downstream of the STW for several water quality variables, suggesting that hydrodynamic dispersion plays a relatively minor role in controlling concentration patterns in this river. (3) In-stream degradation of LAS was consistent with first order kinetics with a rate constant of 0.05–0.06 h−1. (4) Grab sampling is a relatively inefficient methodology for capturing mean concentrations for rivers subjected to highly variable loads, especially when it is restricted to office hours. The inefficiency of grab sampling is more marked for substances (e.g. LAS) which are effectively removed during sewage treatment than for substances which are not. (5) For LAS, diurnal variability in the concentration signal decreases with distance downstream, making grab sampling an increasingly reliable methodology for estimating mean concentrations. (6) 24-h composite sampling is an efficient way of eliminating the effect of diurnal variations in load strength.

Pesticides in stream flow from the 142 ha Colworth catchment in Bedfordshire, UK were monitored from October 1999 to December 2000. About 47% of the catchment is tile-drained and different pesticides and cropping patterns have recently been evaluated in terms of their effect on nutrient and pesticide losses to the stream. The data from Colworth were used to test soil and water assessment tool (SWAT) 2000 predictions of pesticide concentrations at the catchment outlet. A sound model set-up to carry out pesticide modelling was created by means of hydrological modelling with proper simulation of crop growth and evapotranspiration. The pesticides terbuthylazine, terbutryn, cyanazine and bentazone were modelled. There was close agreement between SWAT-predicted pesticide concentration values and observations. Scenario trials were conducted to explore management options for reducing pesticide loads arriving at the catchment outlet. The results obtained indicate that SWAT can be used as a tool to understand pesticide behavior at the catchment scale.

GREAT-ER (Geography-referenced Regional Exposure Assessment Tool for European Rivers) is a model system for predicting chemical fate and exposure in surface waters. The GREAT-ER approach combines a series of well-studied models (for sewers, waste water treatment plants and rivers) with spatial information managed by a GIS. A new version of GREAT-ER (version 2) has been developed which has a number of improved features compared to its predecessor. All components are now implemented as Free Software and make use of Free Software where possible. This simplifies the distribution, use and extension of the system and its components for scientific applications. In addition, two new platform-independent user interfaces have been developed. The first is a simplified web-based interface for easy access without installation and the second is a full-featured desktop version. The model system has been re-designed to increase its flexibility for the modification and extension of its component models. In addition, the file-based data storage system used in version 1 has been replaced by a database management system with a flexible Application Programming Interface (API). This makes it easier for users to share their results, using a new repository (also with web-based and desktop versions) serving as a central communication tool. The output from the new system has been successfully tested against the field-validated models in GREAT-ER 1.0.

A novel and flexible approach is described for simulating the behaviour of chemicals in river basins. A number (n) of river reaches are defined and their connectivity is described by entries in an n x n matrix. Changes in segmentation can be readily accommodated by altering the matrix entries, without the need for model revision. Two models are described. The simpler QMX-R model only considers advection and an overall loss due to the combined processes of volatilization, net transfer to sediment and degradation. The rate constant for the overall loss is derived from fugacity calculations for a single segment system. The more rigorous QMX-F model performs fugacity calculations for each segment and explicitly includes the processes of advection, evaporation, water-sediment exchange and degradation in both water and sediment. In this way chemical exposure in all compartments (including equilibrium concentrations in biota) can be estimated. Both models are designed to serve as intermediate-complexity exposure assessment tools for river basins with relatively low data requirements. By considering the spatially explicit nature of emission sources and the changes in concentration which occur with transport in the channel system, the approach offers significant advantages over simple one-segment simulations while being more readily applicable than more sophisticated, highly segmented, GIS-based models.

Phosphorus (P) is a limiting nutrient in many freshwater ecosystems and increases in its availability can lead to eutrophication. Effective management of P in freshwaters requires quantitative estimates of P supply from all significant sources. A simple GIS-based model, capable of predicting total diffuse source phosphorus export from catchments using readily available data, has been developed. The model is based on the idea of export coefficients but includes the effects of topography (slope and cumulative area), soil type (using the UK Hydrology of Soil Types (HOST) classification) and climate (hydrologically effective rainfall) as well as land use. Uncertainty in key model parameters is accounted for using Monte Carlo simulation which involves random sampling from probability density functions in a large number of iterations. This reduces the need for subjective optimisation of export coefficients. The model has been applied to the Greens Burn catchment, Scotland and predicts P exports within the confidence limits of the measured values.

Two corresponding issues concerning Digital Soil Mapping are the demand for up-to-date, fine resolution soil data and the need to determine soil–landscape relationships. In this study, we propose a Bayesian Network framework as a suitable modelling approach to fulfil these requirements. Bayesian Networks are graphical probabilistic models in which predictions are obtained using prior probabilities derived from either measured data or expert opinion. They represent cause and effect relationships through connections in a network system. The advantage of the Bayesian Networks approach is that the models are easy to interpret and the uncertainty inherent in the relationships between variables can be expressed in terms of probability. In this study we will define the fundamentals of a Bayesian Network and the probability theory that underpins predictions. Then, using case studies, we demonstrate how they can be applied to predict soil properties (bulk density) and soil taxonomic class (associations).

This paper describes a simple model of phosphorus (P) transfer from agricultural land to surface waters which incorporates the effects of spatial variability in catchment properties and uncertainty in model parameter values. TOPMODEL concepts are used to estimate water, solute and sediment fluxes to water bodies. The model predicts the spatial distribution of water table depth and saturation-excess overland flow based on topography. Dissolved P (DP) transfer is assumed to occur vertically in the unsaturated zone and laterally in the saturated zone. Readily soluble P is assumed to decrease exponentially with soil depth. Particulate P (PP) transfers are modelled by estimating overland flow discharge and associated sediment transport capacity. Uncertainty in the distribution of soil surface P concentrations and model parameters controlling the mobility of soil P are incorporated using Monte Carlo simulation. Predicted losses of DP are well correlated with discharge and those of PP are episodic. Highest losses of P tended to be predicted near to the stream where the water table is close to the surface. The combination of a deterministic model core with a stochastic generation of model parameters or state variables provides an attractive way of embracing variability and uncertainty in models of this kind.

Rainfall-runoff modelling is a useful tool for water resources management. This study presents a simple daily rainfall-runoff model, based on the water balance equation, which we apply to the 11,630 km2 Lesser Zab catchment in northeast Iraq. The model was forced by either observed daily rain gauge data from four stations in the catchment or satellite-derived rainfall estimates from two TRMM Multi-satellite Precipitation Analysis (TMPA) data products (TMPA-3B42 and 3B42RT) based on the Tropical Rainfall Measuring Mission (TRMM) from 2003 to 2014. As well as using raw TMPA data, we used a bias-correction method to adjust TMPA values based on rain gauge data. The uncorrected TMPA data products underestimated observed mean catchment rainfall by −10.1% and −10.7%. Corrected data also slightly underestimated gauged rainfall by −0.7% and −1.6%, respectively. Nash-Sutcliffe Efficiency (NSE) and Pearson’s Correlation Coefficient (r) for the model fit with the observed hydrograph were 0.75 and 0.87, respectively, for a calibration period (2010–2011) using gauged rainfall data. Model validation performance (2012–2014) was best (highest NSE and r; lowest RMSE and bias) using the corrected 3B42 data product and poorest when driven by uncorrected 3B42RT data. Uncertainty and equifinality were also explored. Our results suggest that TRMM data can be used to drive rainfall-runoff modelling in semi-arid catchments, particularly when corrected using rain gauge data.

Neglecting hydrodynamic dispersion in models of river water quality has important implications in terms of simplifying the equations which describe the system and for model calibration. In this paper we analyze the behavior of a river system subjected to a time-variable point source input of a degradable solute using frequency response techniques applied to one-dimensional solute transport models, considering the combined effect of amplitude attenuation and phase shift due to hydrodynamic dispersion. We derive exact and simplified expressions that can be used for establishing whether dispersion is negligible at a desired level of accuracy. Finally, we present an application of the analysis to the river Lambro in northern Italy which receives diurnally varying solute loads from an undersized sewage treatment plant. The results obtained show the following: (1) In the case of a time-constant solute input, exact analytical expressions of the error incurred by neglecting dispersion can be derived which show that dispersive transport is negligible in most quasi steady state cases. (2) When solute input varies sinusoidally with time, the error due to neglecting dispersion is strongly dependent on the input frequency and may be significant if the input frequency is high. The error is practically independent of the decay rate coefficient, k, except when k is significantly greater than the input frequency. (3) Ignoring solute transport upstream from the outfall, due to dispersion, in the solution to the advection dispersion equation (ADE) can result in significant errors if the input frequency is sufficiently high. In some cases the choice of the ADE solution employed may have a larger impact than neglecting dispersion altogether.

The hydrological response of a catchment is the result of combined hillslope and channel network delays in the translation of rainfall to discharge at the catchment outlet. In this paper we examine three simple descriptions of the channel network response following an instantaneous hillslope runoff input within a general framework in which the role of network morphology is described using the probability density function (pdf) of network path lengths, and the role of hydrodynamic dispersion is described using a solution of the linear advection dispersion equation. We show that the pdf of path lengths, represented by the area function, can be adequately approximated by a triangular distribution with just two parameters (the mean and maximum network path lengths). The triangular pdf can be combined with a description of residence times in individual paths to derive an analytical expression for the hydrological response of the channel network. This expression provides a better approximation of the network response than both the original geomorphological instantaneous unit hydrograph (GIUH), based on Horton–Strahler ratios, and a single-channel representation, based on the concept of geomorphological dispersion. Using the same expression, we show that geomorphological factors generally prevail in controlling the mean and variance of network residence times. In contrast, hydrodynamic dispersion is shown to have a significant influence on the skewness of the network response over a wide range of catchment characteristics, strongly influencing the hydrograph peak and time to peak values in these cases. Copyright © 1999 John Wiley & Sons, Ltd.

A simple point-source water quality model was constructed to explore hypothetical scenarios of linear alkylbenzene sulphonate (LAS) load variability, chemical degradation kinetics, river discharge and river velocities within the context of planned water quality monitoring in the river Lambro, northern Italy. LAS is an anionic surfactant and a commonly used ingredient of household detergents. Input loads were derived from flow and concentration measurements at sewage treatment works (STW). River travel times are approximated using hydraulic geometry concepts based on existing records of stage, discharge and velocity measurements. The model proved to be especially useful in predicting the impact of overflows from an undersized STW near the top of the catchment on diurnal variations in water quality downstream. Probability density functions (pdf's) of LAS concentrations for a number of points on the river were approximated by running the model stochastically (using a Monte Carlo procedure) over independent frequency distributions of discharge, temperature and input load. The predicted concentration pdf's were then used to estimate the number of random grab and 24 h composite samples, which would be required in order to estimate mean concentrations with confidence. This is especially important in situations where the analysis is expensive and the sample budget is limited. The model results showed that grab sampling was not a viable strategy in the Lambro largely due to the very strong diurnal variability in concentrations induced by the STW overflow. This conclusion was corroborated by field data.

Spatial variability of throughfall depth and mineral ion composition was investigated in a Norway spruce (Picea abies (Karst) L.) plantation at the small plot scale. A regular array of 39 funnel collectors was set up on the forest floor and throughfall volume and concentrations of nitrate, ammonium, chloride, potassium and sulphate were measured on six sampling occasions over a period of 81 days in winter. Considerable spatial variability was found in the patterns of throughfall depth and ion concentrations but for each sampling period there were consistent features in the spatial patterns of these inputs. Throughfall depth increased significantly as a function of increasing distance from the nearest tree bole while ion load was negatively correlated with distance from the tree. The magnitude of the mean ion load for each period depended on mean throughfall depth. Throughfall depth and deposition of ions were also correlated with spatially referenced values of canopy interception capacity estimated using a deterministic model of the canopy hydrology above each collector. Although linear correlations between the canopy cover above each collector (estimated from image analysis of photographs) and the spatial patterns of throughfall volume, concentration and load were significant for only 17% of the time, least-squares fitting of quadratic equations (input vs cover with a maximum input at some intermediate value of cover) were significant for more than half of the cases examined. This study highlights the need for a better characterisation of canopy architecture in order to improve understanding of its role in affecting water and mineral ion deposition on the forest floor.

We investigated variations in the biochemical composition and biodegradation of Scots pine (Pinus sylvestris) needles and litter on four soils with different cation exchange characteristics, ranging from 7 to 92% base saturation. Trends in cellulose, non-crystalline holocellulose concentrations (as measured by a trifluoracetic acid (TFA) hydrolysis technique) and ratios of lignin-to-cellulose and lignin-to-N in needles and litters varied significantly between sampling sites and reflected trends in soil base status under P. sylvestris stands. Although these relationships need to be tested rigorously on a more extensive sample set, clearly the biochemical quality of needles and litter within the same tree species is not uniform. Detailed analysis of the carbohydrate composition of TFA extracts from plant tissues showed that the ratios of selected hexoses (predominantly of microbial origin) to pentoses (mainly from plant cell wall constituents) were between 1 and 15 times lower in decomposing litter than needles. The hexose-to-pentose ratios also increased from 0.117:1 to 0.136:1 (as rhamnose+fucose-to-arabinose+xylose) and from 1.32:1 to 1.67:1 (as galactose+mannose-to-arabinose+xylose) with increasing base saturation of soils. It is concluded that the former ratio most clearly differentiated between the contrasting sites, whereas the latter ratio was more diagnostic of biotransformations during litter decomposition and associated development of microbial biomass. The study highlights a need to further clarify intraspecific differences in tree leaf and litter biochemistry in relation to soil characteristics and challenges the conventional definition of litter quality in terms of C-to-N or lignin-to-N ratio without consideration for the biochemical quality and spatial patterns of a “holocellulose” fraction. The determination of constituent carbohydrates in TFA extracts may provide a useful diagnostic for the potential decomposition of litter and the depolymerisation of structural carbohydrates.

Regulations based on the precautionary principle should undertake a comprehensive assessment of all available scientific and technical data to identify sources of epistemic uncertainty. In the European Union (EU), environmental regulation is required to fulfill the principles established in Article 174 of the EU Treaty, such that it offers a high level of protection and is consistent with the precautionary principle. Pesticides in drinking water are currently regulated by the Drinking Water Directive using a maximum allowable concentration of 0.1 μg/L. This standard (a surrogate zero) was consistent with the precautionary principle when it was originally set in 1980 and remained consistent when retained in 1998. However, given developments in EU pesticide and water policy, international experience in regulating pesticides, and an increasing knowledge of pesticide toxicity, it can be argued that the level of epistemic uncertainty faced by regulators has substantially decreased. In this paper, we examine the extent to which such developments now challenge the basis of European drinking water standards for pesticides and whether, for substances for which there is good toxicological understanding, a regulatory approach based upon the World Health Organization (WHO) Guideline Value (GV) methodology would be more consistent with the principles underpinning European environmental policy.

The estimation of throughfall to forest systems is an important requirement for modelling the hydrology and chemistry of catchments and for understanding moisture dependent processes on the forest floor. Despite widespread recognition that throughfall is inherently variable, relatively few attempts have been made to account for any systematic influence canopy architecture may have over throughfall patterns. A simple Rutter-type interception model is developed here to include spatially distributed parameters which are related to variable canopy cover and interception capacity. Evaporation rates are approximated using the Penman-Monteith equation and canopy drainage is modelled using a negative exponential equation related to canopy wetness. The model was calibrated and validated using field observations of throughfall from a regular array of funnel collectors installed in a small plot in a Norway spruce ( Picea abies ) plantation in southwest England. The observed throughfall patterns had a strong systematic component which was strongly correlated with spatially distributed measured canopy cover and calibrated canopy capacity. The model was able to reproduce the broad trends in observed patterns of throughfall to the plot and was able to predict the cumulative total throughfall input to the plot over 77 days to within 2% of the measured input.

Intact Pinus sylvestris needles and litter from the soil surface were sampled from several stands in SW England growing on soils with different chemical characteristics. The intrinsic chemical and biochemical characteristics of needle and litter samples were measured. The total concentration of lignin, measured by proximate analysis, and the chemical characteristics of the lignin-polymer in P. sylvestris needles and litter, varied between soil types. Lignin concentrations increased from 180 to 316 mg litter g −1 with increasing soil base saturation, although no similar trends in the concentration of lignin in needles were observed. The total concentration of phenylpropanoid moieties, released on CuO-oxidation, increased from 65.6 to 96.1 mg lignin g −1 in needles and 35.3–83.6 mg lignin g −1 in litters, which followed a trend of increasing soil base saturation under P. sylvestris stands. The concentration of NaOH-extractable moieties in needles increased from 1.79 to 2.35 μg sample g −1 with soil base saturation, as did the acid to aldehyde ratios of phenylpropanoid moieties in NaOH-extracts from litters. We conclude that the quality of lignin in needles and litter, as measured by its phenylpropanoid characteristics, varied significantly in P. sylvestris stands. In the light of the evidence from other studies that phenylpropanoid moieties have significantly different rates of microbial decomposition, this aspect of lignin quality should be considered when modelling the rate of litter mass loss and organic matter accumulation in forest soils. Furthermore, we have shown that by combining the CuO-oxidation technique with a NaOH-extraction procedure, phenylpropanoid signatures of a ‘bound’ and a relatively ‘labile’ lignin derived fraction in needles and litters can be more accurately monitored. A more extensive study is required to clarify the apparent relationships between the soil chemical characteristics beneath stands and the quality of lignin in plant materials.