Modeling macropore flow effects on pesticide leaching: inverse parameter estimation using microlysimeters

Macropore flow is a key factor determining pesticide fate, but models accounting for this process need parameters that cannot be easily measured. This study was conducted to investigate the use of inverse techniques to estimate parameters controlling macropore flow and pesticide fate in the dual-permeability model MACRO. Undisturbed columns were sampled at three landscape positions (hilltop, slope, hollow) with contrasting texture and organic carbon content. Transient leaching experiments were performed for an anionic tracer and the herbicide MCPA (4-chloro-2methylphenoxy acetic acid) during a 4-mo period, first under natural rainfall, and then under controlled irrigation in the laboratory. The tracer breakthrough for the liner-textured soil from the hilltop showed strong evidence of macropore flow, resulting in a rapid leaching of MCPA, while leaching was minimal from the organic-rich hollow soil, since macropore flow was weaker and adsorption stronger. The MACRO model was linked to the inverse modeling program SUFI (Sequential Uncertainty Fitting) to enable calibration of nine key model parameters. Based on calculated model efficiencies, MACRO-SUFI gave generally good predictions of water movement and tracer and pesticide transport, although some errors were attributed to difficulties in simulating the effects of soil moisture on degradation and the timing of water outflows. Even after calibration, significant uncertainties remained for some key parameters controlling macropore flow. Nevertheless, the parameter estimates were significantly different between landscape positions and could also be related to basic soil properties. The posterior uncertainty ranges could probably be reduced with a more exhaustive sampling of the parameter space and improved experimental designs.     

Sensitivity and first-step uncertainty analyses for the preferential flow model MACRO

Sensitivity analyses for the preferential flow model MACRO were carried out using one-at-a-time and Monte Carlo sampling approaches. Four different scenarios were generated by simulating leaching to depth of two hypothetical pesticides in a sandy loam and a more structured clay loam soil. Sensitivity of the model was assessed using the predictions for accumulated water percolated at a 1-m depth and accumulated pesticide losses in percolation. Results for simulated percolation were similar for the two soils. Predictions of water volumes percolated were found to be only marginally affected by changes in input parameters and the most influential parameter was the water content defining the boundary between micropores and macropores in this dual-porosity model. In contrast, predictions of pesticide losses were found to be dependent on the scenarios considered and to be significantly affected by variations in input parameters. In most scenarios, predictions for pesticide losses by MACRO were most influenced by parameters related to sorption and degradation. Under specific circumstances, pesticide losses can be largely affected by changes in hydrological properties of the soil. Since parameters were varied within ranges that approximated their uncertainty, a first-step assessment of uncertainty for the predictions of pesticide losses was possible. Large uncertainties in the predictions were reported, although these are likely to have been overestimated by considering a large number of input parameters in the exercise. It appears desirable that a probabilistic framework accounting for uncertainty is integrated into the estimation of pesticide exposure for regulatory purposes.     

Predicting phosphorus losses with the PLEASE model on a local scale in Denmark and the Netherlands

To reduce losses from agricultural soils to surface water, mitigation options have to be implemented as a local scale. For a cost-effective implementation of these measures, an instrument to identify critical areas for P leaching is indispensable. In many countries, P-index methods are used to identify areas as risk for P losses to surface water. In flat areas, where losses by leaching are dominant, these methods have their limitations because leaching is often not described in detail, PLEASE, is a simple mechanistic model designed to stimulate P Losses by leaching at the field scale using a limited amount of local field data. In this study, PLEASE, was applied to 17 lowland sites in Denmark and 14 lowland sites in the Netherlands. Results show that the simple model simulated measured fluxes and concentrations in water from pipe drains, suction cups, and groundwater quite well. The modeling efficiency ranged from 0.92 for modeling total-P fluxes to 0.36 fr modeling concentrations in groundwater. Poor results were obtained for heavy clay soils and eutrophic peat soils, where fluxes and concentration were strongly underestimated by the model. The poot performance for the heavy clay soil can be explained by the transport of P through macropores to the drain pipes and the underestimation of overland flow on this heavy-textured soil. In the eutrophic peat soils, fluxes were underestimated due to the release of P from deep soil layers.     

Phosphorus load to surface water from bank erosion in a Danish lowland river basin

Phosphorus loss from bank erosion was studied in the catchment of River Odense, a lowland Danish river basin, with the aim of testing the hypothesis of whether stream banks act as major diffuse phosphorus (P) sources at catchment scale. Furthermore, the study aimed at analyzing the impact of different factors influencing bank erosion and P loss such as stream order, anthropogenic disturbances, width of uncultivated buffer strips, and the vegetation of buffer strips. A random stratified procedure in geographical information system (GIS) was used to select two replicate stream reaches covering different stream orders, channelized vs. naturally meandering channels, width of uncultivated buffer strips (≤ 2 m and ≥ 10 m), and buffer strips with different vegetation types. Thirty-six 100-m stream reaches with 180 bank plots and a total of 3000 erosion pins were established in autumn 2006, and readings were conducted during a 3-yr period (2006-2009). The results show that neither stream size nor stream disturbance measured as channelization of channel or the width of uncultivated buffer strip had any significant ( < 0.05) influence on bank erosion and P losses during each of the 3 yr studied. In buffer strips with natural trees bank erosion was significantly ( < 0.05) lower than in buffer strips dominated by grass and herbs. Gross and net P input from bank erosion amounted to 13.8 to 16.5 and 2.4 to 6.3 t P, respectively, in the River Odense catchment during the three study years. The net P input from bank erosion equaled 17 to 29% of the annual total P export and 21 to 62% of the annual export of P from diffuse sources from the River Odense catchment. Most of the exported total P was found to be bioavailable (71.7%) based on a P speciation of monthly suspended sediment samples collected at the outlet of the river basin. The results found in this study have a great importance for managers working with P mitigation and modeling at catchment scale.     

Propagation of uncertainties in soil and pesticide properties to pesticide leaching

In the new Dutch decision tree for the evaluation of pesticide leaching to groundwater, spatially distributed soil data are used by the GeoPEARL model to calculate the 90th percentile of the spatial cumulative distribution function of the leaching concentration in the area of potential usage (SP90). Until now it was not known to what extent uncertainties in soil and pesticide properties propagate to spatially aggregated parameters like the SP90. A study was performed to quantify the uncertainties in soil and pesticide properties and to analyze their contribution to the uncertainty in SP90. First, uncertainties in the soil and pesticide properties were quantified. Next, a regular grid sample of points covering the whole of the agricultural area in the Netherlands was randomly selected. At the grid nodes, realizations from the probability distributions of the uncertain inputs were generated and used as input to a Monte Carlo uncertainty propagation analysis. The analysis showed that the uncertainty concerning the SP90 is 10 times smaller than the uncertainty about the leaching concentration at individual point locations. The parameters that contribute most to the uncertainty about the SP90 are, however, the same as the parameters that contribute most to uncertainty about the leaching concentration at individual point locations (e.g., the transformation half-life in soil and the coefficient of sorption on organic matter). Taking uncertainties in soil and pesticide properties into account further leads to a systematic increase of the predicted SP90. The important implication for pesticide regulation is that the leaching concentration is systematically underestimated when these uncertainties are ignored.     

Effects of soil variability and weather conditions on pesticide leaching--a farm-level evaluation

In line with European regulations, Dutch law imposes an environmental threshold of 0.1 microg L(-1) on pesticide concentrations in ground water. During registration, the risk of exceeding this threshold is assessed through simulations for one or a few standard scenarios that do not reflect spatial variability under field conditions. The introduction of precision agriculture, where soil variability is actively managed, can increase control over pesticide leaching. This study presents a step-wise evaluation of the effects of soil variability and weather conditions on pesticide leaching. The evaluation was conducted on a 100-ha arable farm and aimed at identifying opportunities for precision management. As a first step, a relative risk assessment identified pesticides presenting a relatively high risk to the environment. Second, the effect of weather conditions was analyzed through 20 years of simulations for three distinct soil profiles. Results were summarized in cumulative probability plots to provide a probabilistic characterization of historical weather data. The year matching 90% probability (1981) served as a reference to simulate pesticide leaching from 612 soil profiles. After interpolation, areas where concentrations exceeded the environmental threshold were identified. Out of a total of 19 pesticides, isoproturon [N-dimethyl-N'-(4-(1-methylethyl)phenyl)urea], metribuzin [4-amino-6-tert-butyl-3-(methylthio)-as-triazin-5(4H)-one], and bentazon [2,1,3-benzothiadiazin-4(3H)-one, 3-isopropyl-, 2,2-dioxide] showed the highest risk for leaching. Leaching was strongly affected by soil variability at the within-field, field, and farm levels. Opportunities for precision management were apparent, but depended on the scale level at which environmental thresholds were implemented. When legislation is formulated in this issue, the presented step-wise evaluation can serve as a basis for identification and precision management of high-risk pesticides.     

Point- and nonpoint-source pesticide contamination in the Zwester Ohm catchment, Germany

Reducing pesticide loads in surface waters implies identifying the pathways responsible for the pollution. The current study documents the pesticide contamination of the river Zwester Ohm, a 4917-ha catchment in Germany with 41% of the land used for crop production. Discharges and concentrations of 19 pesticides were measured continuously at three locations for 15 mo. The load detected at the outlet of the catchment amounted to 9048 g a.i. The losses represent 0.22% of the pesticides applied by the farmers. The contamination showed a seasonal pattern following the pesticide application times. The wastewater treatment plant system (WWTPS) in the catchment (two wastewater treatment plants [WWTP], 14 combined sewer overflows (CSO), four CSO tanks) emits during dry weather periods purified sewage and during storm events sewage mixed with stormwater runoff into the river. The contribution by the WWTPS to the pesticide load was defined as point-source pollution (PSP). The load was dominated by PSP with at least 77% of the total pollution. No significant interdependencies between intrinsic properties of the pesticides, hydrometeorological factors, and the loads occurring in the stream could be found. Therefore, it is not possible to predict PSP for other catchments based on the results from this study. Whereas 65% of the total load entered the river via the WWTP, a portion of 12% was attributed to the CSO. The study points out that the influence of CSO on PSP should be taken into account in future catchment studies in areas with comparable agricultural structure.     

Impact of data quality and model complexity on prediction of pesticide leaching

Accurate input data for leaching models are expensive and difficult to obtain which may lead to the use of "general" non-site-specific input data. This study investigated the effect of using different quality data on model outputs. Three models of varying complexity, GLEAMS, LEACHM, and HYDRUS-2D, were used to simulate pesticide leaching at a field trial near Hamilton, New Zealand, on an allophanic silt loam using input data of varying quality. Each model was run for four different pesticides (hexazinone, procymidone, picloram and triclopyr); three different sets of pesticide sorption and degradation parameters (i.e., site optimized, laboratory derived, and sourced from the USDA Pesticide Properties Database); and three different sets of soil physical data of varying quality (i.e., site specific, regional database, and particle size distribution data). We found that the selection of site-optimized pesticide sorption (Koc) and degradation parameters (half-life), compared to the use of more general database derived values, had significantly more impact than the quality of the soil input data used, but interestingly also more impact than the choice of the models. Models run with pesticide sorption and degradation parameters derived from observed solute concentrations data provided simulation outputs with goodness-of-fit values closest to optimum, followed by laboratory-derived parameters, with the USDA parameters providing the least accurate simulations. In general, when using pesticide sorption and degradation parameters optimized from site solute concentrations, the more complex models (LEACHM and HYDRUS-2D) were more accurate. However, when using USDA database derived parameters, all models performed about equally.     

SIMULATION OF HERBICIDE CONCENTRATIONS IN STORMFLOW FROM FORESTED WATERSHEDS1

The breakpoint rainfall hydrology and pesticide options of the field scale model CREAMS (Chemicals, Runoff, and Erosion from Agricultural Management Systems) were used to predict average concentrations of hexazinone [3 cyclohexyl-6-(dimethyl-amino)-1-methyl-1,3,5-triazine-2,4(1H,3H)-dione] in stormflow from four forested watersheds in the upper Piedmont region of Georgia. Predicted concentrations were compared with measured concentrations recorded over a 13-month period. CREAMS accurately predicted hexazinone concetrations in the initial stormflow events which also contained the highest concentrations. The model underestimated the hexazinone concentrations in stormflow two months and greater following pesticide application. In a companion study, the daily rainfall option of the CREAMS model was used to evaluate the reltive risk associated with the maximum expected concentration of hexazinone, bromacil (5-bromo-3 sec-butyl-6 methyuracil), picloram (4-amino-3,5,6 trichloropicolinic acid), dicamba (3,6-dichloro-0-anisic acid), and triclopyr {[(3,5,6-trichloro-2-pyridinyl)oxy] acetic acid} in stormflow from small forested watersheds. The model predicted the following order of potential residue appearance in stormflow: bromacil>triclopyr>hexazinone>picloram>dicamba. Subsurface movement of residues via interflow and deep leaching losses are not simulated by the version of CREAMS used in these studies.     

Simulating sulfadimidine transport in surface runoff and soil at the microplot and field scale

To prevent residues of veterinary medicinal products (VMPs) from contaminating surface waters and ground water, an environmental impact assessment is required before a new product is allowed on the market. Physically based simulation models are advocated for the calculation of predicted environmental concentrations at higher tiers of the assessment process. However, the validation status of potentially useful models is poor for VMP transport. The objective of this study was to evaluate the dual-permeability model MACRO for simulation of transport of sulfonamide antibiotics in surface runoff and soil. Special focus was on effects of solute application in liquid manure, which may alter the hydraulic properties at the soil surface. To this end we used data from a microplot runoff experiment and a field experiment, both conducted on the same clay loam soil prone to preferential flow. Results showed that the model could accurately simulate concentrations of sulfadimidine and the nonreactive tracer bromide in runoff and in soil from the microplot experiments. The use of posterior parameter distributions from calibrations using the microplot data resulted in poor simulations for the field data of total sulfadimidine losses. The poor results may be due to surface runoff being instantly transferred off the field in the model, whereas in reality re-infiltration may occur. The effects of the manure application were reflected in smaller total and micropore hydraulic conductivities compared with the application in aqueous solution. These effects could easily be accounted for in regulatory modeling.     

Simulating solute transport in a structured field soil: uncertainty in parameter identification and predictions

Dual-permeability models have been developed to account for the significant effects of macropore flow on contaminant transport, but their use is hampered by difficulties in estimating the additional parameters required. Therefore, our objective was to evaluate data requirements for parameter identification for predictive modeling with the dual-permeability model MACRO. Two different approaches were compared: sequential uncertainty fitting (SUFI) and generalized likelihood uncertainty estimation (GLUE). We investigated six parameters controlling macropore flow and pesticide sorption and degradation, applying MACRO to a comprehensive field data set of bromide andbentazone [3-isopropyl-1H-2,1,3-benzothiadiazin-4(3H)-one-2,2dioxide] transport in a structured soil. The GLUE analyses of parameter conditioning for different combinations of observations showed that both resident and flux concentrations were needed to obtain highly conditioned and unbiased parameters and that observations of tracer transport generally improved the conditioning of macropore flow parameters. The GLUE "behavioral" parameter sets covered wider parameter ranges than the SUFI posterior uncertainty domains. Nevertheless, estimation uncertainty ranges defined by the 5th and 95th percentiles were similar and many simulations randomly sampled from the SUFI posterior uncertainty domains had negative model efficiencies (minimum of -3.2). This is because parameter correlations are neglected in SUFI and the posterior uncertainty domains were not always determined correctly. For the same reasons, uncertainty ranges for predictions of bentazone losses through drainflow for good agricultural practice in southern Sweden were 27% larger for SUFI compared with GLUE. Although SUFI proved to be an efficient parameter estimation tool, GLUE seems better suited as a method of uncertainty estimation for predictions.     

Controls on catchment-scale patterns of phosphorus in soil, streambed sediment, and stream water

Many models of phosphorus (P) transfer at the catchment scale rely on input from generic databases including, amongst others, soil and land use maps. Spatially detailed geochemical data sets have the potential to improve the accuracy of the input parameters of catchment-scale nutrient transfer models. Furthermore, they enable the assessment of the utility of available, generic spatial data sets for the modeling and prediction of soil nutrient status and nutrient transfer at the catchment scale. This study aims to quantify the unique and joint contribution of soil and sediment properties, land cover, and point-source emissions to the spatial variation of P concentrations in soil, streambed sediments, and stream water at the scale of a medium-sized catchment. Soil parent material and soil chemical properties were identified as major factors controlling the catchment-scale spatial variation in soil total P and Olsen P concentrations. Soil type and land cover as derived from the generic spatial database explain 33.7% of the variation in soil total P concentrations and 17.4% of the variation in Olsen P concentrations. Streambed P concentrations are principally related to the major element concentrations in streambed sediment and P delivery from the hillslopes due to sediment erosion. During base flow conditions, the total phosphorus (<0.45 microm) concentrations in stream water are mainly controlled by the concentrations of P and the major elements in the streambed sediment.     

MEASURED AND CREAMS-PREDICTED NITROGEN LOSSES FROM TOMATO AND CORN MANAGEMENT SYSTEMS1

Nitrogen runoff and leaching losses from two tomato and four corn field plots were compared to model predictions by CREAMS, a field-scale model for Chemicals, Runoff, and Erosion from Agricultural Management Systems. The tomato treatments were (1) trickle irrigation with one-half of applied N at preplant and one-half of applied N through the trickle irrigation system and (2) overhead sprinkler irrigation with one-half of applied N at preplant and one-half of applied N in two equal sidedressings. The corn treatments consisted of multiple N applications, minimum tillage, and “conventional” management. Soil type appeared to influence the ability of CREAMS to predict seasonal trends and treatment influences. Model predictions for N losses from tomato and corn treatments that were located on sandy soils often disagreed with measured values. Treatment influences and seasonal trends for N losses from corn treatments that were located on a higher clay content soil were more satisfactorily predicted by CREAMS. Even though model input parameter estimation and measurement techniques may be imperfect, the simulation ability of CREAMS for predicting N leaching losses from systems on deep sands probably needs to be improved. Sensitivity analyses indicated that annual NC₃^?-N leaching loss predictions were either minimally or not affected by changes in saturated hydraulic conductivity. Input estimations of the fraction of soil pore space filled at field capacity and soil organic matter were inversely related to annual NO₃^?-N leaching losses, while potential mineralizable N was directly related to yearly N leaching losses.     

Predicting nitrate leaching under potato crops using transfer functions

Nitrate leaching is a major issue in many cultivated soils. Models that predict the major processes involved at the field scale could be used to test and improve management practices. This study aims to evaluate a simple transfer function approach to predict nitrate leaching in sandy soils. A convective lognormal transfer (CLT) function is convoluted with functional equations simulating N mineralization, plant N uptake, N fertilizer dissolution, and nitrification at the soil surface to predict solute concentrations under potato (Solanum tuberosum L.) and barley (Hordeum vulgare L.) fields as a function of drainage water. Using this approach, nitrate flux concentrations measured in drainable lysimeters (1-m soil depth) were reasonably predicted from 29 Apr. 1996 to 3 Dec. 1996. With average application rates of 16.9 g m(-2) of N fertilizer in potato crops, mean nitrate-leaching losses measured under potato were 8.5 g N m(-2). Tuber N uptake averaged 9.7 g N m(-2) and soil mineral N at start (spring) and end (fall) of N mass balance averaged 1.7 and 4.5 g N m(-2), respectively. Soil N mineralization was estimated by difference (4.3 g N m(-2) on average) and was small compared with N fertilization. Small nitrate flux concentrations at the beginning of the cropping season (May) resulted mainly from initial soil nitrate concentrations. Measured and predicted nitrate flux concentrations significantly increased at mid-season (July-August) following important drainage events coupled with complete dissolution and nitrification of N fertilizers, and declining N uptake by potato plants. Decreases in nitrate concentrations before the end of year (November-December) underlined the predominant effect of N fertilizers applied for the most part at planting acting as a pulse input of solute.     

AN OBJECTWE TEST FOR HYDROLOGIC SCALE1

ABSTRACT: Improving the reliability of parametric hydrologic models (sometimes called cenceptual rainfall-runoff models) in the continuous simulation of runoff from ungaged catchments has been frustrated by difficulties in estimating model parameters from catchment characteristics. An underlying problem is that these models use parameters to represent catchments as a whole, whereas data on catchment characteristics are collected at multiple field locations and are difficult to transform into one measure of collective impact. Subdividing the catchment and calibrating a stochastic parametric model to estimate distributions for the parameters that covered the range of observed streamflow values was found to improve the simulations. This paper presents an optimization of the amount of subdivision to use in simulation with a version of the Stanford Watershed Model using available climatological data. The calibration process assumes that catchment heterogeneity introduces errors that can be reduced by calibrating parameters as spatial distributions rather than single values. Calibrations for three diverse small gaged catchments located in California and in Virginia found the optimal number of subdivisions to range from 4 to 25 and the optimal scale to range from 0.3 to 2.1 mi².     

Field-scale variation in microbial activity and soil properties in relation to mineralization and sorption of pesticides in a sandy soil

Pesticides applied to agricultural soils are subject to environmental concerns because leaching to groundwater reservoirs and aquatic habitats may occur. Knowledge of field variation of pesticide-related parameters is required to evaluate the vulnerability of pesticide leaching. The mineralization and sorption of the pesticides glyphosate and metribuzin and the pesticide degradation product triazinamin in a field were measured and compared with the field-scale variation of geochemical and microbiological parameters. We focused on the soil parameters clay and organic carbon (C) content and on soil respiratory and enzymatic processes and microbial biomass. These parameters were measured in soil samples taken at two depths (Ap and Bs horizon) in 51 sampling points from a 4-ha agricultural fine sandy soil field. The results indicated that the spatial variation of the soil parameters, and in particular the content of organic C, had a major influence on the variability of the microbial parameters and on sorption and pesticide mineralization in the soil. For glyphosate, with a co-metabolic pathway for degradation, the mineralization was increased in soils with high microbial activity. The spatial variability, expressed as the CV, was about five times higher in the Bs horizon than in the Ap horizon, and the local-scale variation within 100 m(2) areas were two to three times lower than the field-scale variation within the entire field of about 4 ha.     

A hydrological modeling framework for defining achievable performance standards for pesticides

This paper proposes a hydrological modeling framework to define achievable performance standards (APSs) for pesticides that could be attained after implementation of recommended management actions, agricultural practices, and available technologies (i.e., beneficial management practices [BMPs]). An integrated hydrological modeling system, Gestion Intégrée des Bassins versants à l'aide d'un Système Informatisé, was used to quantify APSs for six Canadian watersheds for eight pesticides: atrazine, carbofuran, dicamba, glyphosate, MCPB, MCPA, metolachlor, and 2,4-D. Outputs from simulation runs to predict pesticide concentration under current conditions and in response to implementation of two types of beneficial management practices (reduced pesticide application rate and 1- to 10-m-wide edge-of-field and/or riparian buffer strips, implemented singly or in combination) showed that APS values for scenarios with BMPs were less than those for current conditions. Moreover, APS values at the outlet of watersheds were usually less than ecological thresholds of good condition, when available. Upstream river reaches were at greater risk of having concentrations above a given ecological thresholds because of limited stream flows and overland loads of pesticides. Our integrated approach of "hydrological modeling-APS estimation-ecotoxicological significance" provides the most effective interpretation possible, for management and education purposes, of the potential biological impact of predicted pesticide concentrations in rivers.     

Chloropicrin emissions after shank injection: two-dimensional analytical and numerical model simulations of different source methods and field measurements

Understanding the control mechanisms of fumigant movement in soil is a fundamental step for developing management strategies to reduce atmospheric emissions. Most soil fumigants including chloropicrin (CP) are applied by shank injection, and the application process often leaves vertical soil fractures that would potentially cause preferential fumigant movement and increased emissions. This potential transport pathway was evaluated by comparing cumulative emissions and soil air concentrations of CP from direct field measurements with those predicted using analytical and numerical models after assuming either point or rectangle sources for the injected CP. Results clearly showed that shank-injected CP, when treated as vertical rectangle sources, produced cumulative emission losses similar to the field measurements. Treating the shanked CP as point sources caused approximately 50% underprediction than the field measurements. The study also demonstrated that fumigant cumulative emissions can be predicted, with reasonable accuracy, using either analytical or numerical simulations.