Modelling climate change impacts on maize yields under low nitrogen input conditions in sub-Saharan Africa.

Smallholder farmers in sub-Saharan Africa (SSA) currently grow rainfed maize with limited inputs including fertilizer. Climate change may exacerbate current production constraints. Crop models can help quantify the potential impact of climate change on maize yields, but a comprehensive multimodel assessment of simulation accuracy and uncertainty in these low-input systems is currently lacking. We evaluated the impact of varying [CO2 ], temperature and rainfall conditions on maize yield, for different nitrogen (N) inputs (0, 80, 160 kg N/ha) for five environments in SSA, including cool subhumid Ethiopia, cool semi-arid Rwanda, hot subhumid Ghana and hot semi-arid Mali and Benin using an ensemble of 25 maize models. Models were calibrated with measured grain yield, plant biomass, plant N, leaf area index, harvest index and in-season soil water content from 2-year experiments in each country to assess their ability to simulate observed yield. Simulated responses to climate change factors were explored and compared between models. Calibrated models reproduced measured grain yield variations well with average relative root mean square error of 26%, although uncertainty in model prediction was substantial (CV = 28%). Model ensembles gave greater accuracy than any model taken at random. Nitrogen fertilization controlled the response to variations in [CO2 ], temperature and rainfall. Without N fertilizer input, maize (a) benefited less from an increase in atmospheric [CO2 ]; (b) was less affected by higher temperature or decreasing rainfall; and (c) was more affected by increased rainfall because N leaching was more critical. The model intercomparison revealed that simulation of daily soil N supply and N leaching plays a crucial role in simulating climate change impacts for low-input systems. Climate change and N input interactions have strong implications for the design of robust adaptation approaches across SSA, because the impact of climate change in low input systems will be modified if farmers intensify maize production with balanced nutrient management.

Simulating nitrate-nitrogen concentration from a subsurface drainage system in response to nitrogen application rates using RZWQM2.

Computer models have been widely used to evaluate the impact of agronomic management on nitrogen (N) dynamics in subsurface drained fields. However, they have not been evaluated as to their ability to capture the variability of nitrate-nitrogen (NO(3)-N) concentration in subsurface drainage at a wide range of N application rates due to possible errors in the simulation of other system components. The objective of this study was to evaluate the performance of Root Zone Water Quality Model2 (RZWQM2) in simulating the response of NO(3)-N concentration in subsurface drainage to N application rate. A 16-yr field study conducted in Iowa at nine N rates (0-252 kg N ha(-1)) from 1989 to 2004 was used to evaluate the model, based on a previous calibration with data from 2005 to 2009 at this site. The results showed that the RZWQM2 model performed "satisfactorily" in simulating the response of NO(3)-N concentration in subsurface drainage to N fertilizer rate with 0.76, 0.49, and -3% for the Nash-Sutcliffe efficiency, the ratio of the root mean square error to the standard deviation, and percent bias, respectively. The simulation also identified that the N application rate required to achieve the maximum contaminant level for the annual average NO(3)-N concentration was similar to field-observed data. This study supports the use of RZWQM2 to predict NO(3)-N concentration in subsurface drainage at various N application rates once it is calibrated for the local condition.

Application of the Root Zone Water Quality Model (RZWQM) to pesticide fate and transport: an overview.

Pesticide transport models are tools used to develop improved pesticide management strategies, study pesticide processes under different conditions (management, soils, climates, etc) and illuminate aspects of a system in need of more field or laboratory study. This paper briefly overviews RZWQM history and distinguishing features, overviews key RZWQM components and reviews RZWQM validation studies. RZWQM is a physically based agricultural systems model that includes sub-models to simulate: infiltration, runoff, water distribution and chemical movement in the soil; macropore flow and chemical movement through macropores; evapotranspiration (ET); heat transport; plant growth; organic matter/nitrogen cycling; pesticide processes; chemical transfer to runoff; and the effect of agricultural management practices on these processes. Research to date shows that if key input parameters are calibrated, RZWQM can adequately simulate the processes involved with pesticide transport (ET, soil-water content, percolation and runoff, plant growth and pesticide fate). A review of the validation studies revealed that (1) accurate parameterization of restricting soil layers (low permeability horizons) may improve simulated soil-water content; (2) simulating pesticide sorption kinetics may improve simulated soil pesticide concentration with time (persistence) and depth and (3) calibrating the pesticide half-life is generally necessary for accurate pesticide persistence simulations. This overview/review provides insight into the processes involved with the RZWQM pesticide component and helps identify model weaknesses, model strengths and successful modeling strategies.

Documenting the pesticide processes module of the ARS RZWQM agroecosystem model.

We describe the theory and current development state of the pesticide process module of the USDA-Agricultural Research Service Root Zone Water Quality Model, or RZWQM. Several processes which are significant in determining the fate of a pesticide application are included together in this module for the first time, including application technique, root uptake, ionic dissociation, soil depth dependence of persistence, volatilization, wicking upward in soil and aging of residues. The pesticide module requires a large number of parameters to run (as does the RZWQM model as a whole) and it is becoming clear that RZWQM will find most interest and use as part of a 'scenario' in which all data requirements are supplied and the predictions of the system compared with a real (usually partial) data set. Such a scenario may then be modified to examine the response of the system to changes in inputs. It also has significant potential as a technology transfer or teaching tool, providing detailed understanding of a specific agronomic system and its potential impacts on the environment.

The pesticide module of the Root Zone Water Quality Model (RZWQM): testing and sensitivity analysis of selected algorithms for pesticide fate and surface runoff.

The Root Zone Water Quality Model (RZWQM) is a one-dimensional, numerical model for simulating water movement and chemical transport under a variety of management and weather scenarios at the field scale. The pesticide module of RZWQM includes detailed algorithms that describe the complex interactions between pesticides and the environment. We have simulated a range of situations with RZWQM, including foliar interception and washoff of a multiply applied insecticide (chlorpyrifos) to growing corn, and herbicides (alachlor, atrazine, flumetsulam) with pH-dependent soil sorption, to examine whether the model appears to generate reasonable results. The model was also tested using chlorpyrifos and flumetsulam for the sensitivity of its predictions of chemical fate and water and pesticide runoff to various input parameters. The model appears to generate reasonable representations of the fate and partitioning of surface- and foliar-applied chemicals, and the sorption of weakly acidic or basic pesticides, processes that are becoming increasingly important for describing adequately the environmental behavior of newer pesticides. However, the kinetic sorption algorithms for charged pesticides appear to be faulty. Of the 29 parameters and variables analyzed, chlorpyrifos half-life, the Freundlich adsorption exponent, the fraction of kinetic sorption sites, air temperature, soil bulk density, soil-water content at 33 kPa suction head and rainfall were most sensitive for predictions of chlorpyrifos residues in soil. The latter three inputs and the saturated hydraulic conductivity of the soil and surface crusts were most sensitive for predictions of surface water runoff and water-phase loss of chlorpyrifos. In addition, predictions of flumetsulam (a weak acid) runoff and dynamics in soil were sensitive to the Freundlich equilibrium adsorption constant, soil pH and its dissociation coefficient.

Modeling hydrology, metribuzin degradation and metribuzin transport in macroporous tilled and no-till silt loam soil using RZWQM.

Due to the complex nature of pesticide transport, process-based models can be difficult to use. For example, pesticide transport can be effected by macropore flow, and can be further complicated by sorption, desorption and degradation occurring at different rates in different soil compartments. We have used the Root Zone Water Quality Model (RZWQM) to investigate these phenomena with field data that included two management conditions (till and no-till) and metribuzin concentrations in percolate, runoff and soil. Metribuzin degradation and transport were simulated using three pesticide sorption models available in RZWQM: (a) instantaneous equilibrium-only (EO); (b) equilibrium-kinetic (EK, includes sites with slow desorption and no degradation); (c) equilibrium-bound (EB, includes irreversibly bound sites with relatively slow degradation). Site-specific RZWQM input included water retention curves from four soil depths, saturated hydraulic conductivity from four soil depths and the metribuzin partition coefficient. The calibrated parameters were macropore radius, surface crust saturated hydraulic conductivity, kinetic parameters, irreversible binding parameters and metribuzin half-life. The results indicate that (1) simulated metribuzin persistence was more accurate using the EK (root mean square error, RMSE = 0.03 kg ha(-1)) and EB (RMSE = 0.03 kg ha(-1)) sorption models compared to the EO (RMSE = 0.08 kg ha(-1)) model because of slowing metribuzin degradation rate with time and (2) simulating macropore flow resulted in prediction of metribuzin transport in percolate over the simulation period within a factor of two of that observed using all three pesticide sorption models. Moreover, little difference in simulated daily transport was observed between the three pesticide sorption models, except that the EB model substantially under-predicted metribuzin transport in runoff and percolate >30 days after application when transported concentrations were relatively low. This suggests that when macropore flow and hydrology are accurately simulated, metribuzin transport in the field may be adequately simulated using a relatively simple, equilibrium-only pesticide model.

Test of the Root Zone Water Quality Model (RZWQM) for predicting runoff of atrazine, alachlor and fenamiphos species from conventional-tillage corn mesoplots.

The Root Zone Water Quality Model (RZWQM) is a comprehensive, integrated physical, biological and chemical process model that simulates plant growth and movement of water, nutrients and pesticides in a representative area of an agricultural system. We tested the ability of RZWQM to predict surface runoff losses of atrazine, alachlor, fenamiphos and two fenamiphos oxidative degradates against results from a 2-year mesoplot rainfall simulation experiment. Model inputs included site-specific soil properties and weather, but default values were used for most other parameters, including pesticide properties. No attempts were made to calibrate the model except for soil crust/seal hydraulic conductivity and an adjustment of pesticide persistence in near-surface soil. Approximately 2.5 (+/- 0.9), 3.0 (+/- 0.8) and 0.3 (+/- 0.2)% of the applied alachlor, atrazine and fenamiphos were lost in surface water runoff, respectively. Runoff losses in the 'critical' events--those occurring 24 h after pesticide application--were respectively 91 (+/- 5), 86 (+/- 6) and 96 (+/- 3)% of total runoff losses for these pesticides. RZWQM adequately predicted runoff water volumes, giving a predicted/observed ratio of 1.2 (+/- 0.5) for all events. Predicted pesticide concentrations and loads from the 'critical' events were generally within a factor of 2, but atrazine losses from these events were underestimated, which was probably a formulation effect, and fenamiphos losses were overestimated due to rapid oxidation. The ratios of predicted to measured pesticide concentrations in all runoff events varied between 0.2 and 147, with an average of 7. Large over-predictions of pesticide runoff occurred in runoff events later in the season when both loads and concentrations were small. The normalized root mean square error for pesticide runoff concentration predictions varied between 42 and 122%, with an average of 84%. Pesticide runoff loads were predicted with a similar accuracy. These results indicate that the soil-water mixing model used in RZWQM is a robust predictor of pesticide entrainment and runoff.

Herbicide leaching as affected by macropore flow and within-storm rainfall intensity variation: a RZWQM simulation.

Within-event variability in rainfall intensity may affect pesticide leaching rates in soil, but most laboratory studies of pesticide leaching use a rainfall simulator operating at constant rainfall intensity, or cover the soil with ponded water. This is especially true in experiments where macropores are present--macroporous soils present experimental complexities enough without the added complexity of variable rainfall intensity. One way to get around this difficulty is to use a suitable pesticide transport model, calibrate it to describe accurately a fixed-intensity experiment, and then explore the affects of within-event rainfall intensity variation on pesticide leaching through macropores. We used the Root Zone Water Quality Model (RZWQM) to investigate the effect of variable rainfall intensity on alachlor and atrazine transport through macropores. Data were used from an experiment in which atrazine and alachlor were surface-applied to 30 x 30 x 30 cm undisturbed blocks of two macroporous silt loam soils from glacial till regions. One hour later the blocks were subjected to 30-mm simulated rain with constant intensity for 0.5 h. Percolate was collected and analyzed from 64 square cells at the base of the blocks. RZWQM was calibrated to describe accurately the atrazine and alachlor leaching data, and then a median Mid-west variable-intensity storm, in which the initial intensity was high, was simulated. The variable-intensity storm more than quadrupled alachlor losses and almost doubled atrazine losses in one soil over the constant-intensity storm of the same total depth. Also rainfall intensity may affect percolate-producing macroporosity and consequently pesticide transport through macropores. For example, under variable rainfall intensity RZWQM predicted the alachlor concentration to be 2.7 microg ml(-1) with an effective macroporosity of 2.2 E(-4) cm(3) cm(-3) and 1.4 microg ml(-1) with an effective macroporosity of 4.6 E(-4) cm(3) cm(-3). Percolate-producing macroporosity and herbicide leaching under different rainfall intensity patterns, however, are not well understood. Clearly, further investigation of rainfall intensity variation on pesticide leaching through macropores is needed.

Software for pest-management science: computer models and databases from the United States Department of Agriculture-Agricultural Research Service.

We present an overview of USDA Agricultural Research Service (ARS) computer models and databases related to pest-management science, emphasizing current developments in environmental risk assessment and management simulation models. The ARS has a unique national interdisciplinary team of researchers in surface and sub-surface hydrology, soil and plant science, systems analysis and pesticide science, who have networked to develop empirical and mechanistic computer models describing the behavior of pests, pest responses to controls and the environmental impact of pest-control methods. Historically, much of this work has been in support of production agriculture and in support of the conservation programs of our 'action agency' sister, the Natural Resources Conservation Service (formerly the Soil Conservation Service). Because we are a public agency, our software/database products are generally offered without cost, unless they are developed in cooperation with a private-sector cooperator. Because ARS is a basic and applied research organization, with development of new science as our highest priority, these products tend to be offered on an 'as-is' basis with limited user support except for cooperating R&D relationship with other scientists. However, rapid changes in the technology for information analysis and communication continually challenge our way of doing business.