Elsevier

Agricultural Systems

Volume 106, Issue 1, February 2012, Pages 59-71
Agricultural Systems

Extending results from agricultural fields with intensively monitored data to surrounding areas for water quality management

https://doi.org/10.1016/j.agsy.2011.10.010Get rights and content

Abstract

A 45% reduction in riverine total nitrogen flux from the 1980–1996 time period is needed to meet water quality goals in the Mississippi Basin and Gulf of Mexico. This paper addresses the goal of reducing nitrogen in the Mississippi River through three objectives. First, the paper outlines an approach to the site-specific quantification of management effects on nitrogen loading from tile drained agriculture using a simulation model and expert review. Second, information about the net returns to farmers is integrated with the nitrogen loading information to assess the incentives to adopt alternative management systems. Third, the results are presented in a decision support framework that compares the rankings of management systems based on observed and simulated values for net returns and nitrogen loading. The specific question addressed is how information about the physical and biological processes at Iowa State University’s Northeast Research Farm near Nashua, Iowa, could be applied over a large area to help farmers select management systems to reduce nitrogen loading in tile drained areas. Previous research has documented the parameterization and calibration of the RZWQM model at Nashua to simulate 35 management system effects on corn and soybean yields and N loading in tileflow from 1990 to 2003. As most management systems were studied for a 6 year period and in some cases weather had substantial impacts, a set of 30 alternative management systems were also simulated using a common 1974–2003 input climate dataset. To integrate an understanding of the economics of N management, we calculated net returns for all management systems using the DevTreks social budgeting tool. We ranked the 35 observed systems in the Facilitator decision support tool using N loading and net returns and found that rankings from simulated results were very similar to those from the observed results from both an onsite and offsite perspective. We analyzed the effects of tillage, crop rotation, cover crops, and N application method, timing, and amount for the 30 long term simulations on net returns and N loading. The primary contribution of this paper is an approach to creating a quality assured database of management effects on nitrogen loading and net returns for tile drained agriculture in the Mississippi Basin. Such a database would systematically extend data from intensively monitored agricultural fields to the larger area those fields represent.

Highlights

► The Mississippi needs a 45% reduction in nitrogen flux to meet water quality goals. ► Large reductions in N loading require commensurate changes to management systems. ► Site-specific economics of alternative management systems have to be understood. ► Management effect databases integrate observed data, models, and expert opinion. ► Decision support tools can apply such databases to meet water quality goals.

Introduction

Agriculture faces significant challenges meeting the world’s food, fiber, and fuel needs while reducing environmental impacts. The general problem addressed in this paper is how to provide better information to farmers about the impacts of management decisions on offsite agrichemical loadings. The specific question is: How can the research, extension, and conservation community systematically quantify management impacts on nitrogen loading and net returns from tile drained agriculture, and apply that information for field scale decision making?

The National Academy of Engineering identified managing the nitrogen cycle as one of the Grand Challenges of Engineering (NAE, 2008). Increased N raises the risks to human health and the health of water bodies, increases treatment costs for drinking water and contributes to hypoxia in the Gulf of Mexico. According to the Gulf Hypoxia Action Plan 2008 (Mississippi River/Gulf of Mexico Watershed Nutrient Task Force, 2008, p. 22):

Significant reductions in nitrogen and phosphorus are needed. To achieve the Coastal Goal for the size of the hypoxic zone and improve water quality in the Basin, a dual nutrient strategy targeting at least a 45% reduction in riverine total nitrogen load and in riverine total phosphorus load, measured against the average load over the 1980–1996 time period, may be necessary.

and later (p. 44):

Understanding the most efficient and cost-effective conservation practices and management practices to reduce nutrient loads is central to the success of nutrient reduction strategies.

Agriculture is certainly not the only source of nitrogen to the Gulf, but reducing N loading in the northern Midwest is a critical issue, as this area produces high rates of N loading to the Mississippi (Heinz Center, 2002, pp. 46–47). Tile drained agriculture, in particular, requires a focus on N management, as the drains “short-circuit” soil water high in nitrate into streams. Unlike erosion, there are no visible signs, so farmers and conservationists cannot develop an intuitive sense for the relationship between management and N loading. In a study from two Minnesota watersheds, Petrolia and Gowda (2006) point out that controlling N loading in tile drained areas is likely to be a critical component of efforts to control hypoxia in the Gulf of Mexico, that tile drained processes should be explicitly modeled, and that certain nutrient management policies are only effective on tile drained land.

The foundation for understanding natural systems, and the effects of human activities on those systems, is observed data. As a guide to the practical management of those systems however, observed data is almost always limited, so decision making is not straightforward even on intensively studied areas. This paper outlines a process for improving the technical information about management effects on water quality provided to producers.

Farmers focus on agronomic and business issues in agricultural production. To reduce water quality problems from cropland, ultimately farmers need to adopt management systems that reduce N loading on a field by field basis. Delivering technical information requires trained specialists, often private consultants, Extension Agents, and NRCS Conservationists. Many in the conservation community are skilled at distilling research into guidelines or pithy decision rules. However, as the complexity of modern agriculture increases and the list of resource concerns expands, it becomes increasingly difficult for a conservationist to assimilate research findings, interpret, and apply those findings to a particular farmer’s conditions while meeting other job demands. Simulation models provide an option for extending results and applying these to a larger set of fields, but effective and efficient application of models require yet another skill set, investment of time to parameterize, and the outputs can be difficult to interpret, particularly if multiple resource problems are being evaluated. In some cases there is a need for multiple models which further increases the complexity of the problem.

As shown on the top in Fig. 1, currently researchers publish results directly in the literature and Extension Agents and NRCS National, State, and Field Staffs are expected to interpret that information to farmers and help farmers apply that understanding under local conditions. To build a solid scientific foundation for conservation, publishing results in the literature is clearly essential. But when the range of subjects and depth of understanding required of conservationists becomes too large and complex for a conservation agency, let alone a field level conservationist to digest, an alternative approach is needed. We propose that for particularly problematic areas a database of results could capture the important relationships and simplify the Conservationist’s job.

Information technology has advanced to the point where it will soon be feasible for many specialists, with various backgrounds, to collaborate systematically in addressing critical resource management issues that require interdisciplinary cooperation. A new, more integrated and systematic approach is needed that allows for increased collaboration by specialists (Heilman et al., 2002). A specialist or team of specialists would calibrate a comprehensive simulation model to quantify expected management effects, the estimates are then quality assured by local experts and used to populate a database for strategic decision-making. The task of the conservationist is then to interpret the database of estimated effects to farmers either through a specific DSS tool or incorporated into an N index tool (Delgado et al., 2008) specifically for tile drained agriculture. A nine step process for creating and applying such a database is shown in Fig. 2. The rest of this paper describes the application of Steps 1–5 and 8 in that process, based on an intensively monitored field scale research site.

Section snippets

Problem definition (Steps 1 and 2)

The foundation for this study is a dataset from Iowa State University’s Northeast Research Center near Nashua, IA (43.0 N, 92.5 W), hereafter referred to as “Nashua”. The robust dataset includes 14 years (1990–2003) of weather records, corn (Zea mays L.) and soybean (Glycine max (L.) Merr.) yields, tileflow, and N concentration in drainage. According to Bakhsh et al. (2007), slopes at Nashua range from 0% to 8%, which corresponds to the slope groups A, B, and C (0–2%; 2–5%; and 5–9%). For this

RZWQM simulation of 1990–2003 Nashua dataset

N loading in tile drained systems is determined by the combination of tileflow and the N concentration in the tileflow. Fig. 4 shows the RZWQM simulation results compared to the measured crop yield, tileflow, N concentration, and N load for 30 plots over 14 years totaling 420 plot years. There is significant scatter around the 1–1 line for corn yields, in part because of processes affecting yield, e.g., insects and hail. Another issue was an increasing areawide trend in corn yields across the 14 

Simulation, review, and application of the Nashua dataset

The parameterization, calibration, and analysis of RZWQM for the Nashua dataset required several scientist-years, far beyond the skills and time available to conservationists working with farmers. In addition, modification of RZWQM was needed to incorporate lateral flow. Clearly this step in the overall process should be performed by modeling specialists. Table 6 summarizes the efforts made to systematically simulate and review management effects at Nashua and incorporate this information into

Conclusion

We used the RZWQM simulation model and DevTreks budgeting tool to quantify the effects of management systems on farm income and N loading. Two sets of simulations were performed, first for 35 systems through the observed 14 year study period and second for a set of 30 systems over 30 years. We compared the ranking of management systems based on observed data and simulations from RZWQM in the Facilitator multiobjective tool, and found that calibrated RZWQM results led to a very similar ranking

Acknowledgements

We would like to thank Carl Pederson and Ken Pecinovsky of Iowa State University, Cathy Woodard and Caitlin Hall of the University of Arizona, Jim Ayen, Alan Lauver, Barbara Stewart, Michael Sucik, Steve Brinkman, and Hal Cosby of the NRCS in Iowa, Reggie Voss, formerly of the Extension Service, Philip Algreen of Agri Drain, Terry Meade, Bob Jacquis, and Gerardo Armendariz of the Agricultural Research Service.

References (32)

  • S.A. Saseendran et al.

    Simulating management effects on crop production, tile drainage, and water quality using RZWQM–DSSAT

    Geoderma

    (2007)
  • AAEA Task Force on Commodity Costs and Returns, 2000. Commodity Costs and Returns Handbook. Iowa State University,...
  • A. Bakhsh et al.

    N-application methods and precipitation pattern effects on subsurface drainage nitrate losses and crop yields

    Water, Air, and Soil Pollution

    (2010)
  • Boyle, K.P., Heilman, P., Malone, R.W., Ma, L., Kanwar, R.S., in preparation. Social budgeting to improve agricultural...
  • P. Heilman et al.

    How good is good enough? What information is needed for agricultural water quality planning and how can it be provided affordably?

    Journal of Soil and Water Conservation

    (2002)
  • View full text