<< Back to Research Projects

Cropping System Performance Evaluation Based on Remote and Ground Level Data: An Integration of Research Groups

G.A. Bollero, D.G. Bullock, D.S. Bullock, and M.B. Villamil


Over the past decade a substantial amount of characterization data has been collected at the Dudley Smith farm. Some of these efforts include information on geo-referenced soil chemical characteristics and crop yields. With the support of more advanced sampling and analytical technologies the existing information can be improved and meshed with the current efforts at the Dudley Smith farm. When addressing landscape scale issues, no single discipline has the ability to develop and deliver solutions at the required quality and complexity. This initiative offers and opportunity for the development of a research team interested in system performance evaluation based on technologically advanced data acquisition and analysis.


Since 1998, Drs. G.A. Bollero and D.G. Bullock have been members of an interdisciplinary, multi-state project team investigating landscape variability and its effects on system productivity and environmental services. These efforts are being supported by three grants through the North Central Soybean and Research Program (NCSRP), the United Soybean Board (USB), and IFAFS-USDA. Most of the work in Illinois has been done at the UIUC Williams farm at Bellflower, IL and it has consisted on field scale and remote sensor acquisition and analysis of dense data sets. We have established methodologies and algorithms for understanding landscape variability and their relationship with system performance. In addition, this group has pioneered the integration and understanding of remote sensed data with production goals and environmental stewardship. Dr. D.S. Bullock has provided leadership and critical input into the economic analysis of these problems. In addition, Dr. D.G. Bullock and his group are part of the Illinois Precision Agriculture and Remote Sensing Laboratory (IPARSL).

The use of sensors is ideal for studying landscape characteristics because they require little labor to operate, are suited well to repeated measures, and can be linked to GPS and computers for on-the-go spatial data collection. Sensors that directly measure important soil properties are limited at this time, but other sensors can be used to give surrogate measurements of various field characteristics. For example, soil electrical conductivity (EC) and aerial images provide information about important plant-mediating soil properties. We have shown that sensor measurements can be used in conjunction with terrain attributes and statistical procedures to create indirect, but valid, measures of soil properties and to generate spatial soil characteristic indices that correlate to system performance.

The recent introduction of Real-Time Kinematic (RTK) GPS receivers has made possible the automated collection of highly accurate elevation data, thus providing an efficient way of obtaining high-resolution digital elevation models (DEM) of agricultural fields. We have shown that field topography plays an important role in the hydrological response of a catchment to rainfall and has a major impact on water availability to crop production and nutrient leaching potential. The increasing availability of DEMs and advent of computerized terrain analysis tools have made it possible to quantify the topographic attributes (e.g. slope, aspect, curvature, flow path length) of a landscape.

Analysis: The project will be broken into three separate, yet integrated, components including:

Agricultural remote sensing (e.g., remote sensing of crop parameters, estimation of the amount of land in production at any given time, and detection of infestations and diseases among crops) also has the potential to make valuable contributions to the daily efforts of crop producers. The enhanced spectral and spatial resolutions of a variety of new remote sensing systems offer a new means to identify vegetation stress and other anomalies. Furthermore, enhanced satellite imagery, such as IKONOS, is becoming available for input into robust models for estimating in-field agronomic parameters in a cost-effective manner.

We have completed the necessary background research to understand the challenge. We are uniquely qualified because we have the expertise, tools, and equipment required to attract other research groups on and off campus and together develop a multidisciplinary plan of work.

Environmental evaluation

The detrimental environmental impacts of cash crops and livestock production are costs that raise questions about the sustainability of current practices. Agricultural practices are a major source of nitrate (NO3) contamination of ground and surface waters. Nitrogen (N) and NO3 concentrations exceeding Maximum Contaminant Levels (MCL’s -10 mg L-1) have been detected in public supplies and private wells in the Midwest. Moreover, in Illinois 40 percent of the agricultural acreage is drained, creating a very efficient conduit by which NO3 enter streams. In this way, Illinois supplies a large fraction of the N and P loads found in the Mississippi River, contributing to the hypoxia zone in the Gulf of Mexico.

Nitrate contamination of waters is closely related to N fertilization practices. However, livestock production also leads to an increased release of NO3 to ground and surface waters. It has been documented that 70-80% of N consumed by cattle is returned to the pasture mainly in the form of urine. Nitrate concentrations averaged 25 mg L-1 in groundwater draining an intensive grazed area in New Zealand. Thus, within integrated systems such as the one currently implemented at the DS farm, NO3 contamination of waters could be related to N fertilization practices as well as the impact of grazing cattle.

Nitrogen sources (i.e. synthetic fertilizer and excreta) and their interaction with management practices and environmental conditions could lead to excessive NO3 accumulation in the soil profile. Uptake inefficiency also helps to increase NO3 levels in the soil since the crop uses only about half of the N applied to agricultural fields while in the case of animals, the return of N in excreta is also the result of an inefficient utilization of dietary N, being of about 60% under ideal circumstances. Unused N either denitrifies returning to the atmosphere or it remains available in the soil for leaching into streams. As a result, future attempts to improve water quality should focus on source control, specifically, preventing NO3 from entering subsurface flow paths.

The use of winter cover crops (WCC) and pastures, as currently established at DS, seem to be the logical choice for immobilization of NO3 in the soil profile after harvest of the main crops. Many studies have documented the ability of WCC and pastures to reduce the leaching of NO3 and other nutrients from the root zone however, this information is lacking at the integrated system level.

Agricultural landscape research presents great challenges for scientists involved in environmental management. Extrapolation of results found in agroecological studies is a desirable yet a difficult task. Discrepancies in response variables are commonly encountered among locations and also in different years for the same location, accounting for variations in boundary conditions and properties that are not adequately quantified or included in the analysis. In fact, the underlying cause of such variation remains unknown. During a recent Symposium on Landscape Research held in Germany (2001), it become clear the need to identify the underlying processes that occur in the landscape, to understand its inherent variability, to quantify the meaning of different factors, and to integrate those factors across the landscape. Landscape ecology and ecosystem model, based on careful consideration of the major state factors offers an organizing framework that would allow addressing these needs. Major state factors are also referred as the factors of soil formation,climate, relief, parent material, and organisms, interacting in time.

Modern GIS-technology allows the integration of various sources of spatial and attribute data and it has been successfully applied to assess variability in some of the major state factors at the landscape level. If we can integrate the different aspects and appropriate factors within the landscape, we will be able to contribute to a more profound understanding of landscape processes, which will coincide, with a better recognition of the quality and the management of the agroecosystems.


We propose:

  1. To integrate the experience and ideas of UIUC scientists working with remote sensing, spatial variability, and landscape ecology to develop a long term plan for the evaluation of the environmental performance of the current Dudley Smith system. Drs. G.A. Bollero and D.G. Bullock will take the lead to integrate other groups such as Spectral Vision, NCSA, and IPARSL. M.B. Villamil, a Fulbright scholar in the Crop Sciences department, will evaluate existing data and will serve as a liaison for the development of DSystem proposal.
  2. To integrate this new layer of knowledge with the current research and extension structure of the Dudley Smith farm. We believe that a DSystem proposal with a heuristic approach to evaluating system performance at the landscape level will successfully mesh with the strategies currently established at Dudley Smith.
  3. To seek external support: as a DSynergy group we will take advantage of the unique “systems” opportunity that Dudley Smith offers and develop a proposal to be submitted in November 2003 to the National Research Initiative (NRI) integrated systems.