Grant

Justification

Improving NUE and minimizing N contamination of ground and surface waters are issues that are critical to the sustainability of US agriculture and central to the goals of IFAFS. This project will develop and transfer effective, science-based techniques to improve agronomic NUE, increase productivity and farm profit, while reducing N contamination of ground- and surface waters. Remote sensing and ground sampling of soil properties, crop biomass, N, and water status will be used for farm-scale in-season site-specific wheat and corn N management strategies based on an improved understanding of how soils, soil moisture, groundwater, fertilizer N, and crops interact across the landscape and over time. Uniform and site-specific soil- and crop-informed N management strategies will be incorporated into nutrient recommendation decision support tools for developing optimal N application timing and rate recommendations. Project objectives, progress, scientific discovery, and recommendations will be communicated to producers, students, and the general public through the media, the world wide web, field demonstrations, course modules, and liaison with the Neuse Crop Management Project. Extension activities will also provide important opportunities to assess user needs and solicit feedback regarding project activities. Project data will be utilized to train extension agents and crop consultants in the use of the Nitrogen Loss Estimation Worksheet, the NC Interagency Nutrient Management Planning Tool, and project-developed in-season, site-specific N recommendation decision support tools.

The proposed field-scale experiments will be carried out in an area and on soils that represent important and extensive Coastal Plain agroecosystems in NC and elsewhere in the Southeast. Fertilizer N use in these shallow-groundwater agroecosystems has been identified as a substantial component of nonpoint source N contamination of surface and groundwater. One likely contributor to this problem is the use of whole-field average yield goals for making N recommendations. Despite longstanding promotion of BMPs for N and promulgation of environmental regulations to improve water quality, N contamination of Coastal Plain waters continues at unacceptable levels. Increasingly restrictive environmental regulation of agricultural N use may threaten regional crop production and have significant negative impact on grower livelihoods and regional economies. This project will continue development of improved uniform and site-specific soil- and crop-based N management strategies, promote their adoption by growers, and document their impacts on water quality. Some elements of the improved N management strategies that the project will develop, test, and transfer will be specific to southeastern Coastal Plain wheat-soybean-corn cropping systems with controlled drainage. However, techniques developed to realize accurate, site-specific N timing and rate recommendations by integrating ground- and remotely sensed information on soils and crop status will be applicable to a variety of cropping systems around the world. This project will benefit growers by helping them to apply the correct amount of N to meet potential crop demand, thereby improving productivity and profitability. The research will benefit public welfare by reducing N losses to streams and waterways through more accurate predictions of crop N requirements.

Objectives

Ground- and surface water nitrogen (N) contamination from southeastern Coastal Plain agriculture is a regulatory and social issue threatening regional crop production. Our long range goal is to develop and evaluate remote sensing-informed, site-specific, variable-rate N management strategies for their potential to optimize N-use efficiency (NUE) and reduce water contamination in Coastal Plain cropping systems. Our specific objectives are to:

1. Compare conventional uniform N management to intensive strategies that optimize N rate, timing, and placement using remote sensing, crop, and soil information, on both a whole field and site-specific basis.
2. Evaluate management effects on NUE and soluble soil and groundwater N.
3. Continue development of N-management techniques and decision support systems based on remote sensing, crop, and soil information.
4. Demonstrate management strategies and consequences to producers, extension agents, crop advisors, students, the media, and general public through field expos, training programs, course modules, and internet delivery.

Secondary objectives include testing elements of the Nitrogen Loss Estimation Worksheet (NLEW) and the North Carolina Interagency Nutrient Management Planning Tool.

Approach

Research Approach

Trial Location, Field Design, and Treatments
This research will be located at the Lower Coastal Plain Research Station just north of Kinston, NC in the Neuse River Basin. The Lower Coastal Plain Research Station consists of two farms producing flue-cured tobacco, vegetables, corn, cotton, small grains, and soybeans.

A two-year winter wheat/double crop soybean (year 1) - corn (year 2) rotation typical of Coastal Plain agriculture will be established. and continued for the four year life of the project. The research site consists of a 14.7 ha field with four predominant soil types typical of the Coastal Plain. A Goldsboro loamy sand dominates in the NE and SW of the field. This soil transitions to Lynchburg and Norfolk sandy loams in the center of the field. The realistic yield expectation (RYE) database for Coastal Plain soils (Hodges, 2000) assigns RYEs for wheat of 3695, 4031, and 4367 kg ha-1 , and for corn of 7839, 7212, and 8152 kg ha-1, to Lynchburg sandy loam, Norfolk sandy loam, and Goldsboro loamy sand respectively.

Drainage tile lines spaced 30.4 m apart run north and south throughout the field. Lines in the northern and southern halves of the field drain to common lines that exit to the west. Drainage control structures will be placed at these two exit points to allow control of the water table. During the fall and winter when wheat is in the field, the drainage control structures will be managed to prevent flooding during tillering. In March, the structures will be managed to raise the water table as high as is practical for the soybean crop without adversely affecting wheat growth. The structures will be managed through the soybean and corn rotations to reduce the risk of drought stress in these summer crops as much as possible.

The field will be divided into a randomized complete block design with three N treatments replicated 10 - 12 times. These treatments will be: 1) Whole-field RYE management consistent with current regulations and based on the estimated RYE and the N-use factor for the predominant soil type in the field (Goldsboro loamy sand), 2) Whole-field in-season crop- and soil-based management based on field-averaged in-season estimates of optimal N rates determined by remote sensing and ground sampling, and 3) Site-specific in-season crop- and soil-based management based on site-specific estimates of optimal N rates determined by remote sensing and ground sampling. Each treatment "plot" will be 60.8 by 60.8 m or 0.37 ha. Plots will be located so that the drainage tiles run through the plot center. The exact placement and number of replicates will be defined after an intensive soil sampling is made.

Soil Evaluation and Monitoring

For the past ten years, this field has been fertilized (P and K) and limed based on soil type. Ten to 15 soil samples are pooled across each field area with the same soil type (Figure 5) and P, K, and lime applied accordingly to that area. This practice will continue throughout this study. In addition to sampling by soil type, intensive soil sampling on a georeferenced 15 by 15 m grid will be conducted in the Fall of 2000 prior to planting and in 2004 after soybean harvest. These intensive samplings will be used to characterized macro- and micronutrients, pH, soil texture and type. In addition to intensive grid sampling, soil electromagnetic induction (EMI) measurements of soil conductivity will be taken using a Geonics EM-38 on multiple field transects. The Geonics EM-38 can examine depths increments of 0 - 0.75 and 0 - 1.5 m. The NCSU Department of Soil Science intends to purchase a Geophex GEM-2, a multiple- frequency EMI instrument capable of examining multiple depth increments to ~10 m (Won et al, 1996). Electromagnetic induction data will be groundtruthed in part using soil cores extracted during the placement of groundwater sampling wells (see below). These cores will be analyzed in increments to characterize soil properties such as particle size distribution, CEC, and organic matter, bulk density, and their distribution over depth.

Soil- and Groundwater Sampling and Soil Moisture Monitoring

Two well nests will be installed in each plot for sampling groundwater and soil solution for determination of soluble N (NO3 , NH4) by Lachat ion analysis. Each well nest will consist of three screened PVC pipe groundwater sampling wells and three suction lysimeters. The three wells will be placed at different depths, the deepest to remain saturated year-round, the intermediate well to remain saturated during all but the driest periods, and the shallowest to be saturated only during the wettest periods. Groundwater samples will be collected from these wells biweekly and after significant precipitation. Similarly, the three suction lysimeters will be installed to sample soil solution periodically from three depths within surface horizons. Three TDR probes (Moisture Point, ESI) will be installed in each plot at georeferenced locations apart from the well nests for periodic determination of soil moisture in five depth increments from 0 to 120 cm. Using these probes, soil moisture will be measured weekly at solar noon from emergence to physiological maturity of the crop. To help define the amount of soil moisture present at field capacity, hourly TDR measurements will be made following a rainfall event of at least 1.27 cm using the techniques described by Braga and Jones (1998). The amount of soil moisture left at wilting point will be defined by the lowest TDR reading taken from that area of the soil profile when the crop shows clear visual symptoms of water stress.

Aerial IR Thermography

At the same time that weekly soil moisture is measured using the TDR probes, an aerial thermal IR image will be taken using a digital thermal IR as described above. Regression analysis will be performed to quantify relationships between site-specific canopy temperatures and soil moisture, and to delineate zones of low, medium, or high probability for both moisture stress and sufficient plant-available water. Based on research in progress, these zones will be used to help refine in-season estimation of yield potential and N requirements.

Soft Red Winter Wheat N Management

Winter wheat will be planted in October 2000, and 2002. Cultural practices will be consistent with land-grant recommendations for Coastal Plain conventionally tilled wheat. Fertilizer P, K, and lime will be applied by soil type as described above. Pre-plant N (22.4 kg N ha-1) will be applied uniformly. Controlled drainage will be managed to hold the water table as high as is practical for the subsequent soybean crop without adversely affecting wheat growth.

• Uniform realistic yield expectation N management: Consistent with regulatory N recommendation for a field with a predominant soil type of a Goldsboro loamy sand in the Neuse watershed, 123 kg N ha-1 will be applied to large plots in this treatment at GS 30.
  
  
• Uniform in-season crop- and soil-based N management: The GS 25 tiller density averaged across plots in this treatment will be used to determine if a N application is required at GS 25. If the average is less than 500 tillers m-2, 60 kg N ha-1 will be applied to all plots (Weisz et al., 2000). If the average is greater than 500 tillers m-2, no N will be applied. Tiller density will be determined from IR aerial photographs and minimal ground sampling of high and low density areas as described above (Flowers et al., 2000; Heiniger et al., 2000).
  
  Aerial IR photographs will again be taken at GS 30. Green NDVI will be used with ground sampling of tissue N concentration in several high and low N calibration strips placed periodically through the field. The green NDVI will be used to determine a mean N concentration across all large plots in this treatment. The relationship used in Figure 2 (Alley et al. 1996, Weisz 1998) will be used to determine a uniform N rate to be applied to all these plots.
  
  
• Site-specific in-season crop- and soil-based management: The same aerial photographs used to determine whole field in-season N rates will be used for this treatment. However, instead of averaging estimated tiller density or N concentration across plots, the site-specific values will be used. Nitrogen will be applied (in all treatments) with a variable rate spray applicator with a 7.6 m boom making it possible to apply N on a 15.2 by 15.2 m grid. Tiller density at GS 25, and N concentration will be estimated for each of these small grids within large plots in this treatment, and site-specific optimal N rates (Figure 2) applied accordingly.
  
  
• Harvest: Wheat will be harvested in early June 2001 and 2003. Harvest will be completed with a combine equipped with an AgLeader PF 3000 yield monitor and DGPS system. Multiple georeferenced grain samples will be collected from each plot for determination of test weight and and total N.
  
  
• Supplemental Research: In the Fall of 2000, five additional wheat fields located on farms throughout the state will be used to continue refining the site-specific in-season N management system and validation of the remote sensing methods used for determination of tiller density and N concentration. This represents continued refinement of the methods described by Flowers et al. (2000) and Heiniger et al. (2000) upon which this study is based.

Soybean Management

No-till double cropped soybeans will be planted following wheat (June 2001 and 2003). Controlled drainage will be managed through the soybean season to reduce the risk of drought stress. Cultural practices including fertility and pest management will be consistent with land-grant recommendations for Coastal Plain no-till soybeans (Lumpkins, 2000). Fertilizer P, K, and lime will be applied uniformly by soil type. Aerial thermography will be conducted periodically and during a drought. Soybeans will be harvested in October 2001 and 2003 as described above.

Corn N Management

• Continued Development of In-Season N Management Strategies: During the 2001 crop season, corn plots will be established at two irrigated and dryland locations apart from the rotation experiment to continue development of remote sensing methods for determining corn biomass, N content, and development of pre-tassel N management strategies. This work, already in progress, was described above (see section I.F). By the first entry of corn into our rotation in the 2002 crop season, we will have developed and tested a system to optimize pre-tassel N applications based on site-specific in-season knowledge of corn biomass, N content, moisture stress history, moisture stress/adequacy probability, and plant-available soil.
  
  
• Rotation Experiment N Management: Drainage control structures will be managed from soybean harvest through the following corn season to minimize the risk of drought stress. Corn will be planted consistent with land-grant recommendations for Coastal Plain no-till corn (Heiniger, 1999). Fertilizer P, K, and lime will be applied uniformly by soil type.
  
  
• Uniform realistic yield expectation N management: Consistent with the regulatory N recommendation for a field with a predominant soil type of a Goldsboro loamy sand in the Neuse watershed, the total N rate for corn will be 182 kg N ha-1. Consistent with NC recommendations for N fertilization of corn (Lumpkins, 2000), one fourth of the total N (46 kg N ha-1 ) will be applied at planting and the remainder at lay-by (V6-V8).
  
  
• Uniform in-season soil- and crop-based N management: The in-season N management system for corn will be based on three N applications: starter, lay-by, and pre-tassel. The critical determination of how much total N to apply will occur at V12 and will be based on the results of research in progress. Starter N will be applied at a rate of 46 kg N ha-1. At planting, several calibration strips (for aerial photography) will be established in the field with five plots at N rates of 0, 56, 112, 168, 224 kg N ha-1. At lay-by (V6-V8) and pre-tassel, color and color-IR images will be taken of the entire field using the same film and technique described by Flowers et al. (2000). Chlorophyll meter (Minolta Spad Meter) readings will be taken from the calibration plots, as will plant samples to determine biomass, percent tissue N, and total plant N through laboratory analysis. The lay-by N rate will be based on estimation of the N required to carry the crop through tasseling, or approximately 50% of the RYE rate. At pre-tassel, the final increment of the total N rate will be applied only if warranted based color and color-IR evidence. Pre-tassel N will be applied using a "highboy" and drop nozzles.
  
  
• Site-specific in-season soil- and crop-based management: The same aerial photographs used to determine whole field in-season N rates will be used for this treatment. However, instead of averaging estimated N needs across plots, site-specific values will be used. These values will be based on pre-tassel N and biomass, and from information on site-specific risk of drought stress gained through aerial thermography and soil moisture measurements made earlier in the season and during previous crops. Nitrogen will be applied (in all treatments) with a variable rate spray applicator with a 7.6 m boom making it possible to apply N on a 15.2 by 15.2 m grid. Harvest: Corn will be harvested (as described above) in 2002 and 2004. Multiple georeferenced grain samples will be collected from each plot for determination of test weight and and total N.

Data Analysis and Management

Image rectification, analysis, and classification will be done using ERDAS Imagine (ERDAS, Atlanta, GA.) and IDRISI (Clark Labs). Geostatistical analysis and interpolation of all spatial data will be done using SAS (Cary, NC) and GS+ (Gamma Design Software, Plainwell, MI). Georeferenced yield monitor data will be visualized using ArcView (ESRI, Redlands CA), which will also be used for all GIS functions. Individual large plot yields will be determined by averaging all yield values within that plot after eliminating a buffer around the plot edges. Treatment effects on yield, grain quality, NUE, groundwater N, and profitability will be analyzed using classical analysis of variance techniques and mixed-model approaches incorporating estimates of spatial covariance. An economic analysis of treatments will be done using wheat, corn, and soybean enterprise budgets.

Extension Approach

A very significant impact of the project will be realized through the NC State University College of Agriculture and Life Sciences Geographic Information Systems Education Laboratory. This dynamic and well-equipped laboratory provides education and extension training in computer-based information technologies for site-specific nutrient management. Increasing the adoption of decision support systems that rely on remotely sensed data requires an aggressive educational program to increase the population of skilled users of remote sensing data and products. Computer-based educational programs will be developed to demonstrate successful and unsuccessful examples of merging multi-temporal, multi-resolution imagery with field observations in agriculture. Annual workshops for producers, crop consultants, industry personnel, and research and extension faculty will be held on a cost recovery basis.

Soil Science Extension faculty have been working with state agencies to develop the NC Interagency Nutrient Management Planning Tool for creating nutrient management plans. This project will provide data for validating and demonstrating this software. Perhaps most important, our results will provide direct feed-back to the state agencies mandating these N rates.

Project investigators and senior associates will develop case studies based on the proposed conventional and soil- and crop-informed management treatments. We will use these case studies to demonstrate N management strategies and consequences during Extension activities such as field days and meetings. The case studies will be adapted for classroom and internet education, and form the core of a teaching module on using soil- and crop-informed N management in writing nutrient management plans that will be added to the syllabus of the undergraduate class on precision agriculture, "GIS in Agriculture and the Environment". This module would also serve as a stand-alone training on write nutrient management plans.
At the initiation of the project, a World Wide Web site will be developed to communicate with the public about the purpose of the project. Throughout the duration of the project, the web site will be regularly updated. The site will be linked to departmental offerings in soil fertility, nutrient management plan training, precision agriculture, and using GIS as an agricultural management tool. Training will be offered on the World Wide Web through distance education programs. The Department of Soil Science already offers courses on-line, including Soil Fertility. The precision agriculture course, will be available on-line in August, 2000.

Additional audiences will be reached through traditional Extension web and paper publications. Extension bulletins will be prepared based on findings of this project that will educate producers on the use of soil- and crop-informed management as additional BMPs to meet Neuse goals and improve their agronomics.

The Neuse River Basin is the focus of a partnership among growers, industry, and NC State University to reduce N losses from farms. The Neuse Crop Management Project was set up with three years of funding from the Pew Charitable Trust, the USEPA, and the NC Clean Water management Trust Fund. Currently, the Neuse River Education Team is aggressively communicating the potential impacts of numerous BMPs on water quality through publications, producer meetings, and field demonstrations. In cooperation with the Neuse Education Team and the Neuse Crop Management Project, this proposed project will educate local growers about BMPs that improve their bottom line and protect the environment.

As part of the Neuse Crop Management Project, working relationships have already been established with demonstration farms. Water quality data from shallow wells and ditches collected under the Neuse Crop Management Project funding will provide a baseline for monitoring changes in NO3 movement under these soil- and crop-specific precision agricultural activities. These demonstration farms will be available for field tours, Extension agent training, and other educational opportunities explaining how site-specific agriculture can be used to meet the Neuse Rules in the third and fourth years of this funding.

Evaluation and Monitoring

An independent team consisting of NC State University faculty, NC Grower Association board members, county extension agents, and regulatory officials will be recruited to review and evaluate the project on a yearly basis. In addition to project evaluation, we will work with this team to help identify objectively verifiable indicators of project progress toward goals and objectives. The original objectives of this project will provide the foundation for its evaluation.

• Objective 1: Comparison of N management strategies. Grain yield, NUE, and profitability associated with each strategy will be computed. These data will provide clear evidence of the success or failure of this objective.
• Objective 2: Evaluation of N management effects on NUE and soluble soil and groundwater N. Goundwater samples collected on a bi-weekly basis (and following significant rainfall) throughout the project will provide clear evidence of the success or failure of each of the N management strategies.
• Objective 3: Continue development of N-management techniques and decision support systems based on remote sensing, crop, and soil information. The first evaluation point will be Fall 2000, when validation of our remote sensing methods for determining tiller density, and the second year of remote sensing for wheat N concentration will be completed. The second major evaluation point will Summer 2001, when the first year of the wheat program, and the second phase of developing and testing the corn remote sensing system will be completed. The ultimate evidence of success of this objective will be if our research has a direct impact on the NC Interagency Nutrient Management Planning Tool.
• Objective 4: Technology transfer. Several direct measures of the success of this objective are built into this project. Since the NC Interagency Nutrient Management Planning Tool will be used on all farms implementing nutrient management plans, the degree to which this research is incorporated into this decision support tool will provide a direct measure of technology transfer.
  

Another direct measure of success will be the degree to which the Neuse River Education Team adopts our research results into their educational programs. Incorporation of our results into industry, county agent, producer, and university student training and educational modules will provide evidence of success. We plan to disseminate the results of this project through the publication of refereed journal articles and extension materials. Completion of high quality publications will provide an important indicator of project success. On-going project monitoring will be part of monthly project meetings attended by participants including all graduate students, the research associate, and the primary investigators. These regular meetings will be used to address problem solving, report progress, and plan for the coming weeks.