6.0      RESEARCH METHODS - Overview

    This Phase III proposal consists of 10 sub-projects in support of the overall project objectives as listed in Section 4.0.  Specifically, the alignment of the Phase III sub-projects with the project objectives is summarized in Table 1.  Further, a majority of the sub-projects are multidisciplinary in nature with linkages between sub-projects in Phase III, and to sub-projects in Phases I and II.  Section 8.0 outlines a scheme to coordinate project activities for the specific express purpose of disseminating research findings directly to producers in the Commonwealth via the Internet, with backup to be provided using the traditional contacts through the county agriculture agents, meetings and field days.  The following sub-projects summaries provide an overview of work proposed for Phase III.

Outreach Education for Precision Agriculture - One of the many challenges faced by precision agriculture users is managing spatial data.  Too often, the available data sets are not projected in compatible coordinates.  In addition, the software tools to manage the data can be complex and cumbersome.  The goal of this project is to develop an outreach program that will help producers learn to use precision agriculture software (Geographic Information Systems and mapping packages) to properly manage and analyze spatial data.  The outreach program will address issues of data availability and compatibility while providing hands-on educational experience with field data collection, building spatial farm databases, and data analysis.

Kentucky Precision Agriculture Education Consortium - This proposal details the development of a Kentucky precision agriculture education consortium consisting of a multi-disciplinary team of faculty from Murray State University, Western Kentucky University and the University of Kentucky.  The focus of this project is to jointly develop and implement standalone educational modules that are web-based and can be used to teach a dedicated course on precision agriculture, or selected elements could be incorporated into existing courses.

Sensors and Variable Rate Management - Variable rate nitrogen has the potential to improve the profitability of agriculture and reduce environmental losses of nitrogen in Kentucky.  Although previous work has shown that soil type and depth can be used to predict crop response to nitrogen, soil surveys were not created at a scale to micro-manage this variability.  Therefore, we are working to develop methods to utilize remote and ground based sensors to help predict where in fields there will be an agronomic response to nitrogen.  We have two objectives in this proposal: 1) to develop methodologies to collect and calibrate remote and ground based sensor data for precision agriculture applications and 2) to develop methodologies for utilizing sensors for variable rate nitrogen management.

Variability in Grain Crop Yields Based on Landscape Position and Other Attributes - For the past five years, corn yields have been measured in strips across fields where seeding rates were varied according to landscape position at multiple locations.  Landscape positions were used as a surrogate for soil type and erosion as well as differences in available water.  During the past two years, N was also varied at these same positions on a parallel strip.  Yields were affected by both variables resulting in increased net returns with the largest increase attributed to nitrogen.  The objective of this proposal is to evaluate the effect of variable inputs on whole fields.  Correlations between yield and landscape attributes will be used to set seeding and nitrogen rates.  These parameters will be evaluated on a whole field basis.  Economic evaluation will be compared to a similar field where seeding and nitrogen rates are held constant.  This study will be done initially on fields with rolling topographies.

Assessment of Alternative Methods of Applying Precision Deep Tillage - Adverse soil compaction is evident on a 204 acre farm to be cropped using a rotation of notill corn, soybeans and wheat/double-crop soybeans.  The farm will be divided into 1-acre grid cells and 15 soil cone index (CI) measurements will be recorded for each cell.  Remote NIR images of the farm will be acquired using a small remote-controlled aircraft following rainfall events of 4-5 cm to identify areas in which reduced drainage caused by excessive soil compaction is indicated.  Veris electrical conductivity measurements will be made in each cell and evaluated as a means of identifying adverse compaction.  Precision tillage will be applied with shanks on 760 mm centers to a maximum depth of 400 mm on the farm as follows.  The 1-acre grid cells will be randomly assigned to four groups.  The corresponding tillage treatments will be: 1) to the maximum depth at which the mean CI > 1.5 MPa; 2) where below-average corn yield was measured in 2000 and remote imagery and/or Veris measurements confirms likely soil compaction, to the maximum depth at which mean CI > 1.5 MPa; 3) in cells that meet the same criteria as (2), except to a uniform depth of 250 mm; and 4) no tillage.  We expect crop yield to increase in areas of the farm where CI > 1.5 MPa is measured and remedial tillage is applied. Directed sampling of areas of suspected compaction as indicated by remote imagery, Veris measurements and yield map analysis would minimize the cost of CI analysis.  Applying tillage at a uniform depth would make CI analysis unnecessary.  Our objective will be to determine the method of applying remedial tillage that results in the greatest increase of net crop returns.

Equipment Enhancement to Support Variable-Rate Response Surface Development - The primary focus of precision agriculture technology is variable-rate delivery of inputs to optimize the production from specific locations within a field given existing soil water availability, soil type and structure, and nutrient levels.  Heretofore research in this area has been conducted with field-scale machinery.  It must be recognized that much of the precision agriculture equipment has been sold as aftermarket modifications.  To this end researchers have ignored shortcomings of the modified equipment.  These shortcomings include errors in yield estimates arising from less that desirable sensor designs and locations, as well as the errors associated with rate changes and distributions patterns of inputs (e.g. fertilizer application via spinner spreaders, retrofitted yield monitors).  The cumulative effect of these errors is that many research findings are inconclusive.  Acquisition of the equipment as outlined in this proposal, a combine and granular fertilize applicator, will enable researchers at the University of Kentucky to conduct more controlled field experiments involving the variable-application of granular and liquid fertilizers, and variable-rate seeding, jointly with Kentucky producers.  Further, these capabilities should reduce the negative impact of this work on the timeliness of the routine field operations of the cooperators.  Further, the acquisition of the proposed equipment will increase the quality of crop response data through better matching of yield sensors to existing grain combine geometries, specifically the force impetus sensor.  The resulting equipment compliment (existing and new equipment) will better position researchers for intensive field investigations that include seeding and nitrogen rate interactions in corn, along with the development of farm-specific phosphate and potash fertilizer response relationships for grain crops.

Voice Recognition for Concurrent Field Scouting and Machine Operation - Yield monitors and other sensing technologies such as satellite images and soil conductivity measurements are providing producers valuable information about soil landscapes and yield performance.  Unfortunately, the single greatest source of performance information, human observations, remains untapped since a system does not exist to easily reference and transcribe information during field operations without interfering with normal machine operation.  Many farm managers are finding it difficult to perform their own field scouting during the growing and harvest seasons.  Hired helped is not motivated to record/transcribe scouting notes during field operations. This investigation seeks to perfect the use of voice recognition for field scouting concurrently with spraying and harvesting operations.  Current voice recognition software will be modified and a user interface developed to adapt the noisy farm environment.  The system will be evaluated for simultaneous crop performance data collection under field conditions by various cooperators to assess its ability to be integrated into existing farm operations.  The outcome of this investigation yields a technique for Kentucky producers to collect spatial scouting data regarding crop performance for export into most agricultural GIS packages to aid in management decisions.

Break-Even Analysis and Interest Rate Considerations in Precision Agriculture Adoption - The relative profitability differences of precision agriculture are not always clearly or easily determined. The costs of adoption must be offset by increased yields or lower input costs. Break-even analysis with results expressed in terms of yield differentials required to cover additional costs is a straightforward way for farmers to assess precision agriculture adoption choices.  Additionally, all the benefits may not be directly attributed to the individual producer.  Environmental and social benefits that do not accrue to the producer may lead to under investment by producers.  Mechanisms to address this situation equitably include cost sharing or interest rate buy downs on precision agriculture capital investments.  This project will develop quantitative tools using enterprise, partial, and break-even budgeting techniques to address these questions and issues.

Optimal Management Zone Delineation - The identification of profit maximizing management zones, including optimal uniform grid size, is a complex issue central to the successful implementation of variable rate input application.  This vastly important decision has alluded experts but a novel modeling procedure that both identifies the most profitable management zone or grid size and permits economic comparison of alternative decision rules to determine such zones is presented in this proposal.  A multidisciplinary (agricultural economics, agricultural engineering, agronomy) approach is embedded in a mathematical programming model which maximizes profits and includes concepts of nonlinear and integer programming. Actual comparison of alternative decision rules will be made for case studies in Kentucky.  It is hoped that model results will aid in improving management zone delineation rules and lead to the development of farm level decision rules including risks faced by producers.  Consequently, the research proposed in this project lies at the very heart of precision agriculture and represents a first stage of addressing this problem.

Economic Advisory Aids for Precision Agriculture Users - Precision agriculture technologies have the potential to overwhelm producers with information while not providing clear, concise solutions to production or economic problems. This project proposes to develop methods and tools to assist farmers with economic decisions based on yield and input data from precision agriculture data. Procedures will be developed to generate net returns maps, risk maps, and CRP enrollment maps from farmer’s yield map data submitted electronically to UK College of Agriculture personnel.  Resulting maps and recommendations will be transmitted back to the participating farmers.  This project will develop and test these procedures.  If demand is adequate, the potential for offering these types of services on a fee basis analogous to soil testing services will be assessed in a cost/benefit analysis.