Search

Surface reconstruction for scattered data is an ill-posed problem and most computational algorithms become overly expensive as the number of sample points increases. This article studies an effecti...

Field trials conducted in 2008 and 2009 investigated whether plot size affects incidence of white flower anther injury by tarnished plant bug (Lygus lineolaris (Palisot de Beauvois)) in host plant resistance (HPR) evaluations. The three cotton lines evaluated in the trial included a susceptible frego bract line (RBCDHGPIQH-197) and 2 standards, SureGrow (SG) 105 and Deltapine (DP) 393. Samplers monitored white flower anther injury between single row mini-plots embedded within multiple row max-plots. A sub-section of the max-plots was sprayed with insecticides to evaluate these tactics on altering the incidence of white flower anther injury. Plant bug numbers were very low in 2008, while infestation levels were higher in 2009. Significantly higher numbers of flowers with anther injury were observed in both years in the susceptible frego bract line compared to SG 105 and DP 393 lines. In both years, anther injury levels were similar in the max- and mini-plots, with lower levels observed in max-sprayed plots. The white flower monitoring procedure is a consistent indicator of adult plant bug preferences and is not influenced by plot size or interspersions of cultivar lines among plots.

An algorithm is presented to fuse the Normalized Difference Vegetation Index (NDVI) with Light Detection and Ranging (LiDAR) elevation data to produce a map potentially useful for site-specific management practices in cotton. A bi-variate Gaussian probability density distribution is modified to predict an improper probability distribution that also incorporates categorical variables associated with quadrant direction from the population means for the NDVI and elevation data layers. Water availability, influenced by slope and relative changes in elevation (as captured by the elevation data layer), affects crop phenology (as captured by the NDVI data layer). Thus, this fusion procedure results in a map potentially describing the joint effects of NDVI and elevation on cotton growth in a spatial and temporal way.

High density light detection and ranging (LIDAR) imaging has been shown to be able to define yield stability areas of a field for multi-cropping. Since LIDAR imaging is expensive and not widely available, it was hypothesized that medium resolution GPS elevation data which is commonly collected with variable rate technology (VRT) controllers and crop yield monitors could be used in lieu of LIDAR imaging. If proven, growers would be able to construct yield stability maps of their fields without the expense of obtaining LIDAR imaging. After substituting medium resolution GPS elevation data derived from the crop yield monitors, the procedure developed for developing a crop yield stability map was invoked and tested. The hypothesis that medium resolution GPS data could be used in lieu of LIDAR data was found to be invalid as the map generated incorrectly identified both high yield and medium yield areas of the field as low yielding areas as well as the inverse. While disappointing, high resolution GPS data from real-time kinetics (RTK) systems is yet to be tested and may offer an additional avenue to developing crop yield stability maps.

This paper proposes that yield stability patterns exist for multiple crops planted on the same land area over a period of years that growers can use to their advantage in planning crop management strategies using precision agriculture technologies. This study examines the relationship of soil elevation, slope, aspect and curvature to crop yield stability using a digital elevation model of the study area derived from a precise light detection and ranging (LIDAR) image of the farming area and surroundings. Three crop years of cotton and two crop years of corn yields were used to evaluate this hypothesis. The interpolation methods of Inverse Distance Weighted (IDW), simple Kriging and Natural Neighbor found in ESRI's ARCGIS were used to produce crop yield maps. These methods were also compared in the analysis. Simple Kriging gave the best R^2 estimates of yield as a function of elevation, slope, curvature and aspect. When the SAS FastCluster procedure was used to group yield points together using topographical features, the resulting regression analyses R^2 values of yield as a function of elevation, aspect, curvature and slope by cluster number were improved.

A better method is needed to evaluate the effectiveness of precision agricultural practices on research farm fields. We present a novel methodology for formulating the design of an experiment to evaluate the effectiveness of precision agricultural practices. The method combines a georeferenced treatment structure and a georeferenced design structure to build a mixed model that describes and analyzes the site-specific experiment. One or more layers of georeferenced information (obtained by various remote-sensing systems) describing the topography of the research field and its crop attributes may be included as covariates in the mixed model. The concepts of this approach are illustrated through the use of a hypothetical field. Current limitations are also discussed.

Our research has shown that cotton insect pests, specifically tarnished plant bugs, Lygus lineolaris (Palisot de Beauvois) (Heteroptera: Miridae) can be controlled early season in commercial cotton fields in Mississippi, USA, using spatially variable insecticide applications. Technology was developed for using GIS-based map scouting and a technique called the line-intercept method for obtaining low-level insect population counts in both rapidly growing areas of cotton and poorer growing areas. Using these population characteristics in combination with heuristic knowledge of the cotton fields and with the GIS maps, a spatially sensitive map could then be developed that could drive a spatially variable insecticide application for the control of the insect pest. We outline the steps needed to develop an automated technology for overcoming the time-sensitive events for early season control of cotton pests. This technology not only includes software systems for processing multispectral images to spatially variable insecticide application maps for spray controllers in the field but also high-speed wireless local area network (WLAN) technology for automated delivery of these controller application maps and for acquisition of as-applied and harvest maps from the field.

In the processing of light detection and ranging (LiDAR) data, a step error is an abrupt change in estimates of elevation between adjacent strips and must be reduced before building a digital surface model (DSM) of elevation. Existing methodologies in the literature for removing this artifact require an analyst to (1) utilize the sensor and aircraft information of the LiDAR mission, (2) isolate homologous flat surfaces within regions of overlap of adjoining LiDAR strips to estimate the mean offset, or (3) a combination of the two. In this application involving an agricultural landscape, a different methodology was required because the necessary information from the laser scanner or internal navigation system (INS) of the aircraft was unavailable and it was not possible to successfully identify homologous flat surfaces. Therefore, a post-processing, quadratic optimization model was formulated to reduce step artifacts. Using statistics obtained from the geographic overlap of the strips with a benchmark strip, it was possible to determine from the elevation values of the LiDAR point clouds two quantities: the strip variance and the total variance. Using these values and related statistics, the optimization model estimated correction constants, called decision variables, that minimized the among-group variance of the adjoining strips. When the values of these decision variables are added to the point cloud elevations of their respective LiDAR strips, the systematic step errors among adjoining strips are minimized with respect to the elevations provided by the point cloud of the benchmark strip. Decision variable values ranged between -0.087 and 0.078m. The adjusted LiDAR strip point clouds were used to build a corrected DSM of a 638.2-ha agricultural landscape at a spatial resolution of 0.5m. The elevation range of the DSM is approximately 44-81m HAE (height above the ellipsoid), where the higher elevations are the tops of trees. Effectiveness of the optimization model approach to reduce the step errors was evaluated by comparing the DSM before and after adjustment. Several hillshade, gray scale image subsets, and profile plot comparisons between the before and after adjustment of the point clouds of the LiDAR strips illustrate the algorithm's performance in reducing step error effects.

The problem of analyzing georeferenced cotton pest insect samples when a large percentage of the counts are zero is examined. The use of appropriate statistical methods for their analysis is required. To demonstrate this, georeferenced samples (n = 63) of tarnished plant bugs (TPBs; Lygus lineolaris [Palisot de Beauvois] (Heteroptera: Miridae)) were analyzed by three statistical methods and the results were compared. Correlation analysis of the sample counts with 25 classes of cotton growth derived from an unsupervized classification of multispectral imagery was followed by a complete enumeration analysis comprising three scenarios. The first scenario assumed the insect samples were unstratified. A distribution of sample averages was created by complete enumeration of all combinations of samples taken four at a time. The second scenario used imagery of the cotton fields to allocate the samples among three cotton growth categories (marginal, good or best) derived by a supervized classification of the 25 unsupervized classes. The insect samples associated with these categorical habitats were completely enumerated using allocations of 4, 6, 8 or 10 samples at a time from various sample sizes to determine how different allocations affected the results. The mean was not affected, but the standard deviation decreased with increased allocation sizes in all habitats. The third scenario used the two observers and three habitat categories to create six additional complete enumeration distributions by allocating four samples at a time from groups of varying sample sizes. These enumeration distributions are non-parametric estimators of the sampling distribution of: (1) the sample averages of a given sample size when samples are taken from the entire field, (2) the sample averages of a given size when samples are taken from each cotton habitat or (3) the sample averages of a given sample size from samples taken from each habitat by each observer. To support the enumeration analyses, these insect samples were analyzed further by Poisson regression models. These models showed significant differences between TPB counts by the two observers and among the habitats, whereas the observer by habitat interaction was not significant. For every combination of observer and cotton growth category, a Poisson regression model estimated the mean rate of TPB numbers. These means were similar to the corresponding modes of the complete enumeration distributions. The two non-standard methods showed that TPB numbers differed by habitat categories even though there were samples with a zero count, whereas a correlation analysis failed to identify a relationship between TPB sample counts and unsupervized habitat classes.

Designing experiments involves several processes. The first process identifies the experimental units generated by conducting the experiment. The second process is the application of planned treatments to the experimental units. Other processes are the analysis of the experiment and interpretation of results. While traditional experimental small plot designs are useful for investigating many aspects of cotton production, on a large scale, like an entire commercial field, they are difficult to implement because of routine farming operations and topographical variability. As an alternative, by using the capabilities of a variable-rate controller, it is possible to intersect the swath width of the largest farm implement’s boom (or tool bar) along its geo-referenced paths of travel with one or more geo-registered field zones to create experimental units having different shapes and sizes. Defining the experimental units in this manner establishes a site-specific experiment throughout the entire field. Spatial information recovered by Geographic Information System (GIS) processing from these asymmetrical experimental units, when coupled with general linear mixed model methodology, permits the assessment of effects on geo-referenced yield points due to topography, the site-specific and/or traditional farm management practices, and various interactions among them. In this paper, a general methodological approach for analyzing field-sized site-specific experiments is developed and described. An application is demonstrated by analyzing an unreplicated cotton variety trial that included a single site-specific application of a plant growth regulator.

When insect population density varies within the same cotton field, estimation of abundance is difficult. Multiple population densities of the same species occur because cotton fields (due to edaphic and environmental effects) are apportioned into various habitats that are colonized at different rates. These various habitats differ temporally in their spatial distributions, exhibiting varying patterns of interspersion, shape and size. Therefore, when sampling multiple population densities without considering the influence of habitat structure, the estimated population mean represents a summary of diverse population distributions having different means and variances. This single estimate of mean abundance can lead to pest management decisions that are incorrect because it may over- or under-estimate pest density in different areas of the field. Delineation of habitat classes is essential in order to make local control decisions. Within large commercial cotton fields, it is too laborious for observers on the ground to map habitat boundaries, but remote sensing can efficiently create geo-referenced, stratified maps of cotton field habitats. By employing these maps, a simple random sampling design and larger sample unit sizes, it is possible to estimate pest abundance in each habitat without large numbers of samples. Estimates of pest abundance by habitat, when supplemented with ecological precepts and consultant/producer experience, provide the basis for spatial approaches to pest control. Using small sample sizes, the integrated sampling methodology maps the spatial abundance of a cotton insect pest across several large cotton fields.

The hope of precision agriculture is that through more precise timing and usage of seed, agricultural chemicals and irrigation water that higher economic yields can occur while enhancing the economic production of field crops and protecting the environment. The analyses performed in this manuscript demonstrate proof of concept of how precision agriculture coupled with crop simulation models and geographic information systems technology can be used in the cotton production system in the Mid South to optimize yields while minimizing water and nitrogen inputs. The Hood Farm Levingston Field, located in Bolivar County, Mississippi, next to the Mississippi River, was chosen as the test sight to obtain a one hectare soil physical property grid over the entire 201 ha field. The 1997 yield was used as a comparison for the analysis. Actual cultural practices for 1997 were used as input to the model. After the 201 simulations were made using the expert system to optimize for water and nitrogen on a one hectare basis, the model predicted that an increase of 322 kg/ha could be obtained by using only an average increase of 2.6 cm of water/ha and an average decrease of 35 kg N/ha.

This study examines the diapause response in naturally occurring boll weevils under field and simulated field environments of north Mississippi. Squares containing early-stage weevils were collected in July, August, and September and subsamples from each group were installed into similar dynamic environments in the laboratory. In this manner, some weevils experienced uninterrupted photoperiods and temperatures between the field and laboratory (controls), and others experienced a shift forward in time or backward in time between the field and laboratory (treated). Results indicate selective sensitivity among individuals to diapause-inducing or averting daylengths and temperatures during early life stages independent of later stages and during late stages (including adults) independent of earlier stages. For example, it appears that some individuals are sensitive primarily during the early or late life stages, and depending on the environmental cues received at these times, they may or may not develop the diapause phenotype as adults. However, the rates of gain or loss in the acquisition of diapause depend on the intensity and duration of the token stimuli during part or all of the life cycle. If the proper cues are sustained throughout the life cycle, then the expression of diapause in the population will be maximized. Alteration in cues at any time will increase or decrease the percentage of weevils in diapause, with the relative effect related to the intensity and duration of the original stimuli. Once they receive a threshold to diapause-inducing stimuli, populations appear to be more responsive to environmental change. These results may explain some of the variability observed in diapause among studies.

This study examines the relative importance of age at the time of irradiation on attractiveness, mating ability, sperm transfer, prior mating, and longevity as factors of competitive- ness in sterile male boll weevil, Anthonomus grandis grandis Boheman. The amount of sperm transferred by irradiated males appeared to be the most important factor in competitiveness. More sperm was transferred by virgin males irradiated on day 5 than by virgin males irradiated on day 2, and males irradiated on day 5 had greater impact on egg hatch than those irradiated on day 2. The amount of sperm in spermlathecae of females mated to virgin males irradiated on day 5 was indistinguishable from that in females mated to virgin control males. Mating ability of males of all treatments was similar. Comparable numbers of boll weevils were captured in traps baited with males irradiated at 2 or 5 d during the first 4-5 d after irradiation, but thereafter, generally more weevils were captured in traps baited with males irradiated at 2 d or with control males. Attractiveness of males irradiated at 2 d was generally comparable to that of control males. More than 91% of irradiated males individually caged on cotton, Gossypium hirsutum L., plants lived for 10 d in the field, and 40% lived for 14 d; all individually caged control males lived 14 d. When released into cotton fields, however, the numbers of both irradiated and control males declined sharply over 14 d. Thus, the potential for an effective 2-wk life span in the field suggested by the caged study did not appear to apply to laboratory-reared weevils released into cotton fields.

The usefulness of deterministic simulation models has increased in agricultural applications. This increase must be matched with the use of analytical tools that can verify and validate their performance and help interpret their output. Values of model input variables can be selected by the analyst and are analogous to the concept of factors in conventional experiments. However, deterministic models cannot produce independent replications of selected input variables; many statistical analysis procedures cannot therefore be utilized. Consequently, this presentation describes the use of techniques applicable to the analysis of unreplicated, two-level factorial experiments. It is demonstrated that these concepts are a solution to the problem of how to evaluate deterministic simulation model performance and analyse deterministic simulation experiments.

Several common, nonparametric methods of analysis exist for determining significance among survival distributions of K groups or populations. One problem with these methods is that they cannot determine which survival distributions are different from one another if the global hypothesis of equality is rejected (that is, if all distributions are not equivalent). We discuss the answer to this problem and illustrate the use of a little-known, nonparametric approach for making multiple comparisons among K > 2 survival distributions. This procedure can be used with survival data that include censored observations. An observation on the survival] time of an individual insect is censored if the study is concluded before death occurs, or if the insect is lost (for example, escaped) for reasons not related to the experiment. The use of the Sidak inequality to adjust the comparison-wise error rate is discussed. The procedure is illustrated using 2 entomological examples.

Separate executions of deterministic simulation models are not independent replications in the strictest sense and, therefore, it is difficult to statistically analyze model output, since estimates of the error variance for the model are difficult to obtain. However, an analysis of simulation output is desirable. Presented here is a method that permits the statistical analysis of deterministic simulation output or its comparison to other output or experimental data. These analyses are possible when conditions imposed on the computer-generated experiment conform to a two-way treatment structure. The response data can be comprised of (i) model output, (ii) differences between experimental means and the model, or (iii) differences between individual output across treatment levels. The additive main effects and multiplicative interaction (AMMI) model for unreplicated data supplies several statistical tests. Foremost, the characteristic root test indicates if interaction exists between the two treatments. If interaction exists, the AMMI model can be used to determine which treatment combinations are involved. Once sources of interaction have been identified, it is then possible to obtain a good estimate of the error variance, even though these data are unreplicated. Using this estimate of the error variance, it is possible to test hypotheses of interest on model performance. An AMMI analysis comparing the output of a deterministic soil hydrology model to (replicated) experimental data is presented.

Jeffrey Willers, Darrin Roberts, Charles O’Hara, George Milliken, Kenneth Hood, John Walters and Edmund Schuster Genetics and Precision Agriculture Research Unit, USDA-ARS, Mississippi State, Mississippi, Department of Plant and Soil Sciences, Mississippi State University, Mississippi, Geosystems Research Institute, Mississippi State University, Spatial Information Solutions, Starkville, Mississippi, Department of Statistics, Kansas State University, Milliken Associates, Manhattan, Kansas, Perthshire Farms, Gunnison, Mississippi, InTime, Inc., Cleveland, Mississippi Aggeos, Inc., Fulton, Mississippi, Laboratory for Manufacturing and Productivity, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA

Genetic algorithms were used to automate the calibration of complex components of models. The use of genetic algorithms for calibration has broad applicability to both functions and large systems involving several functions. Genetic algorithms are a powerful tool recently developed to “evolve” solutions to problems characterized by massive search spaces. Genetic algorithms use methods analogous to the process of natural evolution. Chromosome analogs were used to represent values to be optimized. The algorithm used selection procedures and genetic operators to produce new chromosome analogs. A “fitness” function was used to evaluate the populations of chromosomes. For our research, fitness was indicated by how well a model simulated observed data. Highly “fit” chromosomes were saved and used to produce new generations and through an iterative process of selection and genetic rearrangement, highly fit strings were evolved that represented near-optimal solutions to parameterization problems. The evaluation of model “goodness-of-fit” affects the behavior of the genetic algorithm. The nature of these effects were investigated by using different indices of model goodness of fit. The effects on search (e.g., how fast an appropriate solution was found) were compared for four measures of goodness of fit. This research represents an innovative approach to solving an important problem, that of the calibration of mathematical models. Its usefulness is demonstrated on specific applications which allowed the determination of calibration parameters for a poikilotherm and photosynthesis models. The implementation of this genetic algorithm is proposed as a stand-alone framework for the development of similar applications for complex, process-based biological models that require periodic parameterization.

Advances in many areas of sensing technologies and the widespread use and greater accuracy of global positioning systems offer the prospect of improving agricultural productivity through the intensive use of data. By nature, agriculture is a spatial science characterized by significant variability in terms of yield and concentration of pests and plant diseases. Consequently, precision agriculture seeks to improve the effectiveness of various types of sensing information to give the grower more data and the ability to design the specific treatments for site-specific management of inputs and outputs. The intensive use of data in agriculture is at a relatively early stage and there remains much opportunity to refine modeling approaches and to build information infrastructure. With the overall goal of optimizing inputs to achieve the maximum output in terms of yield, this paper focuses on the application of a clustering algorithm to field data with the goal to identify management zones. We employ two sets of attributes, first yield and second field properties like slope and electrical conductivity to delineate the management zones. By definition, a management zone is a contiguous area defined by one or more features and may take on many different shapes. Building on the established machine learning approach of k-means clustering, we successfully identify a near optimal number of management zones for a cotton field.

With the proper technology and access to geographical information, it is more important to spend time developing an excellent classification scheme of a remotely sensed attribute of crop and forest vigor than to spend that time collecting multiple samples of insect counts . The ability to define zones from remote sensing images of crop or forest systems provides a vastly improved capacity to assess the sample variability of insect counts. Perspectives on defining zones from remote sensing information, including an examination of some relationships between these zones and insect sample counts, are discussed.

Here we present a methodology used to develop a yield stability map for a field. We proposed that there exist yield stability patters for commercial field crop production which growers can use to optimize crop production while minimizing inputs. The methodology uses multiple years of multi-crop yield monitor data and a high resolution, high density LIDAR (light detection and ranging using a laser light source similar to radar but much higher accuracy) digital elevation map of the field. These data are analyzed using ESRI GIS (geographic information system), ERDAS Imagine image processing, and SAS® (Company and brand names are used for information only and do not represent a recommendation or endorsement by the USDA-ARS) (SAS®2008). statistical analysis to produce detailed, multiple component GIS map which shows transitions for low yielding through medium yielding to the most productive high yielding areas of the field. The statistical procedures make use of krigging, cluster analysis, and simple quadratic regression to produce a statistically sound three-component map showing low-, medium- and high-yield zones in the field. The three-component map is based on a 3 m by 3 m pixel density and provides an immediate visualization tool for the grower to plan crop management practices.

There is a lot of interest in the concept of precision farming, also called precision agriculture or site-specific management. Although the total acreage managed by these concepts is increasing worldwide each year, there are several limitations and constraints that must be resolved to sustain this increase. These include (1) collecting and managing the large amounts of information necessary to accomplish this micromanagement, (2) building and delivering geo-referenced fine-scale (i.e., change every few meters or less) prescriptions in a timely manner, (3) finding or developing agricultural machines capable of quickly and simultaneously altering the rates of one or more agri-chemicals applied to the crop according to a geo-referenced prescription, (4) the need to have personnel stay “current” with advancements in developing technologies and adapting them to agriculture, (5) refining existing and/or creating new analytical theories useful in agriculture within a multidisciplinary, multi-institutional, and multibusiness environment of cooperation, and (6) modification of agricultural practices that enhances environmental conservation and/or stewardship while complying with governmental regulations and facing difficult economic constraints to remain profitable. There are many myths that overshadow the realities and obscure the true possibilities of precision agriculture. Considerations to establish productive linkages between the diverse sources of information and equipment necessary to apply site-specific practices and geographically monitor yield are daunting. It is anticipated that simulation models and other decision support systems will play key roles in integrating tasks involved with precision agriculture. Discovering how to connect models or other software systems to the hardware technologies of precision agriculture, while demonstrating their reliability and managing the flows of information among components, is a major challenge. The close cooperation of the extension, industrial, production, and research sectors of agriculture will be required to resolve this constraint.

This paper discusses the process of developing variable-rate treatment prescriptions and gives specifications for a prototype software system for implementing that process. The process is based on statistical analysis of data from embedded field trials, and incorporates producer preferences in determining a treatment prescription. The system can be used by researchers in agricultural research stations for developing prescriptions for commercial agricultural producers. The specifications provided are general enough to be implemented using a variety of statistical and database packages that are available to researchers. In addition to these specifications we provide online access to source code for implementing the system in SAS. We use this system to develop treatment prescriptions for a commercial cotton farming operation in northeast Louisiana. The prescriptions are based on data from a precision agriculture experiment conducted in 2006. The objective of that study was to compare the effects of five nitrogen rates on cotton lint yield across several soil types for the purpose of developing a variable-rate nitrogen treatment prescription for future use on that farm. Several possible producer preferences were incorporated with the results of the field trial to produce optional treatment prescriptions for the producer.

The use of soil and topography information to explain crop yield variation across fields is often applied for crop management purposes. Remote sensed data is a potential source of information for site-specific crop management, providing both spatial and temporal information about soil and crop condition. Studies were conducted in a 104-acre (42-hectare) dryland cotton field in 2001 and 2002 in order to (1) qualitatively assess the spatial variability of soil physical properties from kriged estimates, (2) compare actual yields with normalized difference vegetation reflectance indices (NDVI) obtained from multispectral imagery and from in situ radiometer data, and (3) predict site-specific cotton yields using a crop simulation model, GOSSYM. An NDVI map of soybean in 2000 obtained from a multispectral image was used to establish four sites in each low, medium and high NDVI class. These 12 sites were studied in 2001 and 12 more sites selected at random were studied in 2002 (n = 24). Site-specific measurements included leaf area index (LAI), canopy hyperspectral reflectance, and three-band multispectral image data for green, red, and near-infrared reflectance wavebands at spatial resolutions of 2 m in 2001 and 0.5 m in 2002. Imagery was imported into the image analysis software Imagine (ERDAS, v. 8.5) for georegistration and image analysis. A 6x6 pixels (144 m2) area of interest was established on top of each field plot site and digital numbers (DN) from reflectance imagery were extracted from each band for derivation of NDVI maps for each of four sampling dates. Lint yield from each plot site was collected by hand and also by a cotton picker equipped with AgLeader yield monitor and OmniStar differential global positioning system. We found plant height, leaf area index, and lint yield were closely associated with NDVI maps and with NIR band values acquired from either an aircraft or handheld (GER-1500) sensor during peak bloom in mid July. Results indicate NDVI and NIR bands could be used to produce estimated field maps of plant height, leaf area index and yield, which offer a potentially attractive mid-season management tool for site specific farming in dryland cotton.

High speed wireless data communication on farms has a large potential application which has already shown economical benefits in the two test farms where we have applied the technology. Because information is entered only once in hand-held devices connected wirelessly to the WLAN, transcription errors are minimized. Application maps which have been hand-carried by high-level technicians or professionals to the application equipment now can be transmitted directly to the equipment, saving significant amount of time and labor of high-level (high-cost) personnel. Similarly, data from the farm equipment can be transmitted directly to the farm base of operations at high speed saving time and money.

A MidTech TASC 6300 variable-rate spray control system was interfaced with a John Deere 4700 spray tractor. The dynamic response of the system was examined by measuring the response time between when a control signal was initiated and when the solution at the nozzle stabilized at the resulting concentration. These measurements were evaluated with three injection dyes at two tractor speeds and three application rates. Three FD&C dyes were chosen to simulate three chemicals in the injection tanks. A total of 1176 samples of solution from the nozzles were collected and analyzed with a spectrophotometer. Their absorbance characteristics were used to determine the response time of the spray system under various testing conditions. The results showed that the overall average delay time of the system was 38.3 seconds, and the rise time was 65.9 seconds. The response time varied with speed, but the product of speed and response time remained almost constant. The system took longer to achieve steady-state when the application rate varied from high to low than when it varied from low to high.