This picture shows the extent and timing of the spray applications used for plant bug/ boll weevil control in a field presented earlier. The time bar shows 10 day interval following emergence. The first application is widely spaced from the second application. This reflects the common management practice prior to the release of the transgenic varieties do not spray early season at close intervals. Too many applications in the early season disrupt the beneficials that keep the Heliothines in check. The graph clearly shows this practice. However, later in the year, plant bugs and boll weevil populations expand and the spray interval needs to close in to keep these populations in check. Both the 1995 and 1996 farm records for this field show four applications and in approximately the same intervals of time after emergence. The spray layers are in different colors for convenience; however, by using a color code and a look-up table, the chemical class of the material used at any time can be illustrated. The graph does convey information about the rate. The thickness of each layer is similar; and here, the thickness represents Bidrin at 1 gal/ 20 acres.
This picture portrays the results from the 1999 experiment. Results from the 2000 effort indicate similar conclusions, but have not yet been graphed. The time bar again shows time in 10 day increments after emergence. During 1999, a transgenic variety was grown in this field. Since this technology was available, I was more comfortable in closing the interval between the early season sprays. The BT cotton would control the Heliothines. The date of the first spray application is similar to that of 1995/96. The chief difference is that this time, we did not spray at the full rate everywhere in the field (as indicated by the regions in gray). Notice that the first two applications are identical, but the third differs. By the time of this application, the imagery was indicated that the cotton in some of these previously unsprayed areas had grown into a state where they were now at risk to plant attack. The sample data at this time indicated that the first two applications had sequenced the age distribution of the plant bugs into a narrow cohort comprised of late instar nymphs or newly emerged adults. Since adults have wings and can fly, I did not want to have new colonization in these previously unoccupied areas. We widened the prescription coverage to the extent indicated by the imagery at this time of the year. We applied the application. It worked. Notice we did not need the historical fourth application.
This is the application map derived from image processing and integration with scouting information. The areas in a red hue are the no-spray, (or reduced rate, depending upon the insect species and its severity) regions of the field. The areas in green are the spray, or full-rate, regions of the field. This map is derived from a spatially registered image and is in turn also registered to the 'earth'. Using GIS software, the map can be processed into a file read by the controller on the sprayer. As GPS equipment on the sprayer records the position of the machine, commands are given to valves and pumps to allow, reduce, or stop the application in different locations throughout the fields.
This shows that it can be particularly useful to create a 'mask' which covers the spectral information in some areas of the field and to the desired extent. We can visually, or subjectively examine the complementary area where the spectral information is allowed to be displayed or printed (if in hardcopy) as shown here with the spray off areas. The graphic can be examined in light of the scouting data acquired at several locations in the same area. By inference, the conclusion can be reached is this a 'good or bad' map. The conclusion for this map was that it was a good map.
This masking next to the spray off areas allow the spray on areas to display their spectral information to allow an evaluation of the complementary areas. Examined this way, we reached the same conclusion; the map is a good one. However, if the judgement was reached that the map is poor, then further processing would be needed. For example, the stronger yellow areas may not need to be sprayed. So, scouting efforts could be employed to check only the more yellow regions of the field. If it was found that they could be omitted, then the application would be revised accordingly. Once a map is ready, it is uploaded into the controller, the sprayer is loaded with the desired chemicals, and put to work in the field. The GIS/GPS equipment on the sprayer determines in which area of the map the machine is located and directs the application of the agri-chemicals as recorded by attributes that describe that location on the computerized map.
All systems have limitations, and a bottleneck is the chief one. Many disciplines and concepts have to be integrated to comprise the development of an optimal farm policy. Each area of input has tremendous information content, and assembling all data into a unified unit is a fortuitous task. There are even subject areas not shown on the this graphic that have to also be considered. I consider this to be the most limiting challenge to the wide-spread adoption of these wonderful technologies recently available to cotton production.