This year’s Spatially Variable Insecticide (SVI) research was based on the experience gained from the 2000 SVI studies conducted at Hardwick and Perthshire Farms (Spectral Visions, 2000). The concept of SVI is founded on the results of similar past studies, as well as the preliminary results of ongoing studies that are using remote sensing to support cotton insect management. These studies were researched prior to beginning the SVI experiments last year. During the literature review, we found that remote sensing to support cotton insect control has been pursued by a number of researchers, mostly at the level of exploratory or basic research. Due to the migratory nature of the insects which affect cotton production, producers and researchers feel that large scale or field level research is needed to fully explore, and understand the potential for utilizing remote sensing technologies to better manage entomological practices at the farm level. The natural environments and frequent movement of these insects cannot be properly duplicated or simulated in basic or plot level studies. We have not found reference to any applied research that uses remote sensing to spatially vary insecticide applications for plant bugs in Southern cotton. In fact, few researchers have taken the use of remotely sensed data to the applied ends envisioned by Ag20/20. One example of current research in cotton insect management was done by Nahum Gat of Opto-Knowledge Systems, is researching hyperspectral imaging for insect-infestations in western cotton (Gat et al., 1999). Perhaps the most thorough basic research for remote sensing in cotton insect management has been done by Fitzgerald et al., (2000) which examined the effects of spider mites and other insects in cotton through a variety of datasets, including thermal imagery. Spider mite damage was also examined through multispectral imagery by Fitzgerald et al., (1999), with some promising results deriving from bit-error filtering on the near infrared band. Another study by Summy et al., (1989) used color infrared photography to detect undestroyed cotton in support of a boll weevil eradication program.

One example of an applied remote sensing cotton insect project is a boll weevil eradication GIS that uses satellite imagery (Smith et al, 1997). Remote sensing in entomology was reviewed by Riley (1989), who discussed a wide range of attempts at insect mapping through aerial photos, radar, airborne and satellite digital imagery, and other data, including research of direct observation of flights of moths, grasshoppers, locusts, and other phenomena. A greater emphasis was placed on inference techniques, that is, detecting the effects or habitat of certain insects to infer their presence. Remote sensing projects for agricultural insect management at the time (Riley, 1989) included mention of corn leaf aphids, milkweed as an indicator of monarch butterfly presence, and post-harvest standing cotton plants as an indicator of boll weevil overwintering sites.

In addition, there are a number of research papers that approach site-specific management techniques for insect control, but not including remote sensing. Parker et al., (1999), for example, evaluated insect management strategies through the use of yield-monitor data but not imagery data. In summary, a review of the literature and the experience gained from past SVI research conducted at Perthshire and Hardwick Farms revealed an opportunity to continue to do groundbreaking applied experimentation in Spatially Variable Insecticide that may have a satisfactory effect on both Southern cotton producers and the remote sensing industry.