Increased cancer risk, adverse reproductive outcomes, and neurotoxicity are among health effects that have been associated with exposure to certain agricultural pesticides. For researchers studying the potential health effects of agricultural pesticides, farmers and other pesticide applicators are relatively easy to identify. But pesticides can drift from their targeted fields, creating hidden, exposed populations in the community. As these populations are usually unaware of what pesticides are being used near their homes, traditional epidemiologic methods may not be feasible. By using GIS, potentially exposed residences can now be identified and mapped in relation to crop fields. Based on the type of crops near homes, researchers are gaining a clearer picture of who might be exposed to which agricultural pesticides.

With a combination of remote sensing data and historical records on crop location, Dr. Mary Ward of the Occupational and Environmental Epidemiology Branch, National Cancer Institute (NCI); Colorado State University's Dr. Jay Nuckols; and their colleagues are using GIS to help identify homes that may have been exposed to specific agricultural pesticides.

"These are studies you can't do with a questionnaire, because people don't know which chemicals have been used on nearby fields," says Ward.

In a January 2000 Environmental Health Perspectives article, Ward, Nuckols, and their coauthors laid the groundwork for using satellite imaging to perform exposure assessment. An integrated system incorporating both GIS and image processing functionality was constructed using GIS software with which they had some history—ArcInfo, ArcView, and Esri Business Partner Leica Geosystems' ERDAS IMAGINE software. Researchers gathered imagery for two summer dates in 1984, covering three Nebraska counties explored in an NCI population-based case control study of non-Hodgkin's lymphoma. From the photos, they set out to determine where different crops had been planted and which pesticides had likely been used proximate to study participants' residences. Explains Nuckols, "Since the images were from 1984, we needed to find a way to ground-truth since we were not able to go to the site and see what was there."

The researchers turned to Farm Service Agency records from the three counties, which contained information on the location and type of crops planted in 1984. The pesticides used on these crops were ascertained from a 1982 survey of pesticide use among farmers conducted by the University of Nebraska Institute of Agriculture and Natural Resources. By estimating chemical use for each crop type, the researchers used ArcView to identify crop fields with a high probability of agricultural pesticide use within a 500-meter radius around each residence.

"The 500-meter buffer zone was chosen as an intermediate distance between the extent of drift likely to occur from aerial and ground-based application methods," says Ward.

"In previous epidemiologic studies of agricultural pesticides, people who lived in town and didn't farm were typically classified as unexposed," Ward explains. "In fact their homes may have been close enough to crop fields to result in exposure from pesticide drift." Indeed, she and Nuckols found that 22 percent of homes overall and 15 percent of homes within towns were within 500 meters of pesticide use. Thus, these residences in towns could have had exposure to agricultural pesticides.

"Within ArcView," says Ward, "we had the capability to zoom in on specific residences, create visual displays of the local landscape, and conduct spatial queries to assess pesticide exposure beyond the farm." Adds Nuckols, "We used to concern ourselves primarily with farmers and their families when we considered these exposures. With the use of this technology, we've moved to a much more integrated view of exposures for the entire population."

Further research in this area will account for the likely extent of pesticide drift occurring during the pesticide application. Additionally, the researchers will validate whether there is a relationship between proximity to crop fields and levels of agricultural pesticides in homes. Ward and Nuckols are working with three research groups to make this determination. A group at the Iowa Geological Survey has created crop maps from satellite imagery for Iowa, and NCI researchers will correlate the agricultural pesticides predicted by the GIS-based approach with the occurrence of pesticides in house dust samples from more than 100 homes. Nuckols and Ward are also working with research groups at the California Department of Health Services and the University of California, Berkeley, to conduct a similar study in the Central Valley of California. As part of the latter research project, Nuckols has received research funding from NCI to determine how best to modify GIS methods for predicting residential pesticide exposure.

For more information, visit ehasl.cvmbs.colostate.edu on the Web or contact Dr. J.R. Nuckols, Environmental Health Advanced Systems Laboratory, Department of Environmental Health, Colorado State University, Fort Collins, Colorado 80523-1681. (Based on an article from Linkage, the newsletter of the Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, U.S. Department of Health & Human Services.)