Location: Crop Production Systems ResearchTitle: Optical and Thermal Methods for Detection of Plant Damage Caused by Tomato Spotted Wilt Virus) Author
Submitted to: University of Georgia Research Report
Publication Type: Experiment Station
Publication Acceptance Date: 1/4/2007
Publication Date: 1/23/2007
Citation: Maw, B.W., Stephenson, M.G., Ritchie, G.L., Mullis, S.W., Thomson, S.J. 2007. Optical and Thermal Methods for Detection of Plant Damage Caused by Tomato Spotted Wilt Virus. University of Georgia Research Report 1-2007. J. Michael Moore, editor. Interpretive Summary: A simple digital camera using an infrared blocking filter was evaluated against other detection methods for detecting the onset of Tomato Spotted Wilt Virus (TSWV) in tobacco plants. Other methods included spectroradiometer (used to detect reflected light in narrow wavebands) and an infrared thermometer. Photographs of healthy and diseased plants included a gray card reference panel to compensate images for changes in sunlight intensity. Several studies to evaluate sensing systems for detection of TSWV were conducted in the greenhouse and field. Some plants were examined for TSWV and labeled using the Double Antibody Sandwich-Enzyme Linked Immunosorbent Assay (DAS-ELISA). The ELISA procedure was especially useful to evaluate the sensitivity of detection methods for determining the onset of disease, as disease may be present but not yet visible. Higher near-infrared reflectance was indicated consistently for healthy plants vs. the diseased plants from processed photographs in the greenhouse, but results were not significantly different for all trials. Statistically significant differences were observed between diseased and healthy plants from both spectroradiometer data and digital photographs in the near-infrared band for field-grown plants. The spectroradiometer was able to detect the onset of TSWV even though visible symptoms were not yet visible on the plant. These plants had previously been found to be infected using ELISA. Stressed plants showed higher leaf temperatures than healthy plants, but not enough samples were taken to determine statistical significance. Procedures for detecting the onset of plant disease using a digital camera and grey card show some promise, but sensitivity needs to be increased for reliable detection of the onset of disease. Confounding effects such as nematode damage and nutrient stress were noted in some experiments, and these effects would need to be accounted for.
Technical Abstract: Reflectance in the near-infrared and thermal wavebands (about 1 to 10 micrometers) has been used to distinguish between diseased and healthy plant leaves. Photographs of leaves from both healthy plants and plants infected with the Tomato Spotted Wilt Virus have shown indications of reflectance value variations according to disease infection. Photographs have been taken of plants in the field during summer as well as plants grown in pots in the greenhouse during autumn and winter. Diseased plants in the field were recognized by visual observations and by ELISA analysis. Diseased plants in the greenhouse were identified by those known to have been inoculated. During the summer additional observations were made with a spectroradiometer and infrared thermometer. Differences have been observed between the reflectance of near infrared radiation from diseased plants and that from healthy plants, the reflectance for diseased plants being lower than that from healthy plants. Strong sunlight is required for the technology to be useful.