D3321: Spectral characterization of cotton plants infested by the spider mite, Tetranychus urticae

Among the arthropods that attack cotton crops, phytophagous mites were considered secondary pests. However, with the increase in the use of phytosanitary products to control primary pests, there was a drastic reduction in natural enemies, which, associated with the occurrence of dry periods in the cotton-growing regions, made phytophagous mites one of the most frequent pests in cotton. Monitoring this pest is a laborious and costly practice. In this sense, new tools have been tested to automate the sampling process. Thus, the objective of this work was to evaluate the potential use of hyperspectral sensing to distinguish between healthy cotton plants and those infested with spider mite, under different periods of infestation. The test was carried out under controlled conditions, with artificial infestation of the mites in cotton plants grown in vases and surrounded by a cage made of “voil” type fabric. Subsequently, after 3, 6, 9 and 12 days, a leaf from each pot was collected to obtain reflectance values and the number of mites found. When analyzing the spectral behavior of all plants in the different periods of infestation, differences are observed in the curves of infested and non-infested plants. To confirm these observations, the average tests performed on the 28 selected bands on these data collection dates indicated a difference between treatments. Thus, it is concluded that it is possible to differentiate between infested and non-infested plants, and the attack symptoms become more evident over the days evaluated.