Detecting drought stress in longan tree using thermal imaging

Thailand is the world's number-one producer of longan fruit. In general, longan production takes place during the dry season under irrigation. Recently, more attention has been given to water-efficient irrigation. Water stress detection by thermal imaging, which is a non-invasive and rapid assessment method, may be an interesting tool for improved irrigation planning. In this study, four potted longan trees were subjected to water stress. Stress responses in terms of stomatal resistance (r(s)) and leaf water potential (LWP) were monitored and compared with a non-stressed control. Based on thermal imaging, the crop water stress index (CWSI) was determined throughout the experiment for all trees and correlations with classical parameters were investigated. A field experiment was also carried out with 20 field-grown longan trees, either subjected to water stress treatment or serving as controls; trees were monitored for rs, LWP and CWSI. Under controlled conditions there was a high correlation between CWSI and both rs and LWP during the entire experimental period. In the field experiment it was found that CWSI was best correlated with rs when images were taken from the shaded side of the leaves. A threshold value of 0.7 for CWSI is proposed to distinguish between stressed and non-stressed longan trees.