Design and Development of a Gas Sensor Array to Detect Salinity Stress in Khasi Mandarin Orange Plants

Soil salinity is a major environmental threat that limits the productivity of plants, and can be a cause of several other secondary stresses such as oxidative stress, osmotic stress, and ionic stress. During stressed conditions, the responses of plants like changes in photosynthetic pigments (chlorophyll and carotenoid concentrations), leaf water accumulation, nutrient imbalance, ion homeostasis, and induction of excessive accumulation of reactive oxygen species occur. The alterations in the emission of volatile organic compounds (VOCs) from the leaves of the plants can also occur due to the induction of salinity stress in plants. An electronic nose (E-nose) can be used to effectively monitor the composition of these leaf-emitted VOCs for the diagnosis of salinity stress induced in the plants. In this paper, the authors describe the design and development of an E-nose system augmented with the necessary pattern recognition algorithm to detect the salinity stress in the Khasi Mandarin Orange plants. Enabling temperature modulation in the system improves the ability of MOS-based gas sensors to selectively detect leaf-emitted VOCs. To train the dataset, machine learning algorithms such as Support Vector Machine (SVM) and Random Forest (RF) classifiers are used. The results show that the prototype that is developed has achieved a maximum accuracy of 98.3detecting salinity stress. The chlorophyll content and chlorophyll fluorescence measurements are also done to verify the stress induction in the target plants.