Laboratory Evaluation of a Low-Cost Metal Oxide Sensor Platform for Monitoring Herbivore Induced VOCs in Maize

Early detection of plant stress and pest infestation is a key enabler of sustainable and precise crop management. Volatile organic compounds (VOCs) emitted by plants provide a valuable, non-invasive means of monitoring physiological changes, yet their implementation in real-time sensing remains limited by the high cost, complexity, and laboratory confinement of conventional analytical methods such as GC-MS (Gas Cromatography - Mass Spectrometry). This thesis investigates the potential of a low-cost MOS (Metal Oxide Semiconductor) gas sensor platform for detecting plant-emitted VOCs under controlled laboratory conditions. The system was tested across two main experiments designed to characterize the sensor response to biological volatile emissions. A preliminary validation was performed using sheep wool samples to assess system performance and stability under known conditions. The main experiment involved maize (Zea mays L.) plants subjected to herbivory by larvae of Spodoptera littoralis (Boisduval), a polyphagous lepidopteran pest widely used in studies of herbivore-induced plant volatiles. The experimental setup allowed simultaneous monitoring of eight plants for approximately twenty-four hours, including a two-hour period of dynamic headspace collection for GC-MS analysis. The results demonstrate that the tested MOS sensors can qualitatively detect the presence of plant-emitted VOCs under herbivory. Although the sensor signals did not allow for discrimination of individual compounds, clear differences in response magnitude were observed between plants exhibiting high and low VOCs emissions, as confirmed by GC-MS. This indicates that commercial MOS sensors are sensitive enough to capture large variations in volatile release associated with biotic stress, even without chemical selectivity. The responses were highly dependent on the degree of feeding activity and overall VOCs concentration, highlighting the need for improved control of biological variability in future experiments. While GC-MS remains indispensable for compound identification, this study demonstrates that low-cost MOS sensors can serve as portable, complementary tools for real-time plant monitoring. Their affordability and simplicity make them promising candidates for integration into distributed sensing networks for greenhouse or field-scale applications. The work establishes an experimental and methodological foundation for future development of low-cost electronic systems aimed at early detection of plant stress through continuous monitoring of volatile emissions, contributing to the broader vision of precision and data-driven agriculture.