ELECTRONIC NOSE USED FOR GLOBAL GAS ANALYSIS, WITH THE PLANT UNDER WATER STRESS.

Abstract

The development of non-invasive methods and affordable tools to investigate plant volatiles, applied to crop protection is considered of fundamental importance in agriculture. This work presents investigations and initial studies, with the so-called electronic nose (E-Nose), which is an electronic system that seeks to emulate the structure of a human nose to detected VOCs. The intensity of the gas emission by soybean under healthy plant and water stress conditions was evaluated. The soybean (Glycine Max) is a crop considered of great economic importance for Brazil, being the main culture of Brazilian agribusiness. For this experiment, a commercial "E-nose” was used and a gas chamber was built, which was internally instrumented with sensors of relative humidity (%), temperature (°C), CO2 sensor (ppm). A vase with dystrophic Red-Yellow Latosol was used, and isolated with aluminum foil. The measurement with the electronic nose was conducted in two intervals, i.e., with the healthy plant (feeding with DI water) and under water stress condition, to obtain the gas emissions of the soybean plant. The natural light Intensity was monitoring with variation in the range of 2.0K Lux to 46.0K Lux. The evaluation of the data collections, from the sensors, was observed a photorespiration pattern, with CO2 production during the day and night, as well as a temperature and relative humidity pattern during all the time of the experiment. Preliminary evaluation of the results showed sensitivity (in ppm) of the set of sensors for the volume being measured, as well as the potential for use of this technique to conduct water stress experiments. Future investigations using soybeans should be carried out in an environment of controlled conditions, and machine learning investigations, using data mining.