Sensoriamento remoto para identificar soja infectada por fungos causadores da ferrugem asiática

Abstract

Since the 1970s, remote sensing (SR) instruments have been applied in the most diverse areas of knowledge. Areas such as the environment, monitoring of agricultural production areas, territorial expansion and occupation / use of soil and water have repeatedly found subsidies for their work in the SR instruments. This technique is based on obtaining information through the interaction of electromagnetic energy (REM) and target objects without physical contact. Currently, there has been an expressive number of scientific studies aimed at the search for relationships between specific characteristics of vegetation and spectral responses along the electromagnetic spectrum. In view of this, the present thesis had the general objective of identifying spectral features associated with the pre-symptomatic phase of soybean crops infected by Phakopsora pachyrhizi. For that, three specific objectives were stipulated: 1) to verify the climatic conditions associated with the occurrence of Asian rust; 2) to identify the biophysical characteristics of the epidemic system in the pre-symptomatic and symptomatic phases; 3) to characterize the spectral behavior of soybeans by spectroradiometry in order to detect changes in reflectance resulting from the occurrence of the fungus Phakopsora pachyrhizi. The experiments were conducted on the premises of the Phytus Institute - Itaara-RS, a partner institution of LabGeotec-UFSM. The activities were carried out concurrently in the countryside areas and in two greenhouses. Initially, two soybean areas (total of 300m²) were made available, divided into two plots measuring 150 m² each. The plots and the two greenhouses were named P01 and P03, P01 being infected without any chemical treatment and P03-Free previously treated area. Hyperspectral data were collected in the field using a Field Spec 3 spectroradiometer (between 400 to 900 nm) on the soybean canopy and in greenhouses on the leaves of the soybean plants conditioned there with the aid of the Probe Leaf apparatus. The biophysical variables (chlorophyll “b” - Chl ”b”, Foliar area index (IAF), Biomass and Number of spores of the fungus) were collected in the field and in greenhouses and processed in the laboratory using destructive methods consolidated by literature. The general analysis of the climatic variables allowed us to conclude that the environmental conditions were favorable during the soybean cultivation period in the study areas, the interval that preceded the infection during the sampling registered average temperatures between 17 and 28ºC, with better growth ranges between 21 and 26 ºC, RH above 80% associated with moderate rainfall (around 20mm) and leaf wetting time between 5 to 10 hours. The results recorded by the biophysical variables (Chl ”b”, Biomass, IAF) and the increase in spores in the atmosphere point to predisposing conditions that enabled the infection process by the action of the fungus Phakopsora pachyrhizi and allowed the disease to occur. The use of hyperspectral remote sensing made it possible to characterize the spectral behavior of soy by spectroradiometry as well as to detect changes in reflectance resulting from the occurrence of the fungus Phakopsora pachyrhizi in the asymptomatic period.