Uso de ráfia pode influenciar os parâmetros agronômicos, a dinâmica da água no solo e a resposta espectral de cafeeiros?

Abstract

Coffee is of great importance to agribusiness, with Brazil being the world's largest producer of this commodity and having a strong presence in the specialty coffee market. Several factors affect coffee productivity, and the use of synthetic or natural ground covers has shown promising results. Raffia, a polypropylene plastic cover, is intended to allow better aeration and maintenance of soil moisture, act as a physical barrier mitigating erosion processes and leaching of nutrients in the soil, and is an excellent option for weed management. However, there are still no studies proving the effectiveness of using this material as ground cover and its effect on phytotechnical aspects in coffee crops. In this sense, the objective of this study was to verify whether the use of raffia can influence agronomic parameters, soil water dynamics, and the spectral response of coffee plants. The experiment was carried out in the municipality of Araguari, Minas Gerais, in a randomized block design with five blocks and four treatments: two strips of raffia placed together on both sides of the coffee plants; two strips of raffia placed on both sides of the coffee rows and 0.5 m apart; two strips of raffia placed on both sides of the coffee rows and 0.7 m apart; and the treatment without raffia. To meet the crop's water demand, irrigation sheets were applied to supplement natural rainfall through a localized drip irrigation system. Soil water dynamics were monitored through tensiometric measurements taken daily. The agronomic parameters were assessed bimonthly after the installation of the raffia, and the productivity data were analyzed over two harvests, in the years 2022 and 2023. Concomitantly with the agronomic assessments, radiometric data were obtained using the CBERS 4A sensor. Supervised classification of the data was performed using the Random Forest algorithm, in order to evaluate data groups composed of spectral bands, vegetation indices, and the combination of bands and indices. Based on the results, it was found that the use of raffia as ground cover does not provide increases in productivity and growth parameters of the coffee plant. However, the combination of multispectral images and vegetation indices works as a complement to field analyses, since they are capable of presenting greater accuracy (overall accuracy = 90%) compared to the isolated analysis of agronomic parameters.