Application of Vegetation Indices to Determine the Reproductive Development of Açaí in the Eastern Amazon

The açaí production chain (Euterpe oleraceae Mart.) has been expanding and bringing with it several innovations and initiatives for the cultivation. Remote sensing technologies allow obtaining multispectral images of the açaí tree, which can be applied in the reproductive phenological determination of the crop, being an effective tool for agricultural management. The present study aimed to determine the relationship between the reproductive development of the açaí tree and the vegetation indices obtained by RPAS (remotely piloted aircraft system) images. The work was carried out in a commercial plantation of açaí, with 7 years old, located in Fazenda Ornela LTDA, municipality of Capitão Poço, state of Pará, Brazil. The experiment was carried out in random blocks, with 8 plots containing 6 clumps each. Each clump consists of 2 stems, totaling 12 stems per plot. All plots received the same cultural treatments. The variables were analyzed: number of spathes, number of inflorescence, number of green bunch, number of black bunch and number of leaves. To obtain orthophotos and assess the NDVI (normalized difference vegetation index) and SAVI (soil-adjusted vegetation index) of the açaí grove, a DJI Mavic 2 Pro RPA equipped with a Mapir Survey 3 multispectral camera was used. The results show that the NDVI and SAVI are correlated with the variable number of black bunch. An estimation model was built to determine the number of black bunches based on NDVI and SAVI. The model explains 47.05% of the variation for NDVI and 30.5% for SAVI, indicating it can predict the reproductive phenological variables and approximate production values of the açaí tree, aiding producers in agricultural management.