Yield prediction in banana (Musa sp.) using STELLA model

Abstract

To overcome the challenges encountered in banana cultivation, such as the high cost of production due to high water consumption by the banana plant, efficient management practices are being adopted. The use of agricultural forecasting techniques is an alternative that has been gaining attention in rural areas. One way to manage and improve agricultural productivity is the use of technologies that allow the monitoring of production. The implementation of computational tools as software to aid processes, such as irrigation management, is gradually taking up space in the agricultural sector. In this light, herein, the present study aimed to develop a model using STELLA 8.0 software to estimate the growth and productivity of irrigated banana ( Musa sp.). For this, the physiological processes and water demand were calculated using reference evapotranspiration (ET₀) and culture evapotranspiration (ET_c) in the first banana cycle for the climatic conditions of the Jaíba Project (Jaíba, Minas Gerais State, Brazil). The data of the climatic conditions were obtained from the National Institute of Meteorology. It was verified that the average monthly ET₀ was 5.78 mm day^-1. In addition, the water requirement of the plant corresponded to a blade equivalent to 65% of ET₀. The verified productivity was 8.93 t ha^-1, which is considered adequate for the simulated conditions. The model responded efficiently to the proposed application and was characterized as a prognostic tool of reality through simplified representation.