Mapping Soil and Pasture Attributes for Buffalo Management through Remote Sensing and Geostatistics in Amazon Biome.

Abstract

Simple SummaryBuffalo breeding in the Amazon biome can contribute significantly to local community development and, thus, is considered an essential income source. However, in Amazon regions, the inadequate breeding of these animals can lead to considerable negative consequences for the environment. Therefore, it is crucial to develop methodologies to improve animal management and grass yield. One of these methodologies is related to Precision Agriculture (PA), adapted for pasture and animal monitoring. Along these lines, we seek to utilize geostatistical techniques and remote sensing applications to better understand Buffalo grazing under a rotating system. In particular, we analyze forage Dry and Green Matter, as well as pH in pasture soils, demonstrating the obstacles against and advantages of the implementation of precise techniques for decision making and increasing grass productivity. We describe ways in which geostatistical soil pH mapping can be conducted, as well as the premises necessary to include remote sensing data in the analysis of pasture variables. Implementing these results in buffalo management systems can contribute to greater productivity and increasingly sustainable livestock.The mapping of pastures can serve to increase productivity and reduce deforestation, especially in Amazon Biome regions. Therefore, in this study, we aimed to explore precision agriculture technologies for assessing the spatial variations of soil pH and biomass indicators (i.e., Dry Matter, DM; and Green Matter, GM). An experiment was conducted in an area cultivated with Panicum maximum (Jacq.) cv. Mombaça in a rotational grazing system for dairy buffaloes in the eastern Amazon. Biomass and soil samples were collected in a 10 m × 10 m grid, with a total of 196 georeferenced points. The data were analyzed by semivariogram and then mapped by Kriging interpolation. In addition, a variability analysis was performed, applying both the Normalized Difference Vegetation Index (NDVI) and Normalized Difference Water Index (NDWI) derived from satellite remote sensing data. The Kriging mapping between DM and pH at 0.30 m depth demonstrated the best correlation. The vegetative index mapping showed that the NDVI presented a better performance in pastures with DM production above 5.42 ton/ha−1. In contrast, DM and GM showed low correlations with the NDWI. The possibility of applying a variable rate within the paddocks was evidenced through geostatistical mapping of soil pH. With this study, we contribute to understanding the necessary premises for utilizing remote sensing data for pasture variable analysis.