Pasture degradation estimates through field data in the State of Goiás, Brazil

Abstract

Grasslands and pastures play a critical role in global food security, biodiversity, and ecosystem services, particularly in tropical regions where they serve as the primary base for livestock production. However, these areas are increasingly vulnerable to degradation due to poor management practices, as observed in the Cerrado biome. This study focuses on pastures in Goiás, a state located in central-western Brazil, which relies heavily on pastures for its agricultural economy. We present a comprehensive field-based assessment of pasture degradation in Goiás, aiming to estimate the area of degraded pasture according to three categories: non-degraded, in process of degradation, and severely degraded. A three-stage cluster sampling design was employed to evaluate 460 field points, incorporating visual assessments of key variables using the novel Pasture Assessment Score (PAS). The sample size was determined to achieve a maximum expected standard error of 4.47% for the estimated proportion of pasture area across the three degradation levels, calculated using the Monte Carlo method. The scores were analyzed using Multiple Correspondence Analysis (MCA) to reduce dimensionality and identify clusters corresponding to degradation levels, which were modeled using Decision Trees. The results revealed that 39.35% (SE = 3.92%) of the pasture area in Goiás was classified as non-degraded, 31.98% (SE = 2.66%) as being in process of degradation, and 28.67% (SE = 3.64%) as severely degraded. These findings underscore the need for targeted management strategies and enhanced monitoring frameworks integrating field-based assessments with remote sensing data to support sustainable pasture management.