Grassland degradation estimates through field data in the state of Goiás, Brazil

Abstract

Grasslands 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, grasslands are increasingly vulnerable to degradation due to poor management practices such as the Cerrado biome in Brazil. This study focuses on grasslands in Goi´as, a state located in central-western Brazil, which relies heavily on pastures for its agricultural economy. We present a comprehensive field-based assessment of grassland degradation in Goi´as, aiming to estimate pasture degradation area according to three categories: nondegraded, 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 a novel Pasture Assessment Score (PAS) system developed specifically for this study. The scores were analyzed using Multiple Correspondence Analysis (MCA) to reduce dimensionality and identify clusters corresponding to degradation levels. Monte Carlo simulations were used to estimate confidence intervals for each degradation class, with standard errors ranging from 3.1% to 4.5%. The results revealed that 39.35% of the pasture area in Goi´as was classified as non-degraded, 31.98% as being in process of degradation, and 28.67% as severely degraded. This study highlights the need for targeted management interventions and improved monitoring frameworks to mitigate grassland degradation. Future works should focus on integrating field data with satellite-derived indices to improve mapping accuracy, and enable large-scale, high-frequency monitoring and detection of early-stage degradation for more effective land management strategies