Multifunctionality of Brazilian pasturelands: A large-scale spatial assessment based on DSSAT simulations and digital soil mapping

• Grant number: 25/09804-2
• Support Opportunities: Scholarships abroad - Research Internship - Doctorate
• Start date: September 15, 2025
• End date: September 14, 2026
• Field of knowledge: Agronomical Sciences - Agronomy - Soil Science
• Principal Investigator: Maria Victoria Ramos Ballester
• Grantee: Andres Mauricio Rico Gomez
• Supervisor: Gerrit Hoogenboom
• Host Institution: Centro de Energia Nuclear na Agricultura (CENA). Universidade de São Paulo (USP). Piracicaba , SP, Brazil
• Institution abroad: University of Florida, Gainesville (UF), United States
• Associated to the scholarship: 24/08446-2 - Pastures sustainable development in Brazil: spatially explicit soil multifunctionality modelling for a climate-changing world, BP.DR

Abstract

Soil is a dynamic and multifunctional natural system, capable of providing multiple ecosystem services strongly modulated by environmental conditions. Due to increasing climate change, understanding how these functions interact is essential for promoting the resilience of agroecosystems. Given that approximately 107 Mha of pasturelands in Brazil exhibit some degree of degradation, they present opportunities for restoring soil multifunctionality (SM). However, there is still a lack of spatial and integrated studies that analyze the response of soils to climate change in the joint provision of their functions. This gap limits the planning of public policies and adaptive management strategies. In this context, this project aims to assess the impact of climate change on soil multifunctionality in Brazilian pastureland, considering their intrinsic potential to produce biomass, water regulation, and climate services. Therefore, we will employ a robust and innovative methodology in three stages: i) we will compile and harmonize legacy soil, climate and remote sensing datasets; ii) we will calibrate the DSSAT model to simulate pasture growth under baseline climate conditions (1980-2013) and future climate change scenarios (2015-2100); and iii) we will use output variables (biomass, soil-profile water content and total carbon balance) as a SM indicators to train a Random Forest algorithm that will produce continuous predictive maps with 30 m resolution. The results of this project will provide useful maps and indicators to guide sustainable management actions, pasture recovery, and climate change mitigation policies in the agricultural sector, aligning science and territorial management on a national scale.