Productive and environmental aspects of integrated agricultural production systems in the Brazilian Atlantic Forest biome

Abstract

Integrated and consortium agricultural production systems can significantly contribute to the mitigation of greenhouse gases (GHG). However, it is still not clear how these systems can affect the quality of carbon (C) in the soil, nor how the dynamics of GHG emissions work depending on the emission rates (drivers). In the present study, we will have two distinct experiments, but which will originate from the following treatments: crop-livestock integration system (ILP), crop-livestock-forest/silvopastoral integration (SSP), marandu grass consortium with forage peanuts (CON) , monoculture of marandu grass (MON) and native forest (MN) belonging to the Brazilian Atlantic Forest region. The experiments will be conducted in the Forage and Pasture sector of the Universidade Estadual Paulista "Júlio de Mesquita Filho", Jaboticabal Campus - SP, 21º14'05'' south latitude, longitude 48º17'09'' W, at an altitude of 615 m, during the years 2024, 2025 and 2026. The soil in the experimental area is classified as a typical Dystrophic Red Oxisol. In experiment 1, assessments will include soil C and nitrogen (N) stocks, C quality indices from fractionation, and the impact of the management of each area on macrofauna and microbial C. Possible interactions between soil physics and C immobilization, and the effects of nutrient cycling on chemical attributes will be evaluated. After harvesting the crops and developing the grass, the forage mass and accumulation, morphological components, leaf area index and chlorophyll content will be determined. In experiment 2, we will evaluate GHG emissions in the soils of the production systems of experiment 1 under laboratory conditions, raising possible interactions between soil variables, including microbiological ones. For both experiments, the design will be completely randomized, in which the years of evaluation and repetitions will be random factors. The assumptions of normality and homoscedasticity will be tested, and in both experiments, multivariate statistics will be used to identify the main explanatory variables. (AU)