Everton Alves Rodrigues Pinheiro, Center for Nuclear Energy in Agriculture, University of Sao Paulo, Piracicaba, São Paulo, BRAZIL and Quirijn de Jong van Lier, Center for Nuclear Energy in Agriculture, University of Sao Paulo, Piracicaba (SP), BRAZIL
Soil hydraulic properties together with hydraulic gradients play an important role in hydrological processes that control the biomass productivity of natural and agricultural ecosystems. In this study, we obtained soil hydraulic properties (SHP) of representative soils from São Paulo State, Southeast Brazil, one of the most important agricultural regions of the Brazilian territory. The SHP were parameterized by inverse modeling of evaporation experiments. These experiments were performed from almost saturated to very dry conditions, with soil-water content monitored by 137Cs gamma-ray attenuation. Upward soil-water flux and water content over time were imported into the HYDRUS-1D model. Using the inverse solution option, SHP were estimated according to the Van Genuchten-Mualem soil hydraulic model (1980). An earlier developed limiting matric flux potential approach (Pinheiro et al., 2017) which takes into account soil, plant and climate characteristics (SHP, root density and atmospheric demand, respectively) was applied to find the limiting pressure head (hlim) for plant transpiration in layered soils. The amount of days required for a certain soil to reach the hlim depends on its inherent characteristic (soil hydraulic properties) and on scenarios in which it is exposed, e.g., atmospheric water demand and crop characteristics. To simulate this, we used the agro-hydrological SWAP model (Soil-Water-Atmosphere-Plant) to predict the time from initial condition to hlim for the top and subsurface layers. The SWAP model was parameterized in terms of soil, crop (maize) and weather. In order to analyze the soil hydraulic strength in keeping water movement to roots at potential rates under different scenarios of water supply, we simulated real and modified rainfall distribution during the crop cycle. Simulations with the calculated hlimwere compared to simulations using the empirical limiting pressure head of the Feddes reduction function.