Simulation of soybean response to drainage facilities and optimization of system design

A study has been made to determine the degree of drainage which maximizes the net return from growing soybeans in soils with an impervious layer close to the surface. To achieve this purpose, the drainage system performance and the crop yield response to each system alternative had to be determined;To simulate the performance of each drainage geometry, two main steps were followed. First, a laboratory investigation using a glass-beads-glycerol drainage model was carried out. The objective was to find and test an equation for predicting water table heights when drains are laid on an impervious layer. Second, a water balance approach to simulate daily water table depths under subsurface drainage facilities was developed. This water balance integrates the determinations of daily excess of soil water and water table depth associated with the drainage geometry being considered;From available experimental data reported in the literature, a soybean yield reduction pattern as a function of various depths and durations of the water table was established. A methodology for simulating crop yield reduction due to water table fluctuation associated with subsurface drainage facilities was also developed;A computer program was developed in Fortran IV G level language. This program integrates the water balance approach for simulating water table depths under subsurface drainage facilities with the model to predict soybean yield responses. Climatological records for a 28-year period for Pelotas, Rio Grande do Sul, South Brazil were used to simulate soybean yield responses for different degrees of drainage;An economic analysis was performed to establish the relationship between levels of drainage investments and benefits from different degrees of drainage. On the basis of this study, it was concluded that the results of the economic analysis do not favor the installation of subsurface drainage. The low hydraulic conductivity (20 cm/day) used in this study was probably the most important factor leading to this conclusion