Acceso usuarios
The amount of soil water at the beginning of the growing season has a large impact on crop yields in rainfed agriculture, especially in semi-arid regions and in years with below-average rainfall in more humid climates. Robust algorithms are needed to estimate soil water storage before planting to aid crop management decisions. The main objectives of this paper are to investigate soil water recharge during the non-growing season (October 20 to May 1) in a semi-arid, temperate ecosystem in south-central Nebraska (USA) and to evaluate empirical models to estimate soil water content at the beginning of the summer-crop growing season. A database of soil water content measurements collected over 5 years at nine locations in south-central Nebraska was used to estimate available water-holding limits in the soil profile and to determine the change in available soil water during the non-growing season. Regression analysis was performed to analyze the relationship among soil water recharge, residual soil water (i.e., soil water content at the end of the previous growing season), total precipitation, and available water-holding capacity (AWHC) in the root zone to 1.5 m. Precipitation storage efficiency (PSE) was calculated as the quotient of soil water recharge and total non-growing season precipitation. Predictive models to estimate soil water content at the beginning of summer-crop growing season were derived from these analyses. A large portion of the variation in soil water recharge was explained by residual soil water and precipitation. PSE averaged 28% across site-years; low PSE values were associated with high residual soil water and/or low AWHC. Two predictive models (linear and linear-plateau) that used residual soil water, total precipitation, and AWHC as independent variables explained 75�80% of the variation in the measured soil water content at the beginning of the summer-crop growing season. These empirical models represent a new tool to estimate soil water content by planting date of summer crops. Site-management conditions such as residue amount and its architecture, tillage system, soil texture, and terrain slope are not currently accounted for in these models and would likely improve predictive capacity.
