Simulating Corn Yields for Two Iowa Soils

Abstract

Corn yields in Iowa often could be improved with better water table management practices, but many soils in this region do not have adequate information about their yield potential. This study was conducted to evaluate the applicability of DRAINMOD in simulating crop yields for artificially drained soils of central and northeastern Iowa. Ten years of field data on crop yields from two drainage experimental sites in Iowa were used to compare relative crop yields with yields predicted by DRAINMOD. Data on saturated hydraulic conductivity, soil-moisture tensions, soil infiltration, and other physical properties were collected either at the experimental sites or in the laboratory on disturbed soil samples. Data on upward flux and infiltration constants for the Green and Ampt equation were calculated by using the infiltration data and soil-moisture retention characteristics of the surface layer. Relative crop yields predicted by DRAINMOD were reasonably close to the observed relative yields for both Iowa sites. Standard error of the estimates of relative yields was 17.41% and the average deviation was 12.99%. The coefficient of determination, r2, between the predicted and measured yields was 0.54. Overall performance of the model suggests that DRAINMOD can be used successfully for predicting yields for different locations if data on site characteristics, soil-water properties, and plant growth functions are available.