Evaluating EPIC-based Harvest Index Approach for Yield Prediction and Response to Soil Water.
Gregory McMaster, USDA-ARS-NPA-SPNRU, USDA-ARS Great Plains Sys Res Unit, 2150 Centre Ave Bldg. D Suite 200, Fort Collins, CO 80526, James C. Ascough II, USDA-ARS-NPA ASRU, 2150 Centre Ave., Bldg. D, Suite 200, Fort Collins, CO 80526, Allan Andales, USDA-ARS, USDA-ARS-Agricultural Systems Research, 2150 Centre Ave Building D Suite 200, Fort Collins, CO 80526, and Deborah A. Edmunds, USDA-ARS, Agricultural Systems Res Unit, 2150 Centre Ave., Bldg. D, Suite 200, Fort Collins, CO 80526.
Variations of the EPIC-based plant growth model have been used in many national modeling efforts such as WEPP, WEPPS, SWAT, and GPFARM. Many different approaches are used in simulation models to predict final crop harvestable yield. The EPIC-based plant growth model uses the harvest index (HI, defined here as the ratio of harvestable yield to total aboveground biomass) approach to calculate yield. Briefly the approach is as follows. The expected final HI is an input parameter for each crop. For optimal conditions, the adjusted (HIA) increases non-linearly from zero at planting to final HI at maturity. A heat unit factor changes the daily estimate of HI so that most of the HI is determined in the second half of the growing season. Only water stress is used to reduce HIA via two mechanisms. The first is a parameter for the sensitivity of a crop HI to water deficits and the second mechanism is the direct affect of the water stress factor (0-1). The most limiting of several stress factors including water, nutrient, and temperature (0-1) is used to reduce biomass accumulation and root growth. Biomass is then partitioned daily among different plant components partly based on the HI. This talk explores the physiological basis of the HI approach and evaluates the approach over a broad range of environmental conditions in the Central Great Plains for several row crops (e.g., wheat, corn, sorghum, sunflower, and proso millet). The long-term objective is to identify areas for modification to improve yield predictions in diverse production environments. The framework used for this evaluation was the EPIC-based plant growth model incorporated into the GPFARM decision support system. Results of the evaluation will be presented.