PARAMETERIZING SHOOTGRO 4.0 TO SIMULATE WINTER WHEAT PHENOLOGY AND YIELD IN THE CZECH REPUBLIC

Authors: ZALUD, ZDENEK - MENDEL UNIVERSITY CZECH; McMaster, Gregory; Wilhelm, Wallace

Submitted to: European Society of Agronomy
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/18/2002
Publication Date: 8/1/2003
Citation: Zalud, Z., Mcmaster, G.S., Wilhelm, W.W. 2003. Parameterizing shootgro 4.0 to simulate winter wheat phenology and yield in the czech republic. European Society of Agronomy. Vol. 19, Issue 4, pg. 495-507

Interpretive Summary: When crop models are developed they normally emphasize the cultivars and conditions for the region in which they were developed. This potentially poses problems in using these models to optimize management practices in new environments because the models may not adequately simulate the new cultivars and conditions. Further compounding the problem is that algorithms describing developmental processes in many models are insufficient to identify optimum management practices, especially those practices which are applied based on crop growth stage. Our objective was to assess whether the developmentally-based simulation model SHOOTGRO 4.0 could be easily parameterized and adequately to simulate winter wheat phenology and yield in the Czech Republic and then be used as a tool for optimizing various management practices. Parameterizing SHOOTGRO for simulating the three crop production regions of the Czech Republic was limited to obtaining general dates of jointing, anthesis, and maturity and maximum number of spikelets and florets per spikelet on the main stem from local agronomists. Field trials in 1998-99 were used to validate anthesis, maturity, and grain yield predictions. Sensitivity analysis was conducted to determine model sensitivity to selected parameterization variables with respect to phenology and yield predictions under selected management practices and environmental conditions. All validation measures suggested that predicting phenology (r2 > 0.70) and yield (r2 = 0.71) was good, with no model bias for phenology and a tendency to over-predict yield. This study suggests that the developmentally-based model SHOOTGRO 4.0 can be parameterized easily and adequately to simulate developmental stages and grain yield in the different crop production regions of the Czech Republic.

Technical Abstract: Crop models normally emphasize the cultivars and conditions for the region in which they were developed. In addition, most models are energy- or carbon-driven models rather than developmentally based. This potentially poses problems in using these models to optimize management practices in new environments because the model may not adequately simulate the new cultivars and conditions. Knowledge of, and algorithms describing, developmental processes in energy- or carbon-based models may not be sufficient to identify optimum management practices, especially those which are applied based on crop growth stage. Our objective was to assess whether the developmentally-based simulation model SHOOTGRO 4.0 could be parameterized easily and adequately to simulate winter wheat phenology and yield in the Czech Republic and then be used as a tool for optimizing various management practices. Parameterizing SHOOTGRO for simulating the three crop production regions of the Czech Republic was limited to obtaining general dates of jointing, anthesis, and maturity and maximum number of spikelets and florets per spikelet on the main stem from local agronomists. Field trials in 1998-99 were used to validate anthesis, maturity, and grain yield predictions. Sensitivity analysis was conducted to determine model sensitivity to selected parameterization variables with respect to phenology and yield predictions under selected management practices and environmental conditions. All validation measures suggested that predicting phenology (r2 > 0.70) and yield (r2 = 0.71) was good, with no model bias for phenology and a tendency to over-predict yield. This study suggests that the developmentally-based model SHOOTGRO 4.0 can be parameterized easily and adequately to simulate developmental stages and