Iowa farmers are faced with determining which management practice is the most economically and environmentally sustainable for their specific farm size, cash flow, and land characteristics. Both integrated crop management and site-specific farming require increasing levels of cost by the farmer, however, the expenditure may lead to increased profits and reduced environmental consequences for some fields.

Since it's impossible to replicate the same conditions on a field to test various practices, how can we know which practices will lead to maximum profits on a particular field?

The answer may lie with crop growth models-computer models-that can be used to replicate various conditions and management techniques.

Computer inputs include soil type, weather, management practices, and crop genetics. These are used to compute daily growth of crop components such as leaves, stems, roots, pods, and seeds. Models can respond to changes in environmental conditions and management practices.

This shows the error between predicted and measured yields across the field.

Crop growth models

In Iowa , models for corn and soybean growth were evaluated recently. Both models have given good estimates of crop growth and yield in research plots located around the state during the 1995 and 1996 growing seasons. These models provide an excellent tool to evaluate the sensitivity of crop growth and yield to changes in management practices, and to study how climatic and management practices impact yields.

In a project funded by the Iowa Corn Promotion Board and the Leopold Center for Sustainable Agriculture at Iowa State University, an interdisciplinary team of ISU researchers use these crop growth models to investigate causes of yield variability.

The corn model was used in a case study of yield variability in a field near Ames. The 40-acre field is divided into 224 grids, where yields have been measured by researchers at the National Soil Tilth Laboratory since 1988. The corn model was calibrated to measured yields for 55 grids for the 1995 growing season, using estimates of soil moisture holding capacity and rooting depth. These properties were then interpolated to the remaining 169 grids, and yield estimates were computed.

The difference between predicted and measured yields for the whole field was less than 1 bushel per acre. Overall, 81 percent of the grids gave errors within 10 percent of measured yields, and 96 percent of the grids had errors within 20 percent of measured yields.

When looking at the causes of yield variability, it appeared that in some grids, soil moisture stress, resulting from shallow rooting depths was a cause. In these grids, an abundance of spring rainfall caused relatively high water tables, which limited predicted rooting depth from emergence to tasseling.

Later during the season, water tables dropped, and limitations in rainfall caused mild to severe water stress, resulting in predicted yield reductions and yield variability across the field. The model also indicated some nitrogen stress in grids with shallow rooting depths and higher leaching rates, which often was related to water stress.

Further study

These models also will compute the environmental impact of nitrogen application in terms of nitrate leaching. This method will be tested on several Iowa farms, and will be used to determine productivity under different management strategies.

Methods are also being developed to estimate the economic return from each management strategy, based on crop response predicted by the models.

On-farm demonstrations are being conducted in 10 fields in Jones and Lynn counties to directly compare the economic return from each management practice.