Role of water stress in yield variability

Yield variability results from complex interactions between the environment, genetics, management, and biotic stress that occurs across a field. Of these factors, water stress is one of the major causes of yield variability in fields. Water stress occurs when roots cannot supply enough water to satisfy evaporative demand of water transpiring from leaves. Root water uptake is a function of soil water availability, root depth and density, and location of roots relative to water in the soil. Plant genetics also can influence root growth and development, and subsequently water uptake, by influencing location of roots relative to water in the soil. Daily evaporative demand driving transpiration and potential water stress is a function of uncontrollable environmental factors, including temperature, relative humidity, and wind speed. Water stress reduces photosynthesis, resulting in reduced crop growth and yield. Minor water stress often occurs for short periods during the hottest part of the day, late in the season. Water stress can vary across a field, depending upon soil type, moisture-holding capacity, and drainage characteristics. The integration of variable water stress each day over the season can result in reduced photosynthesis and variable yield loss across a field. Because of the dynamic nature of water stress, it is very difficult to measure and correlate to final yield.

Background on the project

In work funded by the Iowa Soybean Promotion Board, Iowa Corn Promotion Board, Leopold Center for Sustainable Agriculture, and the USDA Soil Tilth Laboratory, Iowa State University researchers are using crop models to study the role of water stress in creating yield variability. Crop models are computer programs that can integrate the effects of environment, management, genetics, and biotic stress each day during the season to compute how plants respond to daily stress and produce final yield.

A case study

Corn and soybean crop models were used to determine the impact of water stress on creating yield variability in the Baker field near Ames. The field, which has been in a corn-soybean rotation since 1989, is divided into 224 grids with measured yield since 1989. The models were calibrated to test the hypothesis that water stress is a limiting factor for both corn and soybean yields across the field. Comparing predicted and measured yields over several years in a grid can show the degree to which water stress explains yield variability. The corn model was run for the years 1989, 1991, and 1995 and indicated that approximately 59 percent of the yield variability in these three years resulted from water stress. The soybean model, run for years 1992, 1994, and 1996, indicates that water stress explained approximately 69 percent of the yield variability. The model indicates that water stress in low-lying grids probably occurred due to shallow root depth resulting from poor drainage. This could likely be reduced with improved drainage.

What's next?

The next step in this research is to evaluate optimum management practices--such as nitrogen rates and variety selection--to improve profitability in each grid. By using the models, we will be able to compare the economics of different types of field management. For instance, we can compare single-rate nitrogen applications to variable-rate. In addition to this, a multistate and multidisciplinary project jointly funded by the Iowa Soybean Promotion Board and the Illinois Soybean Checkoff Board will provide new relationships in the soybean model, allowing us to consider many variable-rate management alternatives in the future. These new relationships also will allow us to consider interactions of herbicide injury, nematodes, and diseases for variety selection across fields.

Figure 1. Error between predicted and measured soybean yields for 1994.

Error in Soybean Yield Prediction

Figure 2. Error between predicted and measured corn yields for 1989.

Error in Corn Yield Prediction

This article originally appeared on pages 3-4 of the IC-480 (4c Precision Ag Edition) -- April 9, 1998 issue.

Updated 04/08/1998 - 1:00pm