Submitted to: Agronomy Journal
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/31/2007
Publication Date: 5/1/2008
Citation: Jiang, P., Kitchen, N.R., Anderson, S.H., Sudduth, K.A., Sadler, E.J. 2008. Estimating plant-available water using the simple inverse yield model for claypan landscapes. Agronomy Journal. 100:830-836. Interpretive Summary: Understanding how well a soil at a specified location in a field allows for movement of water into the soil (infiltration) and retention of water for plant growth (plant available water capacity or PAW) is basic to explaining crop growth and yield. Both of these soil properties are used in a general way to describe soil differences at large scales, such as contrasting soils among or between regional or political boundaries, but are seldom used at small-scales, such as within a single farm field. The main reason the approach is not used at the smaller scale is due to the expense of measurements. In this study, we proposed that PAW could be estimated using a crop modeling program and on-the-go yield map measurements. This approach previously showed promise when developed and tested on well-drained fields in Wisconsin, but it has not been tested elsewhere. Our interest was to evaluate this method of estimating PAW on poorly-drained claypan soils, found in large portions of northern Missouri and southern Illinois. We found that for claypan soils, yield map information was much less reliable for estimating PAW than with well-drained soils. PAW was best predicted by yield during water-stressed growing seasons, but the strength of the relationship was still not as good as on well-drained soils. For eroded back-slope landscape positions of claypan fields, this method greatly under-estimated PAW. We suspect these findings are because infiltration, and not just PAW, is a dominant variable impacting yield on hydrologically-complex claypan soils. Thus, modifications to this procedure would likely need to include adjustments accounting for variation in infiltration and profile recharge in order to generate more reliable estimates of PAW. These findings benefit producers and their consultants as they look for ways of translating site-specific information into ways to better manage water for crops. The findings benefit the general public as fields are managed for improved crop water utilization so that less surface runoff enters lakes and streams.
Technical Abstract: Plant-available water (PAW) capacity (PAWc) is one of the fundamental soil properties affecting crop yield, yet quantitative determination of PAWc at a field scale has been challenging. A Simple Inverse Yield Model (SIYM) has been devised and shown to be successful in estimating spatially-variable PAWc at a field scale for well-drained soils by matching simulated corn yield with measured yield. For other soils, however, SIYM has yet to be tested. Our primary objectives were to investigate (1) the relationships between PAWc and corn (Zea mays L.) yield, and (2) SIYM performance in estimating PAWc for poorly-drained claypan-soil landscapes. Soil PAWc to a depth of 1.2 m (PAW1.2) was measured at 19 and 18 sampling locations for two claypan-soil fields, Fields 1 and 2, respectively. Corn yield maps of the two fields (total of nine site-years between 1993 and 2003) were used with the model to estimate PAWc. The correlations between corn yield and measured PAW1.2 were significant for only water-stressed years. The regression r^2 values between SIYM-estimated PAW1.2 and the measured PAW1.2 were 0.43 for Field 1 and 0.31 for Field 2 with estimated errors of 18 mm and 50 mm, respectively. The large errors for Field 2 were primarily due to low PAW1.2 estimates at highly-eroded backslope areas, where corn yield was consistently low due to high clay content and hence high soil resistance to water storage. These claypan characteristics introduced additional variability in yield, especially in water-stressed years, which weakened the working assumption of the SIYM that PAW was the primary yield-limiting factor. Thus for areas where claypan dominates, this additional variability in yield needs to be addressed in order to use the SIYM to effectively predict PAWc.