Benchmarking On-Farm Soybean Water Productivity in Nebraska

Impact on Soybean Production and Value

From the large amount of data collected, we will be able to help soybean producers determine which of the commonly used agronomic practices can be fine-tuned to capture more of the yield potential in their specific fields – that is, close their field-specific yield gaps!

James Specht and Kenneth Cassman (2013)

Study Objectives

Our objective was the development of an easy-to-use framework to diagnose and improve soybean water productivity (seed yield per unit of seasonal water supply) in rainfed and irrigated fields in various areas in Nebraska.


We have collected and summarized soybean yield, plus applied irrigation (I) water amount, and nitrogen (N) amount for hundreds of soybean fields as reported by producers to their Natural Resources District (NRD) for as many years as the NRDs has required those producer reports. In addition, we have conducted personal interviews with many producers across the state to also collect yield, water, and N data for specific fields, but in this case we collected much more data on the crop management practices used by the producer in each specific field. This collective set of data allows us to determine what specific practices the "average" Nebraska rainfed or irrigated producer uses in an "average" field that (aside from the field's soil type and that year's weather) might, on average, have been responsible for the gap between the actual yield achieved in that field and the potential yield that was possible for that field based on our soybean crop models. Major findings to date are (1) an earlier planting date is a critical factor for reducing the yield gap, (2) soybean yield response to rain (or rain plus irrigation) "averages" about 3.0 bushels per acre per inch of water, but in some well-managed fields it approaches an estimated water productivity boundary limit of about 3.9 bushels per acre per inch of seasonally available water, and (3) the "average" yield limit boundary (i.e., the upper yield limit in Nebraska currently "averages" about 81 (irrigated) or 62 (rainfed) bushels per acre).

Below is a graph of the producer-reported actual soybean yield (vertical axis) in the fields where the producers provided us with their actual management practices versus the potential soybean yield (horizontal axis) in those fields – the latter yield is estimated based on our crop simulation models, which is an estimate of what the yields could have been with the same weather.

The pink triangles represent the intersection of actual yield obtained by producers for fields planted at some given producer-chosen planting date (again vertical axis) versus the potential yield that could have been obtained if the producer had planted those fields on May 1 each year (Ypo). The yield benefit of early planting seems low (only 2 bu/ac on average), but one must remember that many of the producers with actual yields ABOVE the 63 bu/ac average (horizontal line) are ALREADY planting many of their fields early (many, in fact in the last week of April), so their Ypo pink triangle positions do not move much to the right of their green triangle counterparts. However, below the 63 bu/ac mean, the pink triangles are far to the right of their green triangle counterparts!

Below is a graph of the producer-reported on-farm actual soybean yield (vertical axis) for various rainfed and irrigated fields versus the total amount of seasonal water supply (horizontal axis). Obviously, irrigated fields (blue) are not typically water-stressed during reproductive development because of well-timed irrigated dates and amounts, and thus have the highest yields in this graph. The leftmost red line is a best-fit boundary function for the steepness of soybean yield response to water, reflecting some of the best fields in NE (i.e., 3.9 bu/ac per inch of seasonal water supply). Rainfed and irrigated field data points near that red line reflect two things we all want – the highest yields possible for each inch of water (rain or rain+irrigation), which we call water-use efficiency or WUE. One of our goals in this research is to dissect out of the field data those agronomic practices that are critical for achieving a high WUE.

Note that the median steepness is about 3.0 bu/ac per inch of seasonal water supply. The term median is used to denote the fact, that 50% of the fields fall below and 50% above that median value. You could say, that in a collective way, that a rainfall event increases yield by about 3 bu/ac per inch of received rainfall, or the absence of a rainfall event can decrease yield by the same amount per inch of NON-received rainfall – the line is linear!

Finally, note also the so-called "plateau values" for the two red lines, which typically translate into the degree to which producers can effectively close the yield gap. With irrigation to supplement ill-distributed seasonal rainfall events, yield is higher (here 81 bu/ac) and consistently so than it is in rainfed fields (here 62 bu/ac). This data will be very useful in determining what producers can do in production management to shift their field yields both leftward (to get the same yield for less water), or shift their field yields upward (to get higher yield for the same amount of water), of in the end, shift both leftward and upward to achieve a steeper red line, which means greater WUE but effectively translates into more "crop per drop". Of course, in profitability terms: more revenue, less expense!

Additional Information

Drivers of spatial and temporal variation in soybean yield and irrigation requirements in the western US Corn Belt

Soybean yield gaps and water productivity in the western U.S. Corn Belt