Dustin L. Harrell, Trent L. Roberts, Richard J. Norman, Nathan A. Slaton, Chuck E. Wilson Jr., Tim W. Walker and Gary N. McCauley

Development of a nitrogen soil test has been a goal of soil scientists for as long as there has been a soil fertility discipline. Many soil test extraction procedures have been evaluated over the years, and attempts have been made to correlate soil test values to a crop’s yield response to applied nitrogen fertilizer. Some have had a limited success, but most have not.

Compared with other tests for essential nutrients like potassium and phosphorus, nitrogen soil tests have proved elusive because of constant changes and transformations of nitrogen in the soil and the difficulty of finding a stable pool of extractable soil nitrogen that correlates well with crop response to nitrogen fertilization. Nonetheless, the development of a reliable soil test is essential for the longterm sustainability of production agriculture.

Drill-seeded, delayed-flood rice production is the most common commercial method used in Louisiana, Mississippi, Texas, Arkansas and Missouri. In this type of production system, the initial and largest nitrogen fertilizer application occurs just before the permanent flood is established. The initial irrigation incorporates the nitrogen, and the anaerobic – without oxygen – nature of the submerged soil stabilizes the nitrogen in the system. Because of this, when water is managed properly, rice production has the highest nitrogen use efficiency of any other row crop at about 75 percent.

The high nitrogen-use efficiency of rice makes it an ideal candidate for the development of a nitrogen soil test if a stable, extractable nitrogen pool could be identified that correlates well with rice response to nitrogen fertilizer. To that end, scientists at the University of Arkansas conducted initial work on identifying an extraction method that cor correlated well with mineralizable nitrogen. They identified an alkaline nitrogen-extraction procedure that used direct steam distillation and had the desired correlation. The forms of nitrogen extracted were identified as amino-sugarnitrogen, amino acid-nitrogen, and ammonium-nitrogen. The extraction was coined N-STaR – for nitrogen soil test for rice.

After the stable nitrogen soil test extraction was identified, correlating and calibrating the soil test began. The process started in 2005 with a collaborative effort among scientists at the LSU AgCenter Rice Research Station, the University of Arkansas, Mississippi State University and Texas A&M University. They conducted trials at multiple locations across each state using small-plot, replicated nitrogen response trials. One of the major findings from these trials was that a 0- to 18-inch soil sample would be needed to obtain the strongest relationship of N-STaR extractable nitrogen with the rice response to fertilizer nitrogen on silt loam soils. In the fall of 2008, a completed calibration curve was developed for silt loam soils (Figure 1). Calibration curves were developed based on 100 percent, 95 percent and 90 percent relative rice yield response to nitrogen fertilization.

Once the calibration curve was created, the next phase was to test and validate the N-STaR calibration on commercial rice producers’ fields in each state. In Arkansas, validation studies took the form of small-plot trials on producers’ fields. Arkansas scientists found at 17 of the 18 locations tested that the N-STaR-recommended rates produced yields equal to or greater than the traditional nitrogen rate of 150 pounds of nitrogen per acre on silt loam soils. In 2011, N-STaR became a recommended practice in Arkansas.

Validation trials on commercial rice fields in Louisiana were conducted slightly differently. Scores of soil samples were taken on commercial producers’ fields with silt loam soils. Only soils which had very high N-STaR-extractable nitrogen and low nitrogen recommendations (less than 100 pounds per acre) were selected for the validation trials. These soils were selected in particular because if the current N-STaR soil test calibration would have any potential nitrogen recommendation issues, they would show up in this area of the calibration curve. Of all of the soils in Louisiana evaluated in 2011- 2012, approximately 20 percent of the silt loam soils fell into this category.

Validation trials were conducted on whole fields, generally 20 acres or more. N-STaR nitrogen recommendations from the 100 and 95 percent relative grain yield calibration curves were compared to nitrogen rates that the cooperating farmers historically used on their fields for a particular rice variety. Results of the two years of validation trials are summarized in Figure 2.

In 2011, the on-farm validation trials were excellent. Overall, six of the eight N-STaR recommendations tested produced yields equivalent to or exceeding the producers’ current practice. The 100 percent relative grain yield N-STaR recommendations at the GF&P and R&N Farms were less than one-third of the traditional rates used but produced higher yields.

One reason higher nitrogen rates can often result in lower yields is due to the increased disease pressure often observed with increased rates of applied nitrogen. This was evident with the validation trial at the GF&P Farm in 2011. Sheath blight disease was clearly visible at the 180-pound-per-acre rate and absent in the 60-pound-per-acre treatment.

The 2012 validation trials resulted in three out of the nine N-STaR recommendations producing equal or improved yields compared with traditional rates used by the producers.

Overall, the two years of validation trials indicated that the high N-STaR extractable nitrogen and low nitrogen fertilizer rate recommendation area of the calibration curve tested in the validation trials was correct approximately 52 percent of the time. Because this area of the calibration curve would be used for an estimated 20 percent of silt loam rice soils in Louisiana, it was determined that more data points in this area of the calibration curve would be needed prior to launching the N-STaR soil test as commercial best management practice in Louisiana.

Small-plot trials are being conducted this year and are planned for future years to obtain the needed data. Nonetheless, the N-STaR soil tests show great promise to improve fertilizer nitrogen use efficiency in Louisiana and across the Midsouth in the near future.

Dustin L. Harrell is associate professor and Mosaic Company Professor at the Rice Research Station, Crowley; Trent L. Roberts is assistant professor, University of Arkansas, Department of Crop, Soil and Environmental Sciences; Richard J. Norman, Nathan A. Slaton and Chuck E. Wilson Jr. are all professors in the University of Arkansas, Department of Crop, Soil and Environmental Sciences; Gary N. McCauley is professor, Texas AgriLife Research, Beaumont Research and Extension Center.

The authors recognize the following people for their contributions to the N-STaR validation trials in Louisiana. Without them, this work would not have been possible. James P. Leonards, Ronald P. Regan and Jacob S. Fluitt, research associates, Rice Research Station, Crowley; Stuart J. Gauthier, county agent, St. Martin Parish; Keith A. Fontenot, county agent, Evangeline Parish; Barrett A. Courville, county agent, Acadia Parish; Andrew L. Granger, county agent, Vermilion Parish. Funding for this project was provided by the Louisiana Rice Research Board, USA Rice Foundation, Arkansas Rice Research and Promotion Board, Mississippi Rice Promotion Board and the Texas Rice Promotion Board.

(This article was published in the spring 2013 issue of Louisiana Agriculture magazine.)