Predicting Nitrate-Nitrogen in the Two- to Three-Foot Soil Depth from Nitrate Measurements on Shallower Samples

Preplant soil profile nitrate-N tests are effective for predicting the response of corn (Zeu mays L.) to applied N in Wisconsin. The difficulty of obtaining profile (0 to 3 ft) soil samples is one of the main obstacles to more extensive use of preplant tests. One approach to overcoming this barrier is to develop and use statistical models to predict nitrate-N in the 1- to 3-ft depth from soil nitrate-N measurements in the 0- to l-ft and 1- to 2-ft depths. The study used 2306 soil samples collected over 3 yr from farm fields and 207 samples from a separate 2-yr research study to correlate nitrate-N contents in various soil depth increments. All samples were from medium- and finetextured soils, and the previous crop was usually corn. Six linear regression models that included factors for previous crop, geographical location, and legume and manure N-credits were evaluated. Models were developed to predict second- and third-foot nitrate-N contents from nitrate measurements in the first or second foot. During three climatically different years, distribution of nitrate-N within the 3-ft soil profiles was similar, but total profile nitrate-N content varied significantly. Models using second-foot nitrate-N content to predict third-foot nitrate-N following corn or soybean (Glycine max L.) are reliable (r² = 0.70 and 0.52, respectively) and are suitable for routine use within the preplant test. The average deviation of predicted third-foot nitrate-N from measured values for all previous crops was 12 lb N/acre. Predicted third-foot nitrate-N contents were within 20 lb N/acre of observed values at over 80% of the sites and within 40 lb N/acre of observed values at over 95% of the sites. Prediction of second- and third-foot nitrate-N contents from first-foot measurements is not reliable enough to allow use of a 1-ft sampling depth for the preplant test, unless a greater margin of error in N recommendations is tolerated. A 2-ft sampling depth for the preplant test and use of models to predict third-foot nitrate-N content for corn following corn or soybean is feasible and should promote more widespread use of this test. Research QuestionImplementation of preplant soil nitrate tests in humid regions such as Wisconsin provides an effective method for adjusting corn N recommendations for carryover nitrate-N. Use of the test can provide economic and environmental advantages, but major obstacles to more widespread use of the test in Wisconsin are the time and labor needed to obtain soil samples to a 3-ft depth as required for the test. The objective of this work was to determine if nitrate-N measurements on soil samples from the 0- to 1- or 1- to 2-ft depths could be used to reliably predict nitrate-N at greater profile depths, thereby eliminating the need to collect the deeper samples. Literature SummarySoil nitrate tests are a useful approach for improving N management for corn in humid regions of the USA. Preplant nitrate tests require sampling the corn root zone to a sufficient depth to measure or estimate the amount of residual nitrate-N available for crop use. A minimum sampling depth of 2 ft is frequently recommended for soil nitrate testing and the predictive value of the test is often improved by including the amounts of nitrate in deeper soil layers. Study DescriptionThis study used 2306 soil samples collected over a 3-yr period from farm fields and 207 samples from a separate 2-yr research study to correlate nitrate-N contents in various soil depth increments. These two soil profile nitrate databases allowed evaluation of approaches for predicting 1- to 2-ft and 2- to 3-ft nitrate with two independent data sources. Statistical models were developed from these data to predict second- and third-foot soil nitrate-N contents from first- or second-foot nitrate-N measurements. The predictive models evaluated the effects of previous crop, geographical location, and legume and manure N credits on prediction of subsoil nitrate-N from surface soil measurements. Agreement between the predicted and measured soil nitrate-N contents (pounds per acre) was emphasized more heavily than correlation coefficients alone in selecting the most useful models for predicting subsoil nitrate contents. Applied QuestionsIs prediction of third-foot nitrate from first- or second-foot soil nitrate measurements feasible? The best predictions of soil profile nitrate-N content involved prediction of third-foot (2 to 3 ft) nitrate-N from second-foot (1 to 2 ft) nitrate-N measurements (Table 1). This finding indicates that a 2-ft sampling depth can be reliably used in a preplant soil nitrate test when third-foot nitrate-N is estimated with a predictive model. This study found that predicting nitrate contents in the second- and third foot of soil profiles based on nitrate measurements in the first foot is not reliable. Although first- and second-foot nitrate contents were related, the relationship between first- and third-foot nitrate-N contents was not strong enough to have useful predictive value. Does prediction of third-foot nitrate vary with annual change in total soil profile nitrate content? Total profile (0 to 3 ft) nitrate-N content varied substantially during the three climatically different years included in this study. Distribution of nitrate-N with depth within the soil profiles, however, was similar among years. This indicates that the same model can be used to predict third-foot nitrate-N content across years with varying total profile nitrate accumulation (Table 1). Is prediction of third-foot nitrate influenced by soil-crop management factors? Previous crop strongly influenced the relationship between second- and third-foot nitrate-N content. Separate predictive equations probably will be needed for different previous crops, and methods for predicting third-foot nitrate-N for previous crops of corn and soybean were developed in this study. Although the relationship between second- and third. foot nitrate-N following alfalfa was different than following corn or soybean, the preplant nitrate test has little value in this crop sequence because samples are taken before N in legume residues is released. Table 1Predicted third-foot nitrate-N from second-foot nitrate measurements in 3 gr and using combined data.† Predicted third-foot nitrate-NMeasured second-foot nitrate-N198919901991Combinedlb N/acre------------------lb N/acre------------------- 2020242324 4034384139 6046525953 8062667667No. of samples1455525521249Deviation from normal DreciDitation. in‡ −6.7 + 5.2 + 7.6†Previous crop was corn. ‡Departure from statewide 30-yr mean value of 31.07 in.