Fig. 1.

Locations of 125 on-farm evaluations trials conducted during 4 yr across Iowa to calibrate the late-season N diagnostic tools such as the corn stalk nitrate test and digital aerial imagery of the corn canopy to economic yield response to additional N.

 


Fig. 2.

Sampling areas (5–12 by 30 m) for collecting corn stalk nitrate test samples and extracting green reflectance values of the digital aerial imagery of the corn canopy. Each trial had two N rates: normal (defined as high or low) and normal minus 56 kg N ha−1 (defined as low) or normal plus 56 kg N ha−1 (defined as high). The lower rate is shown as the lighter strips and the higher rate is shown as the darker strips on the image. The sampling areas were selected to represent the wide range of variability in reflectance of the corn canopy within each trial.

 


Fig. 3.

Relationship between corn stalk nitrate test (CSNT) values and yield response (YR) to additional N observed in 125 on-farm evaluation trials conducted in (A) 2010, (B) 2009, (C) 2008, (D) 2007, and (E) data pooled for all 4 yr. The dotted vertical lines show the currently used CSNT categories and dashed horizontal lines show the marginal cost of additional N.

 


Fig. 4.

Relationship between relative green reflectance (RGR) and yield response (YR) to additional N observed in 125 on-farm evaluation trials conducted in (A) 2010, (B) 2009, (C) 2008, (D) 2007, and (E) data pooled for all 4 yr. Relative green reflectance values were expressed as a ratio of green reflectance at the low N rate to that at the high N rate. The dashed horizontal lines show the marginal cost of additional N.

 


Fig. 5.

Average monthly rainfalls for the trial locations studied during 4 yr and average monthly rainfall across Iowa during the last 30 yr.

 


Fig. 6.

Identifying critical values that separate profitable and unprofitable yield response (YR) for late-season measurements of corn N status using (A) the corn stalk nitrate test, (B) relative green reflectance (RGR), and (C) green reflectance (GR) of the corn canopy. The regression lines represent field-level multilevel binary logistic regressions, parameters of which were estimated from intercepts that varied and slopes that were constant across 125 trials conducted during 4 yr. Yield responses >0.31 Mg were considered as profitable due to application of an additional 56 kg N ha−1. Specificity is a percentage of correctly predicted unprofitable YR; sensitivity is a percentage of correctly predicted profitable YR.

 


Fig. 7.

Identifying critical corn stalk nitrate test values that separate profitable and unprofitable yield response (YR) when corn and N prices deviate by 30% from their long-term averages. The regression lines represent field-level multilevel binary logistic regressions, parameters of which were estimated from intercepts that varied and slopes that were constant across 125 trials conducted during 4 yr. All parameters for the field-level models were statistically significant at P < 0.001 level.

 


Fig. 8.

Differences in critical corn stalk nitrate test values separating profitable and unprofitable yield response (YR) for five N management categories created based on a common combination of timing of application and N form: fall-injected liquid swine manure, sidedress N (sidedress anhydrous NH3 or urea–NH4NO3), fall-applied anhydrous NH3 (fall AA), and spring-applied urea–NH4NO3 (spring UAN). The regression lines represent field-level multilevel binary logistic regressions, parameters of which were estimated from intercepts that varied and slopes that were constant across 125 trials conducted during 4 yr. All parameters for the field-level models were statistically significant at P < 0.01 level, except the intercept for sidedress N.