Quadratic Plateau Nitrogen Model
St Clair District SCIA Project - Assessing Nitrogen Requirements in Corn Using the Quadratic Plateau Model

The Quadratic Plateau model was investigated as a method to better predict nitrogen requirements for corn.

Plot layout: Field length strips of several nitrogen rates were established. The treatments consisted of 3 or 4 nitrogen rates below the growers normal rate of nitrogen and one rate above. Check strips (zero nitrogen strips) could have up to 30 lbs. N/ac applied in the starter. Data was analyzed using both the quadratic and quadratic plateau models.

Results From 2000: There were 6 cooperators for a total of 7 plots in 2000. (see Table 1). One plot was lost. Two plots did not level off in yield at high N rates, and the quadratic plateau analysis could not be used. Three other sites generated typical curves for both models, although the wet season increased the variability at one site. The plateau model gave a lower maximum economic rate of nitrogen (MERN) than the quadratic model. The last plot had poor yields with no response to nitrogen. This plot had 60 lbs. N/ac before planting, more perhaps than the MERN would require.

Results From 2001: There were 7 cooperators this year with a total of 8 plots. The 2001 growing season was hot and very dry. Rainfall was variable. Some plots received very little rain and a few received near adequate amounts. This resulted in normal yields in some fields and yields more than 50% below normal in others. High wind in the fall caused some lodging in most plots, but lodging was not significant. The dry growing season made it very difficult to draw significant conclusions from the data. Moisture was the limiting factor in most plots, not nitrogen. Yield responses are due to moisture differences within fields rather than a nitrogen response. Scatter in the data make it difficult to predict the MERN. Variability in the yield responses made it difficult to calculate the quadratic curves, and in most cases impossible to calculate the quadratic plateau. In some of the replications the quadratic curve actually curved up instead of leveling off. The MERN could be determined using the quadratic model in five plots. Only two plots could be analyzed using the Quadratic plateau model. The MERN for these two plots was higher using the quadratic model than using the quadratic plateau.

Comparison of Nitrogen Recommendations - 2000

Summary: The number of plots and participation across the region was excellent. The plots were done well, and most had 2 replications. Unfortunately the weather during the two years of the project was extreme on both ends of the scale. 2000 was very wet and 2001 hot and dry. The valid data sets prove that there are opportunities to reduce the rate of nitrogen in some fields. In 2000 the Seys plot showed a reduction of nitrogen from 180 kg/ha to 110 kg/ha. In 2001 the Annett-Augrim plot indicated a nitrogen reduction from 170 kgN/ha to 100 kgN/ha, and the Buurma plot from 170 kgN/ha to 90 or even 40 kgN/ha (see table above or graphs below). The quadratic plateau may be a good approach to determine the most appropriate rate of nitrogen for a whole field. It does not address variability of nitrogen within a field. This data will be compiled with other nitrogen research in Ontario to further fine tune nitrogen recommendations. With better weather more would have been learned from this project. The region is interested in continuing this project for 2 more years. Returning to some of the same fields from the first 2 years will help to test the MERN's predicted.