Decisions about the amount of nitrogen to apply in wheat and barley are challenging each and every year, as the return per acre is not simply a function of the price of the commodity but also on the quality (grain protein%) of those bushels. There are opportunities to capture premiums for protein but more often than not producers are faced with discounts as the grain protein percentages fall below the market's 14% threshold. While this was already an issue in 2008 with high yields in Northwest Minnesota leading to lower protein, it was greatly magnified in 2009 with producers reporting grain protein percentages of 10% or less.
This issue is not new since it has been long noticed that yield and protein are inversely related. The amount of grain protein produced per acre appears to be relatively constant over years. In high yielding years the extra starch produced simply dilutes the total protein produced per acre, leading to smaller percentages in the grain Unfortunately farmers are not paid for total production of grain protein per acre but rather they are paid for concentration in grain.
In the past two growing seasons nitrogen response trials were conducted along side the spring wheat variety trials in NW Minnesota. Three trials were laid out in both 2008 and 2009. The trial near Hallock in 2009 had to be abandoned due to excess rainfall. The variety 'Knudson' was planted in all trials. Nitrogen was applied as urea and incorporated before just prior to seeding at rates from 0 to 180 lbs of N per acre in 30lb increments. Phosphorus, potassium, and sulfur were applied at blanket rates to eliminate their potential effects on yields. Two foot soil samples were taken prior to treatment application to assess the impact of soil N on crop response. The previous crop was soybeans at all locations except Hallock 2008, which was winter wheat stubble.
Figures 1, 2, and 3 represent relative increases in grain yield, protein, and protein yield (total pounds of protein produced per acre) for all locations. Relative levels were used instead of actual levels to take out some of the field to field variability seen between locations. All locations are color coded on each graph and all comparisons are made between the amount of fertilizer N applied plus the amount of N in a 2 foot soil sample taken in the spring before treatment application.
Yield levels were high both years averaging 100 bu/ac average maximum yields across all locations. Optimum nitrogen rate varied from location to location and was reflective of the amount on N extracted in the 2 foot sample (Figure 1). The overall response to N across locations was to 155 lbs of total N below which the relative yield level dropped off sharply in a diminishing fashion at both locations in 2009 and the Hallock location in 2008 suffered the biggest yield penalty from under-application of N. In general, plots with no nitrogen only produced 20-60% of maximum agronomic yield at these locations. Grain protein responded in a different manner increasing from 80% of maximum linearly to 208 lbs of total N (Figure 2). This shows that on average it took an additional 50 lbs of N to maximize protein over what was needed for yield. When factoring in yield and protein to calculate protein production per acre (Figure 3) it took about 220 lbs of N to maximize the total amount of protein produced. If you factor in the average yield maxima of 100 bushels per acre we would factor 2.2 lbs. of N per bushel, which is close to the current recommendation of 2.5 lbs. per bushel.
The protein yield data illustrates the key problem associated with low grain protein in 2009. With the wet conditions and late planting in the spring of 2009 many producers were not expecting the high yields again that were seen in 2008. Many likely were optimistically utilizing long term yield goals of around 60 bushels per acre. Utilizing our current recommendation system, we would suggest 150 lbs of N per acre for that yield goal. That rate would agree with our N rate required for maximizing yield, but would have clearly undercut protein. Our studies did not show as large decrease in protein as producers experienced but 1) we were only using one variety; and 2) we were using small plots which may have influenced the overall protein levels (not necessarily the response to N) of the crop. But this clearly shows some of the limitations for a yield goal based system; it's challenging to choose the yield for the N recommendation formula that ultimately ends up being harvested in a year like 2009 when Mother Nature decides to skip summer all together (and by doing so creating ideal growing environment for wheat). The underlying challenge is to understand what the probability is that field history and planting date are reliable predictors for yield goal.
So what about any other contributing factors? The fall of 2008 was not ideal, but some farmers were able to apply nitrogen. With the wet spring there likely was a potential for N loss from denitrification on water logged fields or leaching if fields were drained. If the nitrogen was applied late enough so that thee was not much conversion to nitrate this should not have been a major issue early in the spring, but still cannot be discounted. Another key factor that should always be carefully watched is field conditions during fertilizer application. It must be remembered that N in anhydrous ammonia and urea both will go through stages where it is subject to gaseous loss. Proper incorporation or application depth will limit this risk. Two inches of soil is needed to limit this loss but ideally application of anhydrous should be at around 4 inches or more. If urea is surface applied and not incorporated a product like Agrotain has been shown to give a slightly longer window between applications and when the fertilizer should be incorporated mechanically or by rainfall. There also may be some benefits to a product such as Agrotain if urea cannot be applied before seeding and needs to be surface applied after the crop is planted. Air seeders offer more flexibility when fertilizer and seed can be applied in one step, but remember that concentrated bands of fertilizer urea can have negative effects on stand establishment. Increasing the width of the band the fertilizer is placed in or separating the fertilizer band from the seed will help lessen this risk.
Looking ahead to 2010 I am sure we will see N rates increase to avoid the low protein levels seen in 2009. While this reaction is understandable, it relates back to the question posed above -how robust are field history and planting date as predictors of yield potential for the next season. The yield data from the last two years would likely warrant this increase but there are risks involved. These risks include the increased risk of lodging from excessive N applications if yield levels are not the same as previous years. Variety selection therefore becomes even more important as applied amounts on N increase. The second option would be to apply N in-season to boost grain protein levels. Previous research has been shown that this will work we do not have a clear environmental or crop indicator to determine when it is most economical to apply. Most of the previous research has shown that foliar applications of N at anthesis can be used to boost grain protein content. It takes between 30-60 lbs of N to increase grain protein content by 1%. While it may not be the most ideal tool in the toolbox it may be the best if we run into years like 2009.
For more information on foliar N applications for spring wheat go to March 6th, 2006 Crop News article:
Late season applications of nitrogen in spring wheat