Even though the price of nitrogen fertilizer has dropped some in the past year, the lower price of corn means that decisions about N management need to be made carefully, with an eye toward maximizing the return to this critically important input.
The return of dry weather over the past week and the forecast for the coming week has lessened the concern about N loss, though we still need to consider the possibility that some fall-applied anhydrous ammonia might have moved out of the upper part of the soil. If rainfall amounts are average or below-average in the next weeks, some N that has moved down but is still in the soil might still be available to the crop later on.
How much N does the crop need?
The first question on N management is rate ‚Äì how much N will the crop need, and then, how much of it needs to come from the fertilizer we apply? The generally-accepted rule of thumb is that the crop will take up a total of about 1 pound of N for each bushel of yield. We‚Äôve found a similar number in a few studies we‚Äôve done. We normally expect to find the maximum amount of N in plants at maturity, but depending on how the season ends, plant N might peak a few weeks before maturity.
If we expect yield of 200 bushels per acre and accept that the crop will take up 200 lb. of N, does that mean we will need to apply at least 200 lb. of N? Unless we‚Äôre growing the crop hydroponically without soil, or in some growth medium without organic matter, the answer is no. All naturally-developed soils have some organic matter, and many of our more productive soils in Illinois have 3 to 5% organic matter. An acre of soil a foot deep weighs about 4 million pounds, so a soil with 3.5% organic matter in the top foot has about 140,000 lb. of organic matter. About 5% of soil organic matter is N, so this soil would have about 7,000 lb. of organic N in the top foot.
The N in soil organic matter is a tremendously valuable resource, but predicting how much of it will become available to the corn crop each year is not easy. The general rule of thumb is that about 2% of it becomes available, which in our example would be 140 lb. of N per acre. The amount released can range from 1 to 3% per year, depending on soil temperature and moisture, and perhaps to some extent on what microbes are present. And not all of the released N may be available to the crop. Some N is released late in the season after the crop stops taking up N, or into parts of the soil where roots are no longer active. Some released N can also be lost to leaching or denitrification if soils stay wet.
Corn yields without N fertilizer can provide as estimate of how much N the soil provides. In the long-running rotation x N rate trial in place at our Monmouth research center, yields of corn following corn without N fertilizer (treatments stay in the same plots so these haven‚Äôt had N for 34 years) between 1983 and 2015 averaged 77, with a range of 25 to 116 bushels per acre. For corn following soybean, the average was 144 bushels per acre, and the range was 91 to 228. Soybean residue doesn‚Äôt tie up N, so yield for corn following soybean probably measures N availability better than corn after corn. In that case the soil is providing about 140 lb. N per acre per year, which in that soil is roughly 2% of organic N per year.
How much N needs to come from fertilizer?
A great deal of new data were recently assembled and some older data were removed in order to help update the database used by the N rate calculator http://extension.agron.iastate.edu/soilfertility/nrate.aspx to calculate the N rate expected to provide the maximum return to N (MRTN). A preliminary look at the numbers shows that the MRTN will not change a great deal as a result of adding the new data; our very large Illinois database means that it doesn‚Äôt change very fast as new data are added. For corn following soybean in central Illinois, the middle of the MRTN range is around 170 lb. N per acre. It‚Äôs similar to that for southern Illinois, and about 20 lb. less than that in northern Illinois. For corn following corn, the MRTN is about 200 lb. N/acre in all of Illinois.
Across hundreds of trials there is surprisingly little correlation between optimum N rate and yield; we have seen trials where 150 lb. of N or less have produced yields above 250 bushels per acre, and some where it took 225 lb. of N to produce yields of 150 bushels per acre. The only reasonable explanation for such a wide range is that the soil sometimes supplies more N than usual, and that sometimes there is either loss of N (from fertilizer N or organic N or both) or some of the N in the soil is unavailable due to such factors as poor root systems or dry soils that keep water from moving to the roots carrying N with it.
Form and timing
While we can precisely control the amount of fertilizer N we apply, its availability is affected by weather, soil, and crop growth, so is also somewhat unpredictable. Form, timing, and placement of N fertilizer all can affect availability, usually in ways that we can understand. For example, ammonium stays on soil exchange sites so moves little in the soil while nitrate can move easily with water. But if it doesn‚Äôt rain much, nitrate stays in the soil just fine. So knowing the basics of how different fertilizer materials behave can only take us so far; what really happens to N in the soil is heavily dependent on weather, and our predictions regarding N form and timing are little better than our ability to predict weather.
Although predictions of how fertilizer management affects N availability may not be great, we do think that research over a range of sites and years will help us manage better. We started a fairly large study in 2014, sponsored by the Nutrient Research & Education Council (with funds from the Illinois fertilizer checkoff) to look at the effect of N form, timing, and placement on corn yield. Table 1 shows yields averaged over three sites and two years, 2014 and 2015.
Yields averaged across treatments were between 213 and 230 at all sites except DeKalb in 2015, which had cool, wet weather and averaged 185 bushels per acre. The 150-lb N rate was chosen to be less that the rate needed to produce the maximum yields, making yield more sensitive to N availability from each treatment. Rainfall between mid-May and late June was somewhat above average at all three sites in 2014 and well above average at all three sites in 2015. So we think that in all of these trials, especially in 2015, the potential for N loss was greater than normal.
Averaged across sites, yields ranged from 208 to 219, or by only 11 bushels per acre. Yield ranges within individual sites were generally larger than this, but individual treatment effects weren‚Äôt very consistent over sites, so the overall averages are in a relatively narrow range. The inconsistency among sites shows up in the statistical analysis ‚Äì the more the rank among treatments changed among sites, the lower the chances of seeing statistical differences, and the larger the difference has to be before we can say it‚Äôs due to treatment instead of just to chance.
Even with the variability across sites, there were some treatments that gave higher yields than other treatments. Surprisingly, the two treatments that produced the highest yield (219 bushels per acre) were dry forms: urea with Agrotain¬Æ and SuperU¬Æ. Both of these contain urease inhibitor and SuperU contains a nitrification inhibitor as well. But statistically, these treatments did not produce higher yields than the check (UAN injected at planting), and in fact none of the first six treatments listed after the check yielded statistically less than the dry treatments or the check. These included delaying all of the N until sidedress time (V5) and split applications with 100 lb UAN at planting and 50 lb applied as UAN at V5 or V9, or urea + Agrotain broadcast at V5. Waiting until V9 to broadcast urea + Agrotain yielded slightly less than the best treatments. Ammonia (NH3) applied early also yielded among the best treatments, but only if it had N-Serve; without N-Serve it yielded a few bushels less.
Only two treatments yielded significantly less than the check ‚Äì UAN either surface-banded at planting or UAN with Agrotain broadcast at planting. Neither of these is typical; we included them to see how well preserved the N in UAN might be if all applied without injection at planting.
Beginning in 2015 we included treatments with 100 lb. UAN injected at planting and the additional 50 lb. applied as UAN at tasseling time. We applied these as a surface band (dribbled) either in the row middle or into the row. The in-row application yielded more than the mid-row application, but the in-row application did not produce significantly higher yield than the check (UAN injected at planting.)
While we saw some small differences among treatments included in this study, commonly-used timing and forms of N all produced similar yields, even under what we would consider high-loss conditions. This was at an N rate of only 150 lb. N per acre; the higher N rates used by most producers should have provided enough N to the lower-yielding treatments to bring them even closer to the higher yields produced by the better treatments. In fact,¬†the N rate trials included at each site showed that 170¬†lb. of N per acre, applied as UAN injected at planting,¬†increased yields by an average of 5 bushels per acre compared to 150 lb. of N.
While we would have expected larger differences among yields from these treatments, these results that show that both the risk of N loss and the benefit from delaying some of the N or using inhibitors may be a little less than we‚Äôve thought. Getting data from another year or two will help paint the picture more fully, but these results give some reason to be confident that the N management systems in common use all have good potential to provide the crop with N. Adding costs by changing N management, for example by making another trip over the field to apply late N, may not provide a positive return compared to applying all of the N in one or two earlier trips.
It rained on Easter (March 27) this year, which according to the old adage means rain for the next seven Sundays after that. The warm, dry, sunny day most of us enjoyed on April 17 tells us that won‚Äôt happen this year; in fact, the dry pattern that has now been in place for more than a week may persist, even though these is some rain in the forecast. I mention this only as a reminder that some of the N management systems that have been promoted, including late applications of surface-applied N, can mean reduced or delayed availability of fertilizer N to the crop under below-normal rainfall periods. The weather could change, of course, but this is one of the risks, along with increased costs, that increasing the complexity of N management can sometimes bring.