Illinois Fertilizer
Conference Proceedings
January 27-29, 1992
| Home | 1992 Index | Search |
|
|
| Home | 1992 Index | Search |
H.M. Brown, R.G. Hoeft, and E.D. Nafziger1
Current N recommendation systems based on yield goal with adjustment for previous legume crop or manure applications have worked well in many situations. For example, use of the average optimum yield from an eight year study at Monmouth times 1.2 lb N/bu would result in a recommendation within 201bs N/acre of the optimum in 6 of the 8 years (Table 1). The two exceptions occurred in the drought years of 1988 and 1989. While these results provide confidence that the system works well in most years, research and farmer reports of near optimum yield with little or no applied fertilizer N makes one question the system. This is especially true when one determines that lack of expected response cannot be explained on the basis of previous crop, manure application, or poor yield (Table 2). These situations, coupled with economic and environmental concerns require the development of an improved system for determining N needs in the more humid environments of the midwest. A system is needed that would allow farmers to identify those soil conditions where current N recommendation rates should be adjusted.
Soil profile NO3-N levels are being used to predict nitrogen needs in some states in the more humid areas of the midwest. Iowa has adopted the procedure outlined by Magdoff et al., 1984 and Blackmer et al, 1991. Nitrate-N concentrations of soil samples collected from the upper 12 inches of soil when corn is 6 to 12 inches tall are used to set N rates with a full rate being suggested at test levels below 10 ppm NO3-N and an adjusted rate recommended between 11 and 25 ppm NO3-N. Michigan suggests that farmers reduce N rates by 1 pound for each pound of NO3-N found in the upper 2 feet of soil in samples taken in late March or early April (Vitosh et al., 1988). Bundy and Malone (1988) have also demonstrated the value of using NO3-N concentrations in the top 2 or 3 feet of soil in early spring to adjust N rates in Wisconsin.
The objective of this project was to evaluate the accuracy of three systems-
namely yield goal times 1.21b N/bu (YG), preplant soil NO3-N concentration
(PPNT), and presidedress NO3-N concentration (PSNT)- for predicting
nitrogen rates necessary for optimum corn production. Information presented
in this report represents two cropping years.
Experiments were conducted at 48 sites throughout Illinois over the 1990 and
1991 crop years. Sites were selected to provide a range in soil texture, yield
potential, previous crop, and manure application.
Soil samples were collected for NO3 and NH4-N analysis from each
site in late March/early April at depths of 0-6, 6-12, and 12-24 inches. Nitrate
values (ppm) from the upper 24 inches were multiplied by a factor of 8 to estimate
NO3-N content in lb/acre. Samples were also collected when corn was
6-12 inches tall from the top foot of soil at each location and analyzed for
NO3 and NH4-N. Individual plots 10 feet (4 rows) x 50 feet were arranged
in a randomized complete block design using 4 replications.
Nitrogen was applied at sidedressing using six rates of N evenly distributed from 0 to 100% of normal recommendation in 1990 and from 0 to 125% of normal recommendation in 1991 . The recommended N rate for each site was determined by multiplying the soil's productivity level under high management (Fehrenbacher et. al 1984) by 1.2 pounds N per bushel, minus adjustments for previous crop or manure application. Nitrogen treatments were injected as urea-ammonium nitrate solution (28% N) using a 60 and 30-inch knife spacing in 1990 and 1991 respectively.
The center two rows of each plot were thinned to a uniform population shortly after the nitrogen treatments were applied. Ear leaf and grain samples were collected at silking and harvest for N analysis. Grain yield was determined by hand harvesting 30 and 20 feet of the center two rows in 1990 and 1991, respectively and adjusting the shelled weight to 15.5 standard moisture.
Economical optimum yield was calculated for each location as that point on
the quadratic response curve where the slope was equal to a cost:price ratio
of 0.08 using $.20/lb N and $2.50/bu corn (Nafziger et al., 1984). The optimum
N rate was considered to be zero where there was no significant response to
N.
Soil NO3-N content in the upper 24 inches of soil ranged from 9 to 300 lbs N/acre in 1990 and 4 to 1961bs N/acre in 1991. The highest level of NO3-N in both years were observed at sites where manure had been applied the previous fall.
The relationship between NO3-N concentrations in the 0-12 inch zone with that in the 12-24 inch zone was erratic, suggesting that use of the NO3-N concentration in the 0-12 inch zone to, represent concentrations in the upper 24 inches will be of limited value. This would be especially true for those areas where low yields were followed by a wet fall that would result in the movement of carryover N below the upper 12 inch depth, but not out of the rooting zone.
Early June NO3-N concentration in the top 12 inches of soil ranged from 0.3 to 55.4 ppm in 1990 and 1.6 to 51.5 ppm in 1991. The highest levels were observed on sites where manure was applied the previous fall.
Nitrate-N concentrations in the upper 12 inches of soil were less than 10 ppm at 33 of the 48 locations. When one considers only those fields that did not have manure or a previous legume, 85% had NO3-N concentrations less than 10 ppm.
Early spring and presidedress NO3-N concentrations were highly correlated (R2=0.74), but the presidedress samples were about 50% higher than the early spring samples, reflecting the rate of mineralization that occurred between sampling periods.
The relationship between yield obtained on the untreated plot as a percent of optimum yield and NO3-N concentration in the PPNT 0-12 and 12-24 inch samples and the PSNT 0-12 inch samples was very low, (R2=0.08,0.12,0.02 respectively).
Yield Goal
Recommendations projected by using the yield goal system (YG) over-estimated optimum needs by 62 lbs N/acre over all sites with an average excess of 861bs N/acre at 35 sites and an average shortage of 21 lbs N/acre at 10 sites. Further breakout of the experimental locations into responding and non-responding groups allowed a closer evaluation of how well the recommended N rate described the actual optimum needs (Table 3).
Thirty of the 48 experimental sites responded to increasing increments of N fertilizer. Of the 30 responding sites, needs were over estimated by 22 and 35 lbs N/acre when the previous crop was soybeans and corn, respectively. Nitrogen needs were underestimated by 41 and 14 lb N/acre when the previous crop was double crop wheat/bean and vetch, respectively. At the 18 non-responding sites, over application averaged 133 lb N/acre.
Preplant
Use of PPNT results would have resulted in an over application of 261bs N/acre across all sites with an average excess of 57 lbs N/acre and an average shortage of 34 lbs N/acre. When looking at only the 30 responding sites, the PPNT recommended an over application 11 lbs N/acre. Further division of the responding sites into previous crops of soybean and corn shows a shortage of 13 and an excess of 8 lbs N/acre respectively. Over application of 40 lb N/acre would have resulted from use of the PPNT when the previous crop was double crop wheat/beans.
The PPNT detected elevated levels of NO3-N in 9 of the 18 non-responding sites, but not but not at levels high enough to provide optimum nitrogen for the yield produced. Fall application of manure the previous year accounted for 3 of the 9 sites. Low soil NO3-N concentrations measured on the two non-responding sites where alfalfa was the previous crop suggest mineralization of nitrogen from the decomposing alfalfa crop had not occurred when samples were taken.
Presidedress Test
Use of the PSNT would have resulted in an over application of 54 lbs N/acre across all sites with an average excess of 73 lbs N/acre and an average shortage of 22 lbs N/acre.
The PSNT over estimated N needs by 27 and 191bs N/acre when the previous crop was corn and soybeans, respectively. This test recommended 10 lb N/acre less than needed when the previous crop was double-crop wheat/beans.
The PSNT test detected high concentrations of NO3-N on sites with a previous history of manure application. Following soybeans and manure, NO3 levels (47 ppm) were nearly twice that suggested as optimum by Magdoff 1984 and Blackmer 1991. At a site that received fall application of a bedding straw/manure mix, NO3-N levels were 10.8 ppm.
The three systems evaluated were reasonably accurate in predicting N rates for the 30 responding sites. The PPNT consistently predicted lower N rates than either the YG or PSNT. This lower recommendation probably resulted from subtracting a full credit for manure or previous legume from the normal recommendation plus a full credit for the NO3-N found with the PPNT. That portion of the PPNT N03-N that resulted from mineralization of the manure or legume N would thus be double counted.
Of the three systems evaluated, PSNT provided the most accurate estimate of N needed at the responding sites. These improved predictions were the result of limiting the highest rate to 135 lbs N/acre rather than from adjustments based on NO3-N concentrations. The 135 lb N/acre limit is the midpoint of the recommended range for NO3-N levels of 10 ppm or less (Blackmer et al., 1991). Had one used the maximum recommendation, this test would have over recommended more N than either of the other two systems.
Yield of the unfertilized control as a percent of the optimum yield is plotted against soil PPNT and PSNT NO3-N concentrations in Figure 1 and 2. Using the procedure proposed by Cate and Nelson, the data are separated into 4 quadrants. As expected, most of the locations appear in quadrants 1 and 3. Those in quadrant 1 have low soil test and significant response to applied N while those in quadrant 3 have tests above the critical level and do not respond to applied N. When this procedure works perfectly, there are no data points in either quadrant 2 or 4. That was true for quadrant 2 which would contain locations that have a high test, but still respond to added N, but was not true for quadrant 4 that contains 15 locations that have a low test, but did not respond to applied N. Lack of response to applied N is not new for N rate studies (Fox, et al., 1989, Magdoff, 1990), but often such results are discounted as anomalies.
An evaluation of the data from this study indicated that 13 of the non-responding sites could be explained by drought, manure application, or previous crop. Drought conditions limited yield at seven experimental sites over the past two years (Table 4). Optimum yield across the seven sites was below established yield expectations by an average of 39 bu/acre. All seven drought-affected sites did not respond significantly to increasing increments of nitrogen fertilizer. Post-harvest soil samples did not indicate a carry-over or an accumulation of NO3 or NH4-N in the profile to a depth of 3 feet.
Three of the four sites with a fall application of manure were correctly predicted not to respond to nitrogen fertilizer. The fourth site was not correctly predicted by any of the three systems evaluated. An unknown amount of surface applied bedding straw/manure mix (not incorporated) may have been underestimated in nitrogen credit (YG) or the large amount of straw may have delayed mineralization until sometime after PPNT or PSNT samples were taken.
Alfalfa was the previous crop on two of the non-responding sites. Soil test results from neither the preplant nor presidedress samplings detected concentrations of NO3-N great enough to produce optimum yields (Table 5). A greater concentration of NO3-N was detected with the PSNT compared to the PPNT suggesting that N mineralization occurred after early spring soil samples were taken.
The five remaining non-responding sites could not be explained using previous cropping/manure history or environmental conditions. Preplant, presidedress, or post-harvest soil NH4+ and NO3- concentrations did not offer an explanation for the non-response to nitrogen treatments.
Measurement of soil NO3-N concentration that includes easily oxidized
organic nitrogen may increase the predictive ability of soil NO3-N
testing. The current method used to measure soil NO3-N and NH4-N
quantifies only that available at the time the sample was taken. A continuation
of this project for 1992 will consider alternative methods of measuring soil
nitrogen to improve predictability. Evaluation of in-field devices to determine
N recommendations including a chlorophyll and NO3-N meter will also
be made in 1992.
Table 2: Effect of N rate on corn yield. Monmouth, IL
Table 5: Soil NO3-N concentration of samples taken according to the PPNT and PSNT by previous crop
Figure 1. % Relative yield vs nitrate-N (ppm) in upper 24 inches of soil profile
Figure 2. % Relative yield vs nitrate-N (ppm) in upper 12 inches of soil profile
1Research Associate, Professor, and Associate Professor, Department of Agronomy, University of Illinois, Urbana, IL.
Blackmer, A.M. T.F. Morris, D.R. Keeney, R.D. Voss, and R. Killorn. 1991. Estimating nitrogen needs for corn by soil testing. Iowa State University Extension Bulletin Pm-1381. Ames, Iowa.
Bundy, L.G. and E.S. Malone. 1988. Effect of residual profile nitrate on corn response to applied nitrogen. Soil Sci. Soc. Am. J. 52:1377-1383.
Fehrenbacher, J.B., J.D. Alexander, I.J. Jansen, R.G. Darmody, R.A. Pope, M.A. Flock, E.E. Voss, J.W. Scott, W.F. Andrews, and L.J. Bushue. 1984. Soils of Illinois. Agric. Exp. Sta. in Cooperation with the Soil Conservation Service, U.S. Department of Agriculture. Bulletin 778.
Fox, R.H., G.W. Roth, K.V. Iverson, and W.P. Piekielek. 1989. Comparison of soil and tissue nitrate tests for predicting soil nitrogen availability to corn. Agron. J. 81-971974.
Magdoff, F.R., D. Ross, and J. Amadon. 1984. A soil test for nitrogen availability to corn. Soil Sci. Soc. Am. J. 48: 1301-1304.
Magdoff, F.R., W.E. Jokela, R.H. Fox, G.F. Griffin. 1990. A soil test for nitrogen availability in the northeastern United States. Commun. in Soil Sci. Plant Anal., 2:(1316), 1103-1115.
Magdoff, F. R. 1991. Understanding the Magdoff pre-sidedress nitrate test for corn. J. Prod. Agric. 4:297-305.
Nafziger, E.D., R.L. Mulvaney, D.L. Mulvaney, and L.E. Paul. 1984. Effect of previous crop on the response of corn to fertilizer nitrogen. Journal of Fertilizer Issues. 1(4): 136-138.
Vitosh, M.L., D.D. Warnchke, D.R. Christenson, and J.G. Hall. 1988. Soil nitrate
testing : a guide for adjusting Michigan nitrogen recommendations for corn.
Proceedings of the 18th N. Central Extension-Industry Soil Fertility Workshop,
Nov. 9-10, 1988. Potash & Phosphate Institute, Manhattan, KS.
