Illinois Fertilizer
Conference Proceedings
January 22-24, 2001
| Home | 2001 Index | Search |
|
|
| Home | 2001 Index | Search |
R.G. Hoeft, E.D. Nafziger, R.L. Mulvaney, S.T. Mirek, L.C.
Gonzini, and J.J. Warren1
Nitrate levels in excess of the public health standard of 10 mg N/l in public water supplies, along with concern about hypoxia in the Gulf of Mexico, have drawn renewed interest to improving fertilizer N use efficiency. Prior research has indicated that some farmers may be unknowingly contributing to these problems. The objectives of this research were to determine the effect of rate and time of N application on nitrate-N concentrations in water from tile lines; and to evaluate the effect of previous N rate on current N needs, on recovery of fertilizer N by plants, and on fertilizer N transformations in soil. To accomplish the objectives, tile line monitoring systems that record water flow rates and collect water samples on a predetermined schedule were installed at 10 experimental sites. At 11 sites, N rate studies were conducted when the field was planted to corn and at five of the 11 sites, 15N labeled fertilizer was applied to allow us to monitor the fate of fertilizer.
The following observations were evident from the data:
Some Midwestern producers may be unknowingly contributing nitrates to water supplies. On-farm research identified 13 of 77 fields in which corn was nonresponsive to fertilizer N. There was evidence to indicate that these fields had a previous history of high levels of fertilization and/or manure application (Brown et al., 1993). Based on these results, along with work by Torbert et al. (1992) showing that excess fertilizer N is assimilated into organic N compounds, and work by Stevens et al. (1997) demonstrating that these compounds mineralize more easily than native organic matter, we have theorized that these nonresponding sites likely had adequate N release from the soil to meet the needs of the crop. Continued application of optimum or above-optimum N rates on these fields will enhance the potential for nitrate movement through tile line drainage. Buzicky et al. (1983) demonstrated that above-optimum application rates increased the loss of fertilizer-supplied N in tile drainage water and that the problem was even greater when the fertilizer was fall-applied.
A survey of producers in the Sangamon River watershed in 1995 indicated that nearly 70 percent were applying 40 lb N/acre or more above the recommended level for corn, with some applying 100 lb N/acre more than recommended. A 1999 survey in the same watershed indicated that nearly 70 percent of the producers were within 20 lb N/acre of the recommended rate.
The objectives of the project reported in this paper are to:
Ten experimental sites having clearly defined tile systems that drain only that field or a known portion thereof were identified in 1997. Tile line monitoring systems that record water flow rates and collect water samples on a predetermined schedule based on flow rate were installed at each location. At four of the locations, air and soil temperatures are recorded on 5-minute intervals, and precipitation is recorded on 30-minute intervals at all locations. Past cropping records including yield, time and rate of N application, and crop rotation were recorded for each site (Table 1). Other than in the small plot area of the field, the farmers have been encouraged to continue to apply the same rate of N and to continue to manage the field in the same manner as in the past.
Small plot nitrogen rate studies were conducted at seven of the sites in 1997,
the same seven in 1999, and the other four sites in 1998 and 2000, using ammonium
sulfate in 40 lb N/acre increments. The tile system drained more than the field
in question at one of the 11 locations; thus, the yield and soils data are
presented for 11 sites, but tile data for only 10 sites. The 1999 and 2000
treatments were repeated in the exact location in the field as was used in
1997 and 1998. The total N fertilizer application ranged from 0 to 240 lb N/acre
at seven locations, from 36 to 276 lb N/acre at two locations, and from 45
to 285 lb N/acre at two locations. The differential in N rates was due to the
application of DAP by the farmer. The ammonium sulfate was applied at each
location near the time that the farmer made his application. At five locations, 15N-labeled
ammonium sulfate was applied to microplots within each N rate plot at the same
rate of N. The 1999 and 2000 15N plots were
moved within the small plot area to allow us to determine 15N
uptake from the 1999 and 2000 application as well as from the residual from
the 1997 and 1998 application. Corn was planted in mid to late April and thinned
to 29,000 plants per acre at the V-4 growth stage. At maturity, grain was hand-harvested
for yield determination.
At physiological maturity, whole plant samples were collected from the microplot
areas that had received the 15N-labeled fertilizer
and from the 0, 80, 160, and 240 lb N/acre rate plots that had received unlabeled
ammonium sulfate. The whole plant samples were analyzed for total N and 15N.
Following harvest, soil samples were collected to a 4-foot depth from all plots
and analyzed for both inorganic and organic N and 15N.
The water samples collected by the automatic samplers were analyzed for nitrate
and ammonium N and total and soluble P.
The weather differed considerably between the cropping seasons. Temperatures (Figure 1) were cooler than normal and precipitation (Figure 2) was somewhat lower than normal in 1997. In contrast, the winter and spring of 1998 and 1999 were warmer than normal, and precipitation was above normal, especially in May and June of 1998. Temperatures were warmer than normal in the winter and early spring of 2000, but precipitation was below normal for the growing season.
Application of N resulted in substantial yield increase at all locations in each year (Figure 3). Grain yield response to applied N was greatest in 1998, least in 1999, and intermediate in the other two years. Optimum N needed per unit of land area and per unit of production was highest in 1998 and lowest in 1999; on the average, there was a 90 lb/acre and 0.44 lb/bushel difference between the two years. Differential in grain yield response to applied N was due in large part to the influence of climatic variability on soil N release and/or loss. The warm early spring combined with adequate soil moisture would have enhanced the potential for mineralization in both 1998 and 1999. The high unfertilized check yields in 1999 support the theory that mineralization was enhanced in 1999. The enhanced mineralization of 1998 was likely offset by denitrification and leaching that resulted from the excess rainfall that occurred in late May and early June, resulting in lower yield on the unfertilized plots and a higher relative response to applied fertilizer. It is theorized that the cooler soil temperatures in 1997 and the dry soils in 2000 resulted in lower mineralization rates and associated lower yield on the unfertilized treatments. Precipitation was slightly higher than normal in April and August of 1999. However, the potential for this to cause leaching and denitrification was relatively low, as there were no large precipitation events in April through June.
The amount of N needed to obtain the economic optimum yield per unit land area (Figure 4) and per unit of production (Figure 5) tended to decrease within any given year (1998 was an exception) as the historical excess N rate increased, indicating more residual N available from carryover of previous N application. However, if that were the case, such residual must be present in an organic form, as there was no relationship between previous N rate and residual inorganic N in the fall (Figure 6). This would agree with the results of Stevens et al. (1997) that showed that excess N is incorporated into organic forms that are mineralized about seven times faster than native organic forms of N. Lack of relationship between previous N and amount of N needed in 1998 may have been due to the excess precipitation and thus increased N loss in 1998.
Time of N application appeared to have little influence on the optimum N needed
per unit of production (Figure 7). On average, spring-applied
N needed 0.1 lb N/bushel less than either fall or sidedress N (Figure
8).
Fertilizer recovery was determined using the 15N technique at five of the locations. The soil analysis for this portion of the study is completed for 1997 through 1999 and the plant analysis data for 1997 and 1998. Based on those results, the following observations can be made:
Total fertilizer N recovery was higher the later the time of application (Figure 9). This was especially true for plant N uptake and for organic N recovery at the end of the growing season. Inorganic N (primarily nitrate) was higher in the fall of 1998 from N applied the previous fall than from N applied in the spring. Total recovery of N was low in both 1997 and 1998.
While N recovery from fertilizer tended to be slightly higher the later the time of application and the higher the rate of application, in all cases, the majority of the N in plants was soil derived (Figure 10).
Flow-weighted nitrate-N concentrations from tile lines tended to be highest on those fields planted to corn, and they tended to be greatest on those fields that had a history of higher N rates in the past (Figure 12 and Figure 14). The relationship between previous N rate and nitrate-N concentration tended to be strongest in the year following corn than in the corn year.
Total N loss per acre in tile line flow followed a pattern associated with previous N history in the soybean fields in 1998 and 2000 (Figure 13 and Figure 15). The excessive rainfall received in spring 1998 resulted in high N loss per acre from the tile lines, especially those draining the cornfields. There was little relationship between time of N application and either concentration or content of nitrate in tile line water.
Table 1. Characteristics of the experimental sites.
Figure 1.Air temperature for Champaign, IL
Figure 2. Precipitation for Champaign, IL
Figure 3. Corn yield response to applied N - all locations (smaller images all on one page)
Figure 3. Corn yield response to applied N, by location (larger images on individual pages):
| Location 1705 | Location 2722 | Location 2725 | Location 2704 | Location 2720 | Location 3717 | |
| Location 3715 | Location 1833 | Location 2822 | Location 2806 | Location 3827 |
Figure 4. Optimum N per unit of land area as affected by excess N application and year.
Figure 5. Optimum N per unit of production as affected by excess N application and year.
Figure 7. Effect of time of N application and excess N on optimum N per unit of production.
Figure 8. Optimum N per unit of production as affected by time of N application.
Figure 9. Recovery of applied N as affected by time and rate of N application.
Figure 10. Effect of time and rate of N application on uptake of soil and fertilizer N.
Figure 11. Effect of time and rate of N application on residual soil N to a 4-foot depth.
Figure 12. Effect of prior N rate on flow-weighted nitrate-N concentration of tile water - 1998.
Figure 13. Effect of prior N rate on flow-weighted nitrate-N concentration of tile water - 1999.
Figure 14. Effect of prior N rate on flow-weighted nitrate-N concentration of tile water - 2000.
Figure 15. Effect of prior N rate on nitrate loss from tile line - 1998.
Figure 16. Effect of prior N rate on nitrate loss from tile line - 1999.
Figure 17. Effect of prior N rate on nitrate loss from tile line - 2000.
1 R.G. Hoeft and E.D. Nafziger are Professors, Dept. of Crop Sciences; R.L. Mulvaney is Professor, Dept. of Natural Resources and Environmental Sciences; S.T. Mirek is Graduate Research Assistant, Dept. of Crop Sciences; and L. C. Gonzini and J.J. Warren are Senior Research Specialists, Dept. of Crop Sciences, University of IL.
Brown, H.B., R.G. Hoeft, and E.D. Nafziger. 1993. Evaluation of three N recommendation systems for corn yield and residual soil nitrate. In R.G. Hoeft (ed.) Proc. IL. Fert. Conf. pp 43-50.
Buzicky, G.C., G.W. Randall, R.D. Hauck, and A.C. Caldwell. 1983. Fertilizer N losses from a tile-drained Mollisol as influenced by rate and time of 15N depleted fertilizer application. p. 213. In Agronomy abstracts. ASA, Madison, WI.
Stevens, W.B., R.G. Hoeft, and R.L. Mulvaney. 1997. Effect of N fertilization on accumulation and release of readily-mineralizable organic N. In R.G. Hoeft (ed.) Proc. IL. Fert. Conf. pp 65-78.
