Illinois Fertilizer Conference Proceedings January 22-24, 2001 Home 2001 Index Search

Using historically esablished yield variability for the improved prediction of nitrogen fertilizer needs on corn

E.C. Varsa, S.A. Ebelhar, T.D. Wyciskalla, and C.D. Hart¹

Introduction

Increasing numbers of growers have accumulated several years of crop yield data, using yield monitors, in fields where historical patterns of above-average and below-average yields have emerged. That is, high-yielding and low-yielding areas tend to assume somewhat similar patterns year after year. In order to optimize nitrogen (N) fertilization of fields so as to reduce inefficient use of the applied N, it seems plausible to fertilize those areas of fields with high established yields with rates of N that are greater than those areas of fields with lower established yields (Carr et al., 1991; Redulla et al., 1996; Sawyer, 1994). The current practice for N fertilization for a field is to select a rate based upon the past five-year average for the entire field. This approach seemingly overfertilizes low-productivity areas (and enhances N leaching and runoff), while higher-productivity areas are likely underfertilized.

With the advent of variable rate technology (VRT) and recent advances in fertilizer equipment, application rates can be tailored and varied as one traverses the field. It is the objective of this research to determine if agronomic, economic, and environmental benefits can be obtained by varying N application rates across the field as low-, medium-, and high-productivity areas are encountered. This would be compared with the standard practice of uniform N application based upon the average yield for the whole field.

The research contained in this report is described in two parts. First, a whole-field comparison of variable N application will be made with a fixed (uniform) rate of N application. A second phase of this research will evaluate response of corn to rates of N within selected historical low-, medium-, and high-productivity regions of the field.

Materials and Methods

Mr. Kelly Robertson, a farmer and crop consultant located in Franklin County, Illinois with at least five years of yield monitor data on fields of his farm, has agreed to collaborate with us on this research. For the 2000 cropping season, he had a 30-acre field available for use in these studies that was in soybean in 1999 and which has been in a corn-soybean rotation for 10 years. The dominant soil type in the field was a Cisne silt loam with lesser amounts of Hoyleton silt loam. Topography is quite flat, with a slope averaging from zero to two percent. Adequate drainage is a severe limitation in this field, and surface ditching is practiced where possible to remove excess water. Tile drainage is not practiced because of the restrictive claypan layer in the Cisne soil.

A map giving the outline of the field and the normalized yields broken down into low, medium, and high categories is given in Figure 1. The low-yield regions were identified as those with normalized yields that were 90 percent or less; medium (or average yield) regions of the field had normalized yields that ranged from 90 to 110 percent; and high regions had normalized yields that exceeded 110 percent. The areas of the field that were in the low-, medium-, and high-yield categories were approximately 12, 49, and 39 percent, respectively. The average corn yield for this field (excluding drought years) over the past 10 years has been about 160 bushels per acre.

Twenty-four points for sampling were GPS-identified in the gridded map of the field (each grid being 60 ft. by 60 ft.). Of the 24 sampling points, eight points each were located in the low-, medium-, and high-productivity regions (Figure 1). Duplicate samples of the surface soil (0–8 inch depth) taken at the sampling points revealed that the average pH was 6.3 (range 5.4–6.8), available (Bray P1) P was 36 lb per acre (range 17–93), and the exchangeable K was 302 lb per acre (range 183–556). These geo-referenced sampling points were also the locations for other crop and soil evaluations, including pre-sidedress N testing (PSNT), SPAD readings and ear leaf N sampling at silking, stalk nitrate at maturity, grain yield and moisture, and post-harvest soil nitrate analysis to a 3-foot depth.

Strip comparisons of variable rate N and uniform N application were based upon a field average yield of 160 bushels per acre. The uniform N application rate was calculated by multiplying yield x 1.2 less a soybean credit of 40 lb N per acre [(160 x 1.2) - 40 = 152 lb N per acre]. For the purposes of this experiment, we rounded the uniform N rate to 150 lb N per acre. Assumed yields for the variable rate N applications for low-, medium-, and high-productivity zones were 135, 160, and 185 bushels per acre, respectively. Rates of N applied to corn in these yield zones were 120, 150, and 180 lb N per acre (assuming a 40 lb N per acre soybean credit in each zone). Both uniform and variable rate strips were 30 feet wide (12 rows) for the entire length of the field, excluding head lands. Variable N application was accomplished by either increasing or decreasing applicator speed relative to the medium N rate (150 lb N per acre) to achieve the 120 and 180 lb N per acre rate for the low- and high-productivity portions of the field. Anhydrous ammonia was the N source used, and it was applied with a tool bar equipped with shanks spaced between each row. The anhydrous ammonia was applied as a sidedressing to the corn at the five-leaf stage of development.

In a selected portion of the field, where normalized yields of low, medium, and high productivity were closely contiguous to each other, a small-plot, intensive N rate study was conducted within each productivity region. N rate treatments selected were equivalent to 0.8, 1.0, 1.2, 1.4, and 1.6 lb N per bushel expected yield, plus a zero-N check. Additionally, N-Serve was included with the applied N for the 0.8, 1.0, and 1.2 lb N per bushel application rate treatments. A summary of those treatments follows:

All nitrogen treatments were replicated three times within a randomized complete block arrangement in each of the productivity zones. Individual plot sizes were 10 feet (4 rows) wide by 35 feet long. The nitrogen source was 28 percent UAN solution knifed in with a dual-shank applicator between rows 1–2 and 3–4. Application of N was made at the five-leaf stage of development. Measurements taken included stand counts, ear leaf N tissue at silking, SPAD readings at silking, grain yield and moisture, and stalk nitrate at maturity. Additionally, soil cores to a depth of 3 feet were taken to assess for residual soil nitrates. Table 1 shows additional experimental details concerning dates, cropping information, and precipitation.

Results and Discussion

Whole Field Uniform N versus Variable N Results

As shown in Table 1, rainfall received during the 2000 growing season was nearly twice the normal amount expected during the crucial June, July, and August months. No moisture stress was ever observed, and frequently excessive moisture and flooding caused injury to the corn and likely contributed to higher than normal nitrogen losses. An overall yield map of the entire field (Figure 2) showed that the vast majority of the field had a yield in excess of 182 bushels per acre. Comparing the long-term normalized crop yield map (Figure 1) with the 2000 yield map (Figure 2) revealed large areas in the medium productivity category that had exceptionally good yields in 2000. Areas that had a history of poor yields continued to yield poorly in 2000. Mostly, the poor yields were due to drowned-out corn, loss of N, and severe weed competition.
Soil test data and crop response measurements that were taken about the 24 sampling points (shown in Figure 1 and Figure 2) are given in Table 2. Of note was the lower phosphorus (P) soil tests observed as normalized productivity of the soil increased. This was probably a result of repeated, uniform fertilizer P applications over the whole field, but higher crop removal of P through higher yields occurred in the higher-productivity regions of the field. Other soil tests for pH and potassium were rather similar in the three productivity regions. SPAD meter, ear leaf N, and yield were all greater, as expected, as the productivity level of the soil increased. Leaf tissue N composition was well above the critical level (2.75 percent N) where N may cause a deficiency, even for low-productivity sampling points. A yield of 219 bushels per acre was obtained as an average for the 24 sampling points.

A comparison of uniform N versus variable N application revealed mixed results in measurements taken about the 24 sampling points (Table 3.) Generally, uniform N application had higher yields than variable N for the low- and medium-productivity regions, but variable N application had higher yields than uniform N application in the high-productivity regions. The data is somewhat anomalous for the medium sampling points since 150 lb N per acre was the N rate for both variable and uniform N. Higher yields might be expected at the low sampling points because more N was applied at those points with uniform N than variable N. However, at the high-productivity points, more N was applied variably than uniform, which may explain the higher yield observed with variably applied N. From combine yield monitor data, obtained on all strips comparing variable and uniform N application in the field, a yield advantage of 3 bushels per acre was found in favor of variably applied N. The issue of residual nitrates in the soil from the two methods of application have yet to be resolved, since the soil core samples for nitrate have not been analyzed.

Intensive N Rate Studies on Small Plots within Productivity Zones

The response of corn grain yield to increasing N rates within low, medium, and high normalized yield productivity zones is given in Figure 3, Figure 4, and Figure 5, respectively. For the high- and medium-productivity zones, a calculated optimum N rate was very close to the 1.2 lb N per bushel rate of nitrogen. With N-Serve inclusion with the N fertilizer, the optimum N rate was somewhat lower (151 versus 187 lb N per acre) for high-productivity soils. A realistic optimum N rate could not be calculated or predicted for corn grain in low-productivity soils. Extreme variability of the corn stand and vigor within and between plots contribute to the yield variability observed. Over all N rates, N-Serve inclusion with the fertilizer resulted in a higher corn yield in the low-productivity sites.

Tables and Figures

Table 1. Cropping details, measurements, and seasonal rainfall at the Franklin County, Illinois study site in 2000.

Table 2. Soil test and crop response measurements at the 24 sampling points, Franklin County, IL, 2000.

Table 3. A comparison of corn yield with variable and uniform N application across productivity regions of a 30-acre field in Franklin County, IL, 2000. (Data are from 24 sampling point locations).

Figure 1. Normalized crop yield map for variable versus uniform N application, with sample collection points, Franklin Co., IL, 2000.

Figure 2. Corn grain yield map as obtained from the combine yield monitor, Franklin Co., IL, 2000.

Figure 3. Effect of nitrogen rate on corn grain yield, Benton, IL, 2000 (yield level = high).

Figure 4. Effect of nitrogen rate on corn grain yield, Benton, IL, 2000 (yield level = medium).

Figure 5. Effect of nitrogen rate on corn grain yield, Benton, IL, 2000 (yield level = low).

Footnotes and References

¹E.C. Varsa is Associate Professor and T.D. Wyciskalla is Researcher, Dept. of Plant, Soil, and General Agriculture, Southern Illinois University, Carbondale, IL; S.A. Ebelhar is Agronomist; and C.D. Hart is Visiting Research Specialist, Dept. of Crop Sciences, University of Illinois, Dixon Springs Agricultural Center, Simpson, IL.

Carr, P.M., G.R. Carlson, J.S. Jocobsen, G.A. Nielsen, and E.O. Scogley. 1991. Farming soils, not fields: A strategy for increasing fertilizer profitability. J. Prod. Agric. 4: 57-61.

Redulla, C.A., J.L. Havlin, G.L. Kluitenberg, N. Zhang, and M.D. Shrock. 1996. Variable nitrogen management for improving groundwater quality. pp. 1101-1110. In P.C. Robert et al. (ed.), Proceedings of the Third International Conference on Precision Agriculture. Minneapolis, MN. June 23-26, 1996. ASA, CSSA, and SSSA, Madison, WI.

Sawyer, J.E. 1994. Concepts of variable rate technology with considerations for fertilizer application. J. Prod. Agric. 7: 195-206.

Back to 2001 Index

Department of Crop Sciences University of Illinois Extension College of Agricultural, Consumer, and Environmental Sciences University of Illinois at Urbana-Champaign Illinois Fertilizer & Chemical Association Illinois Department of Agriculture

Design by: Crop Sciences Computer and Web Support Group Copyright © 2005 University of Illinois at Urbana-Champaign Email site problems to the webmaster