“WheatPlan” – A Method of Estimating Nitrogen Fertilizer Requirements for Dryland Winter Wheat

Chapter 6 – Fertility, No. 12, Fall 1987

Don Wysocki

Each season, wheat growers must determine the amount of nitrogen (N) they will apply to their crop. This decision involves careful consideration of several factors including the reserve soil nitrogen supply, available soil water, predicted growing season precipitation, yield potential and cost of nitrogen fertilizer, The goal that is commonly sought in applying fertilizer nitrogen is to raise the level of soil N just to the point that a crop will have sufficient nitrogen in relation to the amount of available water. Unfortunately the amount of fertilizer nitrogen required to achieve this goal is difficult to estimate.

Developing a Climate-Yield Nitrogen Model

To assist growers and others in making decisions about fertilizer nitrogen, several Oregon State University researchers have been working to develop a computer software package entitled ” WheatPlan.” The authors of this software are Floyd Bolton, associate professor and Michael Glenn, former graduate student, Department of Crop Science, and Stanley Miller, professor, and Mike Taylor, research assistant, Department of Agricultural and Resource Economics. The software has been released for testing on a limited basis to county agricultural extension agents in dryland wheat-growing areas of Oregon. The program is based on field research conducted by Bolton and others at the Sherman Experiment Station, long-term climatic data from the Sherman Station and information obtained from wheat growers in the area of Moro, OR. The software program is designed to assist in the determination of nitrogen fertilizer requirements for winter wheat in wheat-fallow rotations in the 10- to 15-inch precipitation zone of the Columbia Plateau of eastern Oregon.

The software permits nitrogen fertilizer strategies to be evaluated at spring of the fallow season, seeding and spring of the crop season (topdress). Inputs that users must supply to the program are: monthly ‘precipitation and pan evaporation (inches), previous fertilizer nitrogen applications (pounds/acre), date of seeding and emergence, prevailing price of wheat ($/bushel), prevailing interest rate, cost of nitrogen fertilizer ($/pound), application costs ($/acre)and a calculated productivity index.

Precipitation

Examination of the precipitation and crop yield records at the Sherman Experiment Station, Moro, OR, shows that seasonal precipitation fluctuates around the long-term average and that wheat yields in wheat-fallow rotation are dependent upon precipitation falling both during the fallow and growing season, Using this information Bolton and associates developed a system to indicate dry, normal and wet fallow seasons and dry, normal and wet growing seasons (Table 1). Because of the dependence of crop yield on both fallow and growing season precipitation, the probability of these patterns occurring in sequence is important. The- nine possible combinations of fallow-crop precipitation patterns along with their probabilities for occurrence are shown in Table 2. Bolton conducted a series of field trials to simulate the five fallow-crop precipitation patterns that have occurred most frequently in the past. His work showed that it is possible to use the historical precipitation an yield records and current precipitation to predict a potential crop yield and to estimate the amount of nitrogen required to produce that yield (Table 3).

Table 1. Definition of dry, normal and wet precipitation patterns for fallow and crop seasons at the Sherman experiment Station, Moro, OR (Bolton, OSU).

 Precipitation
(inches)
DryNormalWet
Fallow Season
(14 months Aug.-Sept.)
Less than 10.310.3-14.6More than 14.6
Crop Season
(10 months Oct.-July)
Less than 8.68.6-12.6More than 12.6

Table 2. Combinations of fallow and crop season precipitation patterns and their probability of occurrence, Sherman Experiment Station, Moro, OR (Bolton, OSU).

Precipitation pattern
fallow-crop
Cumulative Precipitation
(inches)
% probability of occurrence
Dry-dry15.2-18.93.1
Dry-normal16.7-22.813.9
Dry-wet20.7-26.14.6
Normal-dry17.3-23.116.9
Normal-normal18.9-27.233.8
Normal-wet22.9-30.49.2
Wet-dry21.6-25.83.1
Wet-normal23.2-29.513.9
Wet-wet27.2-32.71.5

Table 3. Optimal yields at the optimal N fertilizer level for the five meet common fallow-crop precipitation patterns, Wasco, OR (Bolton, OSU).

Pattern
fallow-crop
Yield
(bu/acre)
Optimum applied nitrogen
(lb/acre)
ActualPredicted Deviation
Dry-normal46.039.5+6.540
Normal-dry51.049.7+1.354
Normal-normal50.253.1-2.954
Normal-wet60.060.6-0.667
Wet-normal53.9*68.7-14.854
*A heavy frost in mid-October caused winter kill and loss of early growth

Soil Nitrogen

Some portion of the nitrogen required by a wheat crop will be supplied by mineralization of soil organic matter. The contribution from mineralization will vary on soil and environmental conditions but should not be overlooked. Results of field trials conducted by Bolton at the Sherman station showed that the amount of nitrogen mineralized during both the fallow and crop periods was strongly affected by the precipitation pattern (Table 4). Consequently, the precipitation pattern not only affects grain yields but also the amount of nitrogen made available through mineralization. Using inputs for precipitation the program estimates the amount of soil nitrogen that is released for plant growth during both the fallow and crop season.

Soil Water Storage

Although the amount of precipitation occurring during the fallow period is extremely important, only the water that is stored in the soil will be available for plant growth. The amount of fallow season precipitation stored in the soil is a function of many factors including: precipitation amount and intensity, temperature, wind, humidity, radiation and soil depth. A reasonably accurate measurement of stored soil water at seeding requires soil sampling to the rooting depth throughout a field. This is time consuming and expensive, and therefore not practical on a large scale. Bolton found that stored soil water at seeding can be estimated reasonably well using monthly precipitation and pan evaporation rates. Pan evaporation rates are accumulated on a daily basis during the frost-free period at all official weather stations in Oregon and elsewhere. The program using a mathematical model and inputs for both monthly fallow season precipitation and evaporation internally estimates stored soil water at seeding to predict potential grain yields.

Date of Seeding and Emergence

Both the date of seeding and emergence can strongly influence yields of winter wheat. Work by many researchers over the years has shown that late planting or delayed emergence will result in yield reductions. Input on both date of seeding and emergence are required by the program to adjust expected yields when planting is late and emergence is delayed or both. Bolton states that in general, delays in stand establishment by 30, 45 or 60 days can result in yield reduction as much as 5, 10 or 60 bushels/acre respectively.

Table 4. Influence of precipitation patterns on mineralization of nitrogen during the fallow and crop season (Bolton, OSU).

Precipitation
pattern
Nitrogen Mineralized
(lb/acre)
Fallow Season
Dry56.8
Normal44.0
Wet65.6
Crop Season
Dry-normal38.5
Normal-dry34.4
Normal-normal32.0
Normal-wet26.5
Wet-normal21.0

Productivity Index

Seldom under a given set of precipitation conditions can different growers be expected to produce the same yields. Factors such as soil type, soil depth, topography, management practices, weeds and diseases vary sufficiently from farm to farm or even from field to field, so that each location is unique. To account for these differences it was necessary to develop a method of adapting the program to the conditions of each user. This is achieved through use of the productivity index (PI). The PI is a comparison of the water use efficiency (WUE) of the user’s farm or field to that at the Sherman Experiment Station (location of the program’s data base) over the same time span. WUE is determined by dividing the total precipitation from harvest to harvest by the yield (Table 5). The WUE is a measure of a particular site’s ability to convert water to grain. Bolton suggests that a minimum of 5 but no more than 10 years of crop data be used to calculate WUEs. Time spans shorter than this may be unduly weighted by an unusual season and those longer than this may be outdated because of changes in varieties, production practices or technology. The PI is determined by dividing the user’s WUE by the WUE at the Sherman station for the same period of time (Table 5). The program uses the PI to fine-tune program output for the user’s conditions.

Economic Factors

Cost of application, interest rates, price of nitrogen and price of wheat are economic factors that influence the return from applications of fertilizer nitrogen. This flexibility permits the user to consider numerous scenarios for various combinations of prices, costs and yields. This is useful for developing fertilizer strategies as wheat prices rise and fall.

Table 5. Determination of water use efficiency and productivity index.

Harvest yearFallow+crop season precipitation
(inches)
Yield
(bu/acre)
WUE
Sherman Experiment Station
197821.7955.02.53
197923.7445.61.92
198025.7562.32.42
198129.2152.31.79
198226.2259.52.27
198331.0979.82.57
Average WUE=2.08
Hypothetical Farm
197821.5841.01.90
197925.4934.01.33
198027.3652.01.90
198131.4180.02.55
198229.3954.01.84
198332.4262.01.91
Average WUE=1.90
Productivity Index
PI=WUE hypothetical farm/WUE Sherman Station=1.90/2.08=0.91

Summary

One of the most difficult management decisions that wheat growers make each season is the amount of nitrogen to apply. Because of the dynamic nature of N in the soil and the unpredictability of rainfall prescribing the optimum level of N is difficult. To assist wheat growers in the 10- to 15-inch precipitation zone of eastern Oregon in making N fertility decisions, researchers at Oregon State University have developed a computer software package. The program requires simple inputs that producers can easily obtain from their own records and local weather stations. It requires an IBM PC or compatible, DOS 2.x and 256K RAM. The software has been released on a limited basis for testing and refinement by county agricultural extension agents and should be ready for general release in the near future.