News & Events - Archived News

Fertilizer application is an integral part of growing crops, and determining how much to use has been the subject of much research over the years. Although its still not an exact science, many useful guidelines and recommendations have become generally accepted.

Many fields, however, are variable, so the common practice of taking soil samples from different areas of a field then mixing them all together doesnt adequately describe the fertility needs for specific areas of the field it just gives an average.

Until the advent of the Global Positioning System (GPS), there was no practical way to address the problem of varying the amount of crop nutrients applied at different areas in a field.

By using the GPS technology and related computer hardware, the exact location of each soil sample can be determined. Instead of mixing all the samples from an entire field to get an average, each sample can be individually analyzed and maintained as a unique data point. A specially equipped fertilizer applicator can then be used to vary the amount of nutrients applied to different parts of the field. This practice is referred to as variable rate application or site specific farm management.

In 1997, Dr. Gerry Anderson began a Precision Agriculture study at the USDA-ARS Northern Agricultural Research Laboratory in Sidney to examine the benefits of variable rate application on irrigated crops. The goal was determining if fertilizer costs could be reduced while maintaining or increasing yields in other words, increasing production while reducing production costs. In addition to increasing production while reducing costs, it is generally assumed that more precise use of fertilizers will also be good for the environment.

In the fall of 1997, 69 soil samples were taken from 58 acres, and variable rate fertilizer maps were created for nitrogen, phosphorus and potassium. A custom operator with a specially equipped fertilizer applicator was hired to apply the fertilizer. The field was divided into eight strips about 240 feet wide. Half the strips had a uniform rate of fertilizer applied, while the other received variable rate applications determined by soil samples.

In 1998, five more fields were added to the study and the same procedure was used. The soil samples were on an approximate one-acre grid with 274 sample locations in 300 acres. At each sample location, two 4-foot (where possible) cores were pulled. The cores were cut at 12 inches and 24 inches to divide them into three sections. The top foot from both cores was then mixed together to form one sample. The sections of cores from the second foot were also mixed together to form another sample, and the bottom two feet of the cores were mixed to form a third sample. These samples were then analyzed separately.

In 1998, the study focused on wheat. In 1999 four fields were planted to sugar beets and two in corn. In 2000, the rotation was reversed with four fields in corn and two in sugar beets. Yields were determined by taking hand samples of the sugar beets and wheat at each of the sample locations so sugar and protein could be measured. Yields were also weighed from most fields on a strip-by-strip basis. Corn yield was determined from combine yield monitor data and load slips.

In 1998, the year with only 58 acres in the study, wheat didnt show a significant difference in bushels per acre or protein between the two treatment types. In 1999, the corn test was damaged by high winds on Halloween, and no conclusive data could be formulated because about 30 bushels per acre were knocked to the ground.

One field was about 25 percent harvested when the wind hit. The other field that was harvested after the wind showed about a 10-bushel yield advantage in the variable rate strips compared to the check. The corn in 2000 showed a 4.44-bushel advantage to the variable rate application, which was not statistically significant.

Sugar beets showed the greatest benefit from the variable rate fertilizer application. If sugar beets are over-fertilized with nitrogen, sugar content can decrease while impurities increase. In other crops, yield typically levels off as N rates increase above optimum levels the fertilizer may be wasted, but the crop will not usually suffer although lodging may occur at very high levels.

The results from the two years of sugar beet trials seemed to bear this out. The average yield of the variable rate test strips was 2 tons/acre higher than the yield of the control strips. The increase ranged from a low of .83 tons to a high of 3.58 tons per acre.

While the percent sugar was nearly the same (0.04% difference) on both treatments, the smallest tonnage increase was accompanied by a decrease in sugar percent. The smallest field out of the six showed a loss of about 80 pounds of sugar per acre, which would, of course, translate into a loss of revenue. This was in contrast to the average increase in pounds of sugar per acre of 796.

Fertilizer costs were decreased by about 10 percent. A larger decrease was expected, but after sampling fields it was found to be a common scenario where the bulked soil sample that represented the entire field to determine the uniform fertilizer application rates would show the need for little or no added potassium.

However, the grid sample would reveal an area that was considerably below the recommended levels and so would require a healthy application of potassium to meet the guidelines set by the formulas governing the variable rate application.

The precision ag team at the Sidney lab believes these results show promise. The barriers are the high cost of application equipment and related software, the extra labor and soil sampling costs and probably most important the amount of time needed to make it all work. The precision ag team is currently evaluating strategies to make the practice more cost effective.

This spring, an analytical spectral device will be used to map fields and hopefully identify zones that can be used to reduce the number of soil samples needed to adequately characterize a field. Another possible solution is using remote in late August to determine the nitrogen content in sugarbeet leaves (which contain a considerable amount of nitrogen that is returned to the soil and used by the next years crop).

By determining the spatial variability and relative quantity of nitrogen in the beet leaves, perhaps we can make an inexpensive but useful fertilizer application map for the crop following beets.

Another idea is being investigated to reduce the problem of delaying fall tillage while waiting for soil test results. Because areas of the field that test low for phosphorous and potassium one year are likely to test low the year after as well, a reasonably accurate spread map could be made using soil test results from the previous year.

About 65 percent of the expected nitrogen required would also be applied with the phosphorous and potassium right after harvest. After the soil samples were analyzed, the producer would have all winter to generate the spread map for the remaining nitrogen required. This would still let the producer benefit from the ability to variably apply nitrogen but would not delay fall tillage work for the week or so it takes to get the sample results.