Features Agronomy Cereals
Real-time variable rate nitrogen application inches closer

Ground based technologies, such as GreenSeeker, that use optical sensors to measure variability in crop growth may be an important breakthrough in precision farming. The technologies operate in real-time, eliminating the need for extensive data collection and interpretation to develop management zones and prescription fertilizer maps.

“From what we’ve seen so far, there are several opportunities worth investigating. In moist areas, you could apply all your nitrogen fertilizer for average yield at time of seeding, and then top up with additional post-emergence N to cover spatial variability and to improve yield,” says Chris Holzapfel, who is working on developing the technology for the Canadian prairies at the Indian Head Agricultural Research Foundation (IHARF) in Saskatchewan. “In drier areas, it might be used as a risk management tool, where a lower amount of N is applied at seeding and then post-emergence N applied if growing conditions are good.”

GreenSeeker technology uses active optical sensors that emit visible and near infrared light onto the crop canopy, and measure the canopy’s reflectance of emitted light. The reflectance values are then used to calculate the Normalized Difference Vegetation Index (NDVI), which is an indirect measurement of the crop’s above ground growth. By comparing the NDVI of the crop being evaluated to that of an N-rich strip in the field, the technology can be used to address field variability without developing management zones, as was previously done with GPS and yield mapping.

GreenSeeker technology was developed at Oklahoma State University. Generally, six sensors are mounted across the front of the boom on a high clearance or dribble band sprayer. As the applicator moves across the field, a built-in microprocessor analyzes the NDVI readings and determines the N requirements that are needed to meet full yield potential. Pre-determined algorithms calculate the amount of N required. The information is relayed to the rate controller to provide variable rate N application in real-time as the applicator moves across the field.

Chris Holzapfel is developing the algorithms for GreenSeeker technology. Photo By Bruce Barker.

In theory, a farmer would seed one strip of his field with a very high N rate to simulate a N-rich environment where the crop’s N requirements are being satisfied. He would drive the GreenSeeker sprayer over the N-rich strip to set the upper NDVI limit. Any NDVI values recorded on the rest of the field that exceed that value would receive no additional N. NDVI values that are less on the field would receive additional N as a surface dribble band of UAN based on its yield potential relative to the yield potential of the N-rich strip.

Ground truth research necessary for the prairies
Although the technology has been available for several years in the US, research has to be conducted on the prairies on each crop to develop the algorithms to calibrate the application rates. These algorithms are mathematical calculations that correlate the NDVI readings with crop response to added N under varying fertility levels.

At IHARF, Holzapfel is working with Guy Lafond of Agriculture and Agri-Food Canada to develop the algorithms necessary to implement GreenSeeker technology on the prairies. They are furthest along on canola and wheat, and just getting started on winter wheat, barley, oats and durum wheat.

In the early stages of the work, Holzapfel established an empirical equation that allowed him to estimate yield potential during the growing season. He used small plots seeded with various N fertilizer and seeding rates to establish plots with different yield potentials. The NDVI of each plot was measured and plotted against grain yield. To refine the data to take into account the normal increase in vegetation as the crop grows, Holzapfel divided the yield and NDVI relationship by the number of growing degree days accumulated between planting and sensing. These relationships provide the basis for the N application algorithms.

The next step was to put the algorithm into practice in small plot research. The plots were established in 2005 and carried on through 2007 in the case of canola and spring wheat. Other crop plot research was established more recently. The treatments included four fertilizer applications consisting of a typical farmer practice, a N-rich treatment used to calibrate each field, a split application with a fixed rate of post-emergence N, a split application with a variable rate of N using the GreenSeeker technology, and a reduced rate of N at seeding without further post-emergence N. The N-rich treatment was generally at 133 to 150 percent of recommended rates to ensure that N was not the limiting factor in crop growth.

At several sites, fertilizer N inputs were reduced with the GreenSeeker technology without affecting grain yield. “Overall, the results of these experiments to-date show that the GreenSeeker sensor can be a useful tool for evaluating N status in canola and fine-tuning post-emergence N application rates,” explains Holzapfel.

Field trials show potential
Field scale evaluations on the algorithms and GreenSeeker technology were also conducted at Indian Head. Six boom-mounted GreenSeeker sensors and N-rich reference strips were used to assess post-emergence, variable rate applications of UAN on canola and cereal yields and fertilizer use, compared with applying all fertilizer N at seeding or split applications using fixed rates.

Similar to the plot studies, the variable rate GreenSeeker technology usually reduced fertilizer application rates without compromising grain yields. However, because of the added cost of the post-emergence N application, the variable rate application was rarely more profitable than the fixed rate applied at seeding. “It’s hard to beat one pass seed and fertilizer application in drier areas,” explains Holzapfel. “There’s still a lot to sort out on what is the best application for this technology.”

Holzapfel says they are reassessing their approach to post-emergence N because the more profitable sites were those where variable rate N increased yield rather than saved on fertilizer cost. He says this might indicate that rather than using the technology to reduce
fertilizer rates, it might be more useful to use it to increase yield potential in moister growing areas or in years where there is good potential to substantially increase yields.

“We propose that this technology has the potential to be a useful tool for estimating the potential increase in grain yields that producers may achieve with post-emergence N, and allow them to make a decision as to whether doing so would be profitable,” summarizes Holzapfel. “If the increase in grain yield that can be achieved is not sufficient to cover the cost of the added N and the extra pass over the field, no further action would be recommended, while if the increase is sufficient, it may be profitable for the producer to apply post-emergence N.” -end-