Nicolas Tremblay’s team uses complete weather stations to record seasonal characteristics, including rainfall amount and distribution over time, which are used in the equations to predict nitrogen rates. Photo by Edith Fallon, AAFC.
Finding that “just right” rate between too much nitrogen (N) fertilizer and too little isn’t simple. Fortunately a new tool is being developed to improve N rate recommendations for corn. Tests so far show using the tool could help growers save money by reducing their fertilizer inputs and achieve higher yields by adding more N when it is truly needed.
Many changing factors influence a crop’s N needs. “In most fields, about half the nitrogen the crop takes up comes from the soil. Soil organic matter breaks down and releases nitrogen that gets taken up by the crop and then recycled back to the soil. Organic nitrogen breakdown varies with the weather, soil conditions, and the interaction between the soil conditions and the weather. So the supply changes from year to year and place to place,” Keith Reid, a soil scientist with Agriculture and Agri-Food Canada (AAFC),” explains.
“In addition, the amount of nitrogen the crop needs varies from year to year and place to place because a crop that is growing well needs more nitrogen than a crop that isn’t doing as well.”
Also, both the N released by soil organic matter and the N applied as fertilizer can be lost to the surrounding environment through leaching and gaseous losses, and those losses vary depending on such factors as the weather and soil characteristics.
“Further complicating the picture is some of the variation happens before we would normally apply nitrogen fertilizer and some happens after. So nitrogen rate predictions are never going to be 100 per cent accurate because we can’t account for the strange things the weather might do later in the growing season,” Reid says. “So how do we integrate all of those things in that intricate dance, and zero in on what’s the right amount of nitrogen to apply?”
Reid and Nicolas Tremblay, a research scientist with AAFC, are working on answering that question through their project to develop a N fertilizer rate prediction tool. The initial version of the tool will provide N recommendations for corn. Down the road, the plan is to add other crops like wheat, potatoes and canola to the tool.
“The project’s overarching goals are to help farmers improve their bottom line by not applying more nitrogen than the crop needs and also not shorting the crop by applying too little nitrogen, and at the same time to reduce the environmental impacts of applying too much nitrogen,” Reid notes.
“Using the tool should remove a lot of the uncertainty around what is the right amount of nitrogen to apply. That would mean growers would be less likely to apply ‘insurance’ rates of nitrogen. We know it is a pretty direct loss off your bottom line if you leave 20 or 30 bushels of yield in the field that you could have had if you had put on a little extra nitrogen. That is a powerful incentive to make sure you don’t short the crop,” he adds. “At the same, if growers routinely put on an extra 20 or 30 pounds of nitrogen ‘in case,’ that is an economic drain because it’s an expense they didn’t need to make. Over enough acres and enough years, that adds up.”
Reid explains if the crop didn’t need those extra 20 or 30 pounds, it represents about a pound-for-pound loss of N to the environment. Nitrate leached into the groundwater is a concern for drinking water quality. There are also concerns nitrates in water harm some types of amphibians and other aquatic organisms. And N can be lost to the air as nitrous oxide, a greenhouse gas.
“Agriculture is one of the sources of those nitrogen forms that get into the environment, and it is part of agriculture’s responsibility to try to minimize that,” Reid says.
Developing the scientific underpinnings
Tremblay’s recent research studies have laid the scientific foundations for the tool. He notes, “We have conducted a suite of projects over about the last decade to sort out the critical factors involved in the selection of optimal nitrogen rates to be applied to a crop.” Through these studies, Tremblay and his research group have identified key factors including soil texture and rainfall, as well as other factors that also play a role. And they have quantified the complex, interacting effects of these factors on corn’s response to N.
Reid explains the research work involves meta-analysis of data from many N response trials. “Instead of looking at the average results from the trials, we’re looking at the variations in the results to determine what is causing some sites to need more nitrogen and some sites to need less, to see how that is related to soil type, weather conditions, crop management, previous crop and so on. Along with that meta-analysis, we’re using some advanced soil and crop growth models to extend the information from the trial locations into other environments.”
He adds, “Nicolas has done some really good work on developing an index of abundant and well-distributed rainfall – did you receive enough rainfall, not too much or too little, and how evenly was it distributed over the given time period? That parameter does a good job of encapsulating those two rainfall aspects into a single factor.” This factor has helped to improve the accuracy of the equations to predict N needs.
As a result of all these studies, the researchers have been able to put together the most important decision rules into a model that could be used as a decision-support system to help growers select the N rate for their corn crop.
Tremblay’s research team tested the system in 2013 and 2014 at 18 fields across the corn-growing region of Quebec. “Our goal was to compare our model’s recommendations with what the grower was applying in his normal management,” Tremblay notes.
The results from these tests have been quite positive. “In 2013, on average, we had the same yield level as the grower but with 15 per cent less nitrogen applied,” he says. “And in 2014, we had an average profit of $42 per hectare. This profit was due either to a 24 per cent reduction in nitrogen fertilizer for the same yield as the grower’s, or to a five per cent yield increase with a 29-kilogram nitrogen supplement when [higher nitrogen rates were] required by our system.”
The researchers are now in the process of putting the decision-support system into a web-accessible tool for growers.
“The grower will be able to locate the field he wants to fertilize on a map on the Internet. Then, based on the GPS location defined, there will be information in the background on soil and rainfall that will be considered by the software to calculate the optimal nitrogen requirement of that particular field at that particular time,” Tremblay explains.
The grower will also need to input the field’s soil texture, along with other information such as the field’s previous crop, soil organic matter level and tillage system. The tool will use data from Environment Canada to determine rainfall the field recently received and the forecasted rainfall for the area.
Reid thinks the tool will likely provide more accurate predictions if it is used to predict fertilizer rates for sidedressing or later applications, rather than applications around planting time. “The later the fertilizer application is delayed, the more of the weather variability we can take into account. But I hope the model has some value for preplant applications because a lot of growers [apply all of their N as a preplant application].”
Reid is working on a way for the tool to provide recommendations that address in-field variability in N requirements, perhaps through the use of optical sensors, like GreenSeeker technology. “A lot of studies are being done using optical sensors to sense the variability across a field and to change nitrogen rates on the go, but those studies really haven’t linked with soil-based indicators or weather-based predictions,” he notes. “I’m looking at how we can actually integrate the information from the optical sensors in the field with what we know about the soils and the weather conditions, so we can go to that next level of precision in being able to account for both spatial and temporal variability.”
Over the coming months, the researchers will continue to fine-tune the web platform for the tool and to further strengthen the model’s accuracy with new statistical analysis. AAFC is also working on an agreement with Environment Canada so the web-based tool can access the necessary weather data.
Because the federal government doesn’t deliver tools directly to farmers, AAFC will be collaborating with other agencies to deliver the tool. Tremblay expects AAFC will eventually set up an agreement with a private company to make the tool available to users and to maintain and improve the tool in the years ahead. Reid thinks provincial governments might also be interested in the tool; for example, Ontario might want to incorporate some of the tool’s equations into the Ontario Corn Nitrogen Calculator.
The tool is expected to be ready for corn growers in 2016.