Holzapfel: Evaluating In-Season Yield Potential and Nitrogen Fertilizer Requirements in Canola using the GreenSeeker Sensor
Date: 2007
Term: 3 years
Status: Completed
Researcher(s): Chris Holzapfel, Indian Head Agricultural Research Foundation and Judy McKell, Indian Head Agricultural Research Foundation
SaskCanola Investment: n/a
Total Project Cost: n/a
Funding Partners: n/a
Project Summary
Sensor-based N management appears to have potential for enhancing agronomic N-use efficiency in canola production and, provided that the risks and benefits of sensor-based N management are managed appropriately, economic profitability for canola producers. Optical sensors such as the GreenSeeker™, have the potential to be used as tools to help producers decide whether or not topdressing N is warranted and, if so, how much additional N is required to optimize yields.
The greatest challenge in nitrogen (N) fertilizer management in Saskatchewan is determining rates that are appropriate for the specific soil and growing conditions at or before seeding when most N fertilizer is applied. The difficulty in choosing optimal N rates is that yield potential, soil N availability and N losses all tend to be difficult to predict and variable across the landscape as well as from one year to the next. Applying fertilizer N rates that are higher than the minimum rate required for maximum yield results in reduced economic profits and N-use efficiency, along with increased potential for the N to be lost from the soil-crop system with negative environmental consequences.
In regions where moisture availability frequently becomes limiting during the growing season, splitting N fertilizer applications between seeding time and in-crop topdressing applications can be used to more accurately match the total amount of N applied with crop demands. Optical sensors, such as the GreenSeeker™ sensor, have potential as tools to help producers better match N inputs with crop demands, resulting in enhanced nitrogen use efficiency and economic returns. These optical sensors measure the crop’s normalized difference vegetation index (NDVI), which is an indirect measure of biomass, N-uptake, and yield potential.
The overall objectives of this study were: 1) to investigate the potential for estimating canola yield potential using canola NDVI measurements early enough in the growing season to still achieve a yield response to topdressing N, and 2) to evaluate the feasibility of using optical sensor measurements and high N reference crops to determine topdress N rates relative to banding the entire N requirements at the time of seeding.
In developing the empirical NDVI-canola yield equations, researchers used data from small plot experiments with a range of N rates and seeding rates, at Indian Head, Swift Current, Brandon and Ottawa. The study results show that it is possible to estimate canola yield potential over a wide-range of environments and plant populations using in-season
NDVI measurements. Similar to corn, NDVI for canola increased with time as the crop progressed through the vegetative growth stages, peaked during the mid-reproductive stages and the start of flowering and decreased to a certain extent as the crop matured.
Figure 1. Relationship between canola seed yield and NDVI divided bygrowing degree days (base temperature 58C, GDD5) accumulated between seeding and sensing for Brandon, MB, Indian Head, SK, and Ottawa, ON (thin-Black - Black soil zones) and sensing dates where the canola was between growth stages 2.5 and 4.2 (Harper and Berkenkamp 1975). Dashed lines are the upper and lower 95% confidence limits (n = 1678).
Source: Can. J. Plant. Sci. 89:1149-1160.
The second objective was to examine the feasibility of using optical sensors and reference strips where N is not limiting to determine N topdressing requirements relative to the conventional practice of banding the crop’s entire estimated N requirements at seeding. Sensor-based N management or variable rate N (VRN) was evaluated at Indian Head and Scott in 2005-07 along with several other N management treatments including the predominant practice of banding the crop’s entire N requirements beneath the soil surface at seeding.
Overall the study results show that sensor-based N management performed well compared to the conventional practice of banding the entire estimated N requirements at seeding. On average, 15-53% less N was applied to the VRN treatments relative to banding predetermined amounts of N at seeding and overall there were no significant differences in grain yield observed between the two treatments. However, it is important to consider soil moisture conditions at the time of the N topdressing application along with optical sensor measurements when deciding whether or not to topdress N.
The study results show that sensor-based N management appears to be a feasible option for canola production in western Canada and has potential to increase agronomic N-use efficiency over the long-term, especially in the Black soil zone. Optical sensors such as the GreenSeeker™, have the potential to be used as tools to help producers decide whether or not topdressing N is warranted and, if so, how much additional N is required to optimize yields. However, for the practice to be economically viable, the value of the yield gains and/or N fertilizer savings must be sufficiently large to cover the added cost of the extra field operation.
Scientific Publications
Holzapfel, C. B., Lafond, G. P., Brandt, S. A., Bullock, P. R., Irvine, R. B., James, D. C., Morrison, M. J. and May., W. E. 2009. Optical sensors have potential for determining nitrogen fertilizer topdressing requirements of canola in Saskatchewan. Can. J. Plant. Sci. 89: 411 – 425.
Holzapfel, C. B., Lafond, G. P., Brandt, S. A., Bullock, P. R., Irvine, R. B., Morrison, M. J., May, W. E and James, D. C. 2009. Estimating canola (Brassica napus L.) yield potential using an active optical sensor. Can. J. Plant. Sci. 89: 1149_1160.