A simulation of variable rate nitrogen application in winter wheat with soil and sensor information - An economic feasibility study

• Published in: Agricultural systems Vol. 192; p. 103147
• Main Authors: Pedersen, Michael Friis, Gyldengren, Jacob Glerup, Pedersen, Søren Marcus, Diamantopoulos, Efstathios, Gislum, René, Styczen, Merete Elisabeth
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01-08-2021
• Subjects: Canopy sensors; Crop modelling; Differential gross margin; Precision agriculture; Soil profiles; Variable rate N application; Soil profiles; Precision agriculture; Canopy sensors; Crop modelling; Variable rate N application; Differential gross margin
• ISSN: 0308-521X; 1873-2267
• DOI: 10.1016/j.agsy.2021.103147

Variable rate nitrogen (N) management strategies are often based on information about soil texture or information from canopy sensors, mounted on ground-based vehicles or satellites. However, disentangling the effect of each information type on N management strategy with experimental studies is often difficult, as results are only valid for the specific experimental conditions as well as the weather conditions for specific years. An alternative to this is to use deterministic crop growth models to generate a wider range of weather x treatment combinations.
This study examines if ‘static’ soil profile information or ‘dynamic’ canopy sensor type information provide a better basis for decision making concerning N-application at the subfield level.
The DAISY model was used to simulate crop growth in a five-year crop rotation on six soil profiles found in a heterogeneous sandy loam field. A range of management descriptions and simulations were made using 5 × 500 years of synthetic weather data with each crop in a five-year rotation set at the first year of the five parallel simulations. Simulated growth variables were used as proxies for a ‘dynamic’ canopy sensor information system. The differential gross margin was then calculated for a range of price relations between fertilizer (model input) and wheat yield (model output), including wheat price adjustments according to protein content. From regressions and backward induction analysis, the N application that maximizes the expected grain revenue minus fertilizer expenditure was estimated for four information cases; Case 1) Uniform application, assuming no prior information, Case 2) application based on soil type information, Case 3) application based on canopy sensor information and Case 4) application based on combined soil and canopy sensor information.
Findings from this study indicated that decisions with soil information alone provide an annual differential gross margin of variable rate application (without considering cost of information and technology) between 3.88 and 13.30 € ha−1 across price and soil variation. This margin approximately doubled with applications based on canopy sensor information and further doubled again with applications based on both soil and canopy sensor information.
Thus, knowledge of the soil has the potential to improve interpretation of sensor signals for fertilization planning. The results may guide developers to decide on what type of information should be included in their decision support systems.
[Display omitted]
•Usually variable rate application (VRA) of N is based on sensors or static information. So far, studies of VRA net returns have shown mixed results.•By focusing on differential gross margins, this study examines if soil or canopy sensor info provide the best basis for making VRA decisions.•The best N strategies were found by using backward induction on outputs from a crop simulation model, used as proxies for real world sensor info.•Gross margin increases up to 13 € ha−1 based on soil info. This margin doubles with sensors and further doubles with both soil and sensor info.•Findings suggest a combined use of info to obtain higher net returns and adoption of VRA, if both types of info can be acquired at low cost.