Precision farming insights – Part 2: First step in Variable Rate Technology
In a previous article, I wrote about what the best steps might be for applying nitrogen-based fertilizers at variable rates, and today I’m bringing something practical to demonstrate what I discussed back then.
After several years of working with various farmers on variable rate fertilizer applications, using NDVI images as a reference but also correlating them with personal observations and in-field decisions, this year I had to demonstrate the positive results of this approach to a farm that cultivates approximately 20,000 hectares in Romania. For this phase, we selected an area of about 1.000 hectares, all cultivated with wheat but from varied geographic regions. There was a distance of over 80 km between the easternmost and westernmost fields.
In February, we studied satellite images and, with the help of specific applications, set a few points of interest in each field – where we had both high and low NDVI values – to concretely determine the reasons for these high or low NDVI index values, with in-field observations. For more details on what and how we did this, see the previous article.
After all the field visits, we were able to make a fertilization decision, which, for all fields, was to apply more fertilizer in areas with lower vegetation values. The total amount of fertilizer used was the same as what would have been applied in the usual, non-variable rate approach. The fertilizer used was lime ammonium nitrate, with doses ranging between 200 and 280 kg per hectare. Of course, there were also areas where no fertilizer was applied at all, though these were not significant in size.
In this case, the application was done using a machine that’s somewhat atypical for a granular fertilizer spreader—the Rauch Aero GT, which has the distinctive feature of being a pneumatic, row-based distributor. The machine was controlled using ISOBUS technology, and the application prescriptions were loaded manually from a USB drive, although there was also the option to send them directly from the office to the monitor.
The first results were checked approximately one month after application, and already the variations were much less pronounced in the field, but the final verdict was to be given by the harvest maps. I also need to mention that the fields in question are not irrigated, but they benefited from a satisfactory rainfall regime during that period (e.g., 88 liters per square meter was the accumulated amount of precipitation on one field in March and April).
But as I said, the final test came at the time of harvesting, when it was very clear to everyone that the maps from the combine show a much more uniform field in terms of production. For example, the field whose productivity map is shown below has an area of over 133 hectares.
In the yield map from the harvest, a clear uniformity of production can be observed, and in the image above, the yellow lines visible are caused by the initial furrowing to create access roads for the trucks carrying the harvest.
For a better view of a field where we applied variable rate and its correlation with the harvest map, see the following image, where you can clearly observe that there are still small variations at harvest; however, in my opinion, nothing major.
For confidentiality reasons, I cannot disclose the yields obtained, but in any case, the goal of this experiment was to achieve crop and harvest uniformity, not necessarily to increase it.
Applying nitrogen at variable rates based on satellite images is now very likely to be implemented across the entire farm, as the results speak for themselves.
If you’re interested in trying this yourself or need help, don’t hesitate to write to me at iulian@stefan.com.ro.
