No one knows their own paddock like the grower. Hours spent on the seeder, sprayer and header gives plenty of time to form an understanding of parts of a paddock that consistently perform well and those that aren’t as productive. However, using Variable Rate Application (VRA) of inputs to target underlying variability within a paddock can be a challenge, even with large potential cost savings or production benefits.
There are many potential sources of variability in a paddock and surveying the area from the seat of a tractor or header can provide an indication of variability in outcomes such as yield and growth, but it doesn’t reveal the underlying causes of this variation. Variable rate fertiliser application relies on the ability to effectively target differences in response to nutrients that can then make an economic difference to production.
As part of an ongoing focus on precision agriculture, BCG is undertaking a trial across eight paddocks this season to help demonstrate an approach to assessing the potential benefits of VRA. This includes assessing the claim that many paddocks in the southern Mallee are too uniform to justify VRA.
Trial paddocks were selected in pairs by four growers: one that they identified as ‘variable’, the other as ‘uniform’. Test strips with varying rates of nitrogen and phosphorus have been applied to map responsiveness across each paddock.
Figure 1. An example of the paddock zones used to target soil sampling transects. Different colours correspond to areas with different ‘average performance’ when satellite data from several years is stacked on top of each other using the ‘PA Stack’ online tool. The three soil sampling transects taken across the trial area in the paddock are indicated in white. Resolution of 10m/ pixel.
So far, soil sampling has been undertaken on three different transects in each paddock, and NDVI (crop greenness) was collected at approximately GS30 using a drone to give an initial idea of what response is being observed. This data has shown some interesting results (Figure 2).
Figure 2. Graph showing variability from the average NDVI (%) along a strip down each of the paddocks selected by a grower. Greater variability can be seen along the strip in the variable paddock where NDVI values change by close to 12% along the strip whereas that change is only around 2% in the uniform paddock.
Expectedly, farmers have been accurate in identifying their relatively variable or uniform paddocks. This holds true for both overall variability in NDVI along a control strip, as well as nutrient responses indicated by comparing strips that differ in nitrogen (N) or phosphorus (P) rate.
However, variability to some degree is still apparent in both soil test results and in NDVI from the uniform paddocks. Whether this level of variability will translate into economic differences in yield remains to be seen: it’s possible that they could still be large enough to justify a VRA approach.
Some results have also been unexpected, and could indicate the value of using a test strip approach to assessing paddock variability. One of the ‘uniform’ paddocks appears to have at least as much variability in P response than the related ‘variable’ paddock, even though it is indeed less variable in all other measures.
It’s too early to draw any conclusions, but these initial results certainly show some potential for the situation to be less clear-cut than it might appear to the naked eye. This also reinforces the benefits of tools like test strips when thinking about VRA.
Ultimately, whether each paddock is a viable candidate for VRA of nitrogen or phosphorous will become more clear at harvest, when we will perform a similar analysis of yield maps. Stay tuned!
