Dr. Christy van Beek, senior soil scientist at AgroCares and expert in soil fertility, shares what 20 years of experience has taught her about fertiliser recommendations in a series of blogs. In this third blog she explains the five most common techniques of making fertiliser recommendations:
[Blog 3] So many roads to Rome
Back to the techniques of making a fertiliser recommendation. Again, there are some decisions to make, starting with the fundamental approach of the recommendation. This is followed by the input parameters used and the way they are derived.
Fertiliser recommendations basically serve to restore or maintain the productive capacity of the soil. Therefore, they commonly consider the soil status and the amount of nutrients withdrawn by crops. To do so, one can use one of the following methods:
1. Balance method
Apply as much as is withdrawn from the system taking into account managed (e.g. harvest) and un-managed (e.g. erosion) losses. This method is sound and clear, but assumes that the soil is currently in good condition. If this is not the case, this approach will ‘recycle poverty’.
2. Threshold method
When the content of a certain element in soil is below a specific threshold; it is applied to the level in which it is not insufficient anymore. This method is widely used, but the correct determination of the threshold is difficult, nor does it consider interactions between nutrients.
3. Trial and error
Different combinations of fertilisers are applied and one looks at the highest yield response. Although this is not very scientific; the approach is often used.
4. Yield response method
This method is based on field experiments in which a relationship is determined between the amount of fertilisers and the yields obtained per crop. Apart from the costs of doing extensive field trials, the method is also only valid within the area of the experiment.
5. Simulation modelling
Empirical, semi empirical and physiological relations are used to calculate the exact fertiliser requirement. This method is actually only applicable for scientists and requires substantial amounts of data.
Combination of different techniques
Often, the methods above are combined. For example, a threshold is corrected with a balance method, or the yield response curve is used for simulation modelling. Many examples of the above methods can be found in Ethiopia, which is currently reviewing its fertiliser policy. Notwithstanding the method, all except the balance method require soil data.
Soil data can be obtained from wet chemistry laboratories or from ‘dry chemistry’, which uses spectrometers like the technology developed at SoilCares.
More information about making fertilizer recommendations?
Would you like to learn more? Discover what makes a proper fertiliser recommendation in the first blog from the series or read about the factors that influence the adoption of fertiliser recommendations by smallholder farmers in the second blog.
Text written by Dr. Christy van Beek