Data analysis aids input application
YIELD maps are seen by many as an expensive way of confirming the presence of problems which farmers already know exist.
Rabbits and compaction around headlands, and drainage problems are the usual villains, says Murray Lark of the Silsoe Research Institute. But a little more number crunching on the computer can reveal valuable information on which to base future inputs, he maintains.
Interpretation is complex. Many factors including pests, diseases, weeds, nutrients, soil physical conditions and management may limit crop yield, he notes.
The trick is to break the field down into generalised sub-regions, by identifying areas which produce consistently high yields, average yields and low yields.
This is done by the computer, which analyses clusters of yield information gathered over several years. It then produces a map, breaking the field down into several sub-regions.
Each sub-region shares similar yield trends over several seasons, so is likely to be affected by similar limiting factors, providing a basis to treat each as a distinct management unit, he says. For example, sub-regions show a strong correlation with soil type, allowing further investigation by sampling to identify the factors limiting yield.
"It offers a simpler framework for soil sampling. A few samples can be taken from one of each of the sub-regions, rather than having to sample the whole field." That provides a basis for varying fertiliser inputs across the field, says Dr Lark.
Analysing yield trends from a series of yield maps enables fields to be divided into areas which share the same limiting factors. Inputs can then be adjusted, says Murray Lark of Silsoe Research Institute.
Next week we continue our coverage of the British Crop Protection Council conference in Brighton, focusing on biological control, pest and disease monitoring and seed treatments.