Reducing Variability in Pea Crop Yield Using Multi-Spectral Imagery

Introduction

Peas are an important crop grown in New Zealand, especially in Canterbury, which is the main growing region for peas. Peas play a vital role in maintaining the biological and economic sustainability of New Zealand’s arable industry. They enhance soil fertility and structure, provide breaks for disease in arable rotations, and provide good monetary return if high yields are achieved. Many farmers have concluded peas to be very hard at making profits off due to variability in yield for a number of reasons. This has created a decrease in areas sown in peas, which as a flow on effect, can lead to a less sustainable model of intensive cropping via the loss of biological and economic advantages.  Peas also provide and an increase in the use of synthetic fertiliser and pesticide application.

In order to investigate the variability in pea crop yields, Dronescape were tasked with flying a 26ha pea crop with a Rededge Multispectral sensor in order to gain information on plant health implications which may have been causing variability in yield. The end result was to find out these factors affecting the target crop, and how these may be mitigated and overcome. The crop was flown in early December which is during the flowering stage. From the ground, no variability in stand establishment and maturity was observed with the naked eye, apart from around the entrance to the paddock where heavy machinery had caused compaction and in turn reduced growth

After flying the area with our UAV equip with our sensor, we then processed the data using the Atlas software post-production tool.  A number of interesting observations were made and Dronescape agronomist Ben Beachen was able to draw conclusions as to why variability of yield would occur and to consult recommendations on how to mitigate these influences.

Image 1 – DSM (Digital Surface Model)

Image 1 revealed significant variation in topography and elevation with variation in metres above sea level ranging from 101m in the blue area, to 86m in the dark red area. The DSM shows the natural water flows within the paddock and how this would affect drainage and surface ponding. Straight away we realised that drainage was one of the compounding issues that caused variability in the pea crop yield, especially when the soil type of the area was taken into account, which consisted of an impeded drainage Temuka clay loam.

Peas prefer free draining soils and are sensitive to compaction affected crop establishment, root growth, and even stand maturity, however, they can be successfully grown on heavier soils providing there is effective drainage and good soil structure maintenance.

The low point shown in the darkest red area is an altitude of 86m above sea level while the purple areas reach 101m above sea level.

                                                                                                                                              Image 2 – NDVI Reflectance Map

Image 2 shows the NDVI (Normalised Difference Vegetation Index) which gives a general indicator of plant biomass. The map shows there are a number of areas with poor biomass, which when compared to the DSM, show a correlation to the theory that low spots in the paddock may cause crop yield variably due to surface ponding and poor drainage, resulting in poor crop establishment.

                                                                                                                                               Image 3 – Chlorophyll Map

Image 3 signifies areas of leaf chlorophyll deficiency in red. As a plant matures, leaf chlorophyll levels drop as a reduction in energy requirements due to leaf growth taking place as the plant begins to go reproductive. As the pea crop had flowered and pods where beginning to form, the observed low chlorophyll levels seen in Image 3 are expected. It can be seen that the areas that have low biomass from assumed poor establishment from surface ponding, correlated to more leaf chlorophyll (seen in the areas ranging from yellow to blue). This brought us to the conclusion that these areas will have a delayed maturation of pea pods, in turn creating a reduction of yield as under-mature peas will be rejected at a factory processing level.

Recommendations

After drawing conclusions to variability in pea yields from the paddock of interest the next job was to come up with a solution to the problem with actionable recommendations. In this case the issue causing the biggest variation in yields comes down to poor drainage, resulting in surface ponding, and poor establishment as a consequence. To a lesser degree, compaction from uncontrolled traffic during establishment and harvesting of peas and other crops is likely to be causing some yield variation. From these conclusions the following recommendations to land improvement and management practises where made:

Precision Levelling

Precision levelling is a land improvement that can increase the productivity and value of an agricultural cropland by improving field drainage or increasing the efficiency of surface irrigation.  The cost of precision grading represents a long-term investment in farm assets. Precision levelling is the operation of shaping the surface of land to predetermined grades so that the surface slopes to a drain or is configured for efficient irrigation water application. This operation is typically performed by tractors pulling dirt buckets or scrapers that pick up soil in the high points in a field and deposit it in the low points. Dirt scraping operations are controlled by laser equipment that enables the slope of a field to be cut to a specific grade. The potential benefits from precision land levelling can be significant in reducing surface water ponding , reducing scouring during heavy rainfall events, and optimising crop water usage, which in turn reduces variability in crop yield.  This is however a costly operation to carry out therefore benefits over cost must be weighed up when deciding on the best way forward.

Controlled traffic management

Peas are a crop that are very sensitive to soil compaction, particularly where it can restrict root growth and development. Optimum crop production is achieved in soils that allow biological activity through aerobic conditions. When heavy machinery traffic is not managed in cropping situations, the air is squashed out of the soil which leads to compaction. Controlled traffic management is based on the principle of using permanent compacted wheel tracks to support all load bearing machinery in order to minimise the spread of wheel tracks from heavy machinery throughout the paddocks. Machinery are configured to have the same wheel axle width in order for this the work. Under this system the remainder of the soil is not affected at all by load bearing machinery so soil remains in the best possible condition for crop production.

Conclusion

In conclusion, the camera enabled us to identify and highlight areas of interest within the pea crop that displayed variability that would normally be near impossible to build a picture of through the naked eye. By using the infrared image from a birds eye view we were able to pinpoint the areas of variability with efficiency and accuracy as opposed to ground methods. The data also gave us the ability to draw correlations between the associated factors such as drainage, chlorophyll content, and crop biomass, and the crops variability. The overall conclusion of this project is that this data can be used to draw precise recommendations on improving crop yields.