The development of a practical method for removing insects from large samples of grain

Abstract

Pest detection in grain, particularly during transit, is of great commercial importance. Current methods will not detect insects at low population densities and often cannot estimate levels of infestation accurately. This limits management options and encourages the prophylactic use of pesticides.

Much research is and has been directed at improving methods of detection. However, the results so far either offer very limited improvements or still require much more development before they are suitable for commercial application. Following a thorough review of the available options, it was concluded that some method of separating insects rapidly from large (5kg or more) samples of grain, offered the most practical route to improved detection.

Prototype machines using sieves or aspiration to separate insects from grain were built and tested. The best results were obtained with a flat-bed, reciprocating sieve and development was concentrated on this design. The rate of flow of grain through the machine had a profound effect on the recovery of insects but the angle of the sieve and the amplitude of shaking had much less effect. The optimised prototype siever was able to give consistent recoveries of two important species of grain pest close to 100%.

A pre-production prototype machine was then built and tested. This machine incorporated all the design principles of the first prototype but also was designed to meet operating and safety standards required of a commercial machine. Tests showed that this unit could process 10kg of wheat in 1.8 minutes, recovering almost 100% of check insects in the grain. It proved capable of detecting infestations at population densities of 0.2 insects/kg and also gave a reliable estimate of population densities. Similar results were also obtained with barley. The machine proved easy to use.

The performance of the pre-production prototype represents a large improvement over current, commercial methods, and also offers more effective detection than many of the new approaches currently being investigated. Samplex Ltd now market a machine based on this project.