Wheat insect species infesting

An automated NIRS system capable of scanning individual grains containing late-instar larvae of S. oryzae, R. dominica, or S. cerealella at the rate of 15 kernels/min has been developed in the United Kingdom (Chambers et al., 1998). The system was effective and could detect the infestation irrespective of the type/class of wheat, its protein content (range 11.32 16.2%) and moisture content (range 10.0 13.2%). The minimum detectable size of the insects by NIRS varied between species. As identified by x-ray analysis, the NIRS system has been shown to detect R. dominica as small as 1.1 mm2 with 95% level confidence, whereas for S. oryzae it was 2.0 mm2, and for S. cerealella 2.7 mm2. For a particular insect species, the accuracy of detection increases as insect development proceeds. Accordingly, in S. oryzae the accuracy of detection of first instar larvae was 10%, second instar larvae 24%, third instar larvae 82%, fourth instar larvae 95%, and the accuracy was 100% for pupae and adults. 

---