Risk mapping of redheaded cockchafer (Adoryphorus couloni) (Burmeister) infestations using a combination of novel k-means clustering and on-the-go plant and soil sensing technologies

The redheaded cockchafer (Adoryphorus couloni) (Burmiester) (RHC) is a serious pest of improved pastures in south-eastern Australia and current detection relies on pasture damage becoming visible to the naked eye. Various precision agriculture sensors are able to delineate spatial variability in soil texture and moisture content as well as numerous contributing factors to the photosynthetic 'vigour' of pastures, namely biomass, canopy architecture and species composition. The aim of this paper is to seek to determine whether the same technologies can be used to identify paddock zones prone to RHC infestation. This study investigates the association between data generated by a CropCircle™ (an active optical plant canopy sensor (AOS)), an EM38, (an electromagnetic induction soil sensor), and third instar RHC larvae counts. Results indicate that the red wavelength reflected component of the AOS from the pasture canopies offered the most accurate model of third instar RHC larvae count (residual mean square error = 1.04).