| Abstract | Of the nineteen canegrub species in Australia, greyback canegrub which occurs from Plane Creek northwards is the most important. Growers rely heavily on insecticides for greyback canegrub management, and effective insecticidal treatments are now available for both plant crops and ratoons. However these treatments are expensive, and there is no system that allows growers to strategically apply insecticides to only those fields which really need treatment. The aim of this project was to test a system which would allow growers to vary their treatment decisions as circumstances changed. In a previous SRDC-funded project (BSS257), BSES Limited developed a set of models which predict numbers of greyback canegrubs one year ahead. Required information includes canegrub numbers in the current year and presence of visible grub damage in canefields. The Mt Kinchant Grower Group engaged BSES as a consultant to implement this system on Group farms, to test the predictive system and evaluate the costs and benefits of a grub-management consultancy that could be used by other growers in the industry. BSES monitored each of the 10 farms within the Group from 2008-2010. Canefields were sampled for canegrubs by BSES in April-May of each year – 78 fields in 2008, 80 in 2009 and 46 in 2010. Twenty stools were dug in most of these fields and grubs identified and counted. A sample of at least 50 grubs was then reared to adult and causes of any deaths were diagnosed (identifiable pathogens are Adelina, Metarhizium and milky disease); disease levels were very low in both 2008 and 2009 while grubs from 2010 are still being reared. Fields were inspected before harvest and any visible damage recorded; aerial photographs were taken in 2008 and 2009 to help locate grub damage. Gappy ratoons that may indicate grub damage were recorded after harvest. The locations of grub-infested stools and grub damage were recorded in a GIS layer. Maps were printed showing the status of fields on each farm in terms of current insecticidal protection, grub numbers (for sampled fields) and visible damage (for all fields on the farm). The risk of grub attack in the following year was quantified using the predictive models. Group members received a package each year that included the field-status maps, farm report and treatment recommendations. There was general agreement between trends of actual and predicted grub numbers in 2009 and 2010 but with a lot of unexplained variation, particularly in 2010. Treatment decisions tended to err on the conservative side, which is not necessarily a bad thing. Damage was low on most farms during the project. Unexpected damage was only observed in a small number of fields, and that damage was localised and light in almost all cases. This project allowed the Group to have input into the type of information that growers require from a canegrub-management service, and has allowed the service to be costed and its functionality evaluated. Data collected in the project will be used to fine-tune the predictive models. |
