Operational research analysis of patch spraying

Abstract Sprays to control infestations of major cereal weeds such as wild oats and blackgrass are expensive and these weeds frequently appear to occur in patches. Spraying only those patches with weeds thus appears attractive from both an environmental and economic viewpoint. This paper describes a model that considers a field in patches and simulates the germination and control of weeds. The model shows that apparently extreme observed patchiness occurs naturally from normal whole-field weed-control strategies. Threshold levels for whole-field spraying are calculated and shown to be a function of both herbicide cost and effectiveness. The profitability of patch spraying comes mainly from not spraying patches with no weeds and is greatest where there are a small number of seeds per patch, this being the proper measure rather than seeds m−2. At 120 wild oat seeds per patch and 91% average herbicide kill, the value of patch spraying is £101 ha−1 over 10 years which, for a e200 ha farm, is equivalent to a capital cost of £13 320 that can be justified on equipment to enable patch spraying, from a purely economic viewpoint without considering environmental advantages. At 1000 blackgrass seeds per patch, the corresponding capital cost is £5540; these need patches 12 m2 in area. The larger the patch, the less capital cost that can be justified. The more effective the herbicide, the greater the value of patch spraying and the higher the justifiable threshold. The saving is insensitive to the spray threshold level and ± 5 plants m−2 has little effect on the overall cost.