FRAMEWORK TO SUPPORT DECISIONS ON CHEMICAL PEST-CONTROL UNDER UNCERTAINTY, APPLIED TO APHIDS AND BROWN RUST IN WINTER-WHEAT

An approach is presented for supporting decision making on chemical pest control in an uncertain environment. Timing of chemical control is considered as a decision problem, which is analysed in terms of objectives, strategies, selection criteria, and a model of the system dynamics. Return on expenditure and insurance are considered to be two generally applicable objectives in pest control. These objectives are made operational by defining the profitability of strategy A compared with strategy B as the probability that A results in lower costs than B, and the risk associated with a strategy as the 0.90 quantile of the probability distribution of costs associated with that strategy, respectively. The authors propose to express the performance of relevant strategies in terms of profitability and risk. Risk-neutral damage thresholds where expected costs of no chemical control at any time just exceed those of immediate application of a pesticide, constitute a yardstick for assessing a strategy's performance. The approach is illustrated using a deterministic model of the pathosystem aphid-brown rust-winter wheat. Uncertainty is modelled as random inputs. No control and immediate pesticide application represent the only relevant strategies, as postponing chemical control decreases profitability and increases risk. The decline in risk attributable to pesticide application is larger for brown rust than for aphids. At risk-neutral damage thresholds the probability of positive returns on pesticide application amounts to only 30%. The causes of this bias to chemical control in recommendations for risk-neutral decision makers are discussed. Current recommendations for both aphids and brown rust appear to presume that farmers become more risk-averse as crop development advances.