Monte Carlo simulation method for forecasting the timing of pest insect attacks

Abstract

The timing of attack by pest insects can vary greatly both from region to region and from year to year because the rates at which insects complete their life cycles depend mainly on temperature. In addition, variation between individual insects in their rate of development can lead to a spread of activity. A simulation method, based on rates of insect development, has been produced for forecasting the timing of attack by pest insects. Variability is incorporated using the ‘same-shape property’ which implies that the coefficient of variation of the rate of insect development is independent of temperature. The method is feasible because it uses a fixed number of individuals from one generation to the next and simulates the timing of events rather than the population dynamics of the insects. The method is sufficiently flexible to allow insect resting phases to occur at the appropriate instant for each individual. Its effectiveness in predicting the behaviour of the cabbage root fly ( Delia radicum), the carrot fly ( Psila rosae), the bronzed blossom beetle ( Meligethes aeneus) and the large narcissus fly ( Merodon equestris) is demonstrated for one site in one season. To be of practical use, the method has been designed to use standard meteorological data. Estimation of soil temperatures at 6 cm and the derivation of equations for estimating diurnal variation are described in two appendices. The basis of the method is the integration of a temperature-dependent rate function over time. A third appendix describes a quicker and more accurate method of integration than that usually employed in pest forecasting work.