Some problems and approaches to modelling insect phenology
The development of models that predict the timing of important events in insect life cycles for control purposes has attracted considerable research attention in recent years. Approaches to phenological modelling range from simple physiological-time based models and static models, to more complex models that are process orientated. The use of hypothetical data to verify internal model assumptions in the absence of adequate observational data is illustrated. The implications of the rate summation or Kaufmann effect and the use of small and large samples of historical data for model validation are discussed. Predictive sample reuse methodology is used to validate a simple model based on long-term insect, climate and host-plant phenology data. Model validation indicated that possible interactions between host plant phenology, weather and insect population events merit close examination.