Abstract

Growers of floricultural crops are still dependent on chemical insecticides because of low consumer tolerance to damaged produce. More knowledge of natural interactions between insects and their host plants would allow insect pests to be controlled in a more environmentally friendly and effective way. This thesis analysed the system of biological interaction between the pest Frankliniella occidentalis (Western Flower Thrips) and its host plant Chrysanthemum x morifolium. Mathematical prediction models were developed to describe the growth and development processes of the plant and the insect population, on the assumption that there is a temperature-dependent interaction between flower growth and development and thrips population growth. An introductory study developed a method for linking flower growth and development that was used in subsequent studies. The second study aimed to predict growth responses of non-infested chrysanthemum flowers to temperature and irradiation, in order to distinguish between temperature and thrips effects on chrysanthemum flower size. Since F. occidentalis feeds on both flowers and leaves, a model constructed in the third study served as a prediction tool for food and habitat distribution for the thrips. By recording the final leaf length, leaf area distribution during the growing period could be re-constructed. The model was later included in the large population growth model as a leaf canopy model. The effect of three temperature regimes on early population growth of F.occidentalis was investigated in detail. The results showed that the observed rapid increase in population size could not be correlated with flower opening. The relative reproduction rate of F. occidentalis changed exponentially, probably depending on changes in population density and different reproductive strategies. Therefore, in a final model, population density played an important role during early population growth, whereas a decrease in food supply, in terms of damaged leaf area, controlled population decline at the end of the growth period. At lower temperatures (approx. 20˚C), the population decline could be simulated by simply manipulating food supply, while at higher temperatures (approx. 26˚C), the model underestimated population decline, indicating the need for a stress or migration factor in the system.