**4. Conclusion**

98 Zoology

that had migrated from an infected orchard into the newly planted orchard would die after feeding (if an individual was virulent it might transmit HLB, but would not reproduce on the tree). After this period, individuals migrating from the infected orchard would reproduce in the newly planted orchard. The number of newly infected trees was lower in the plantings treated with systemic pesticide than in those with the negative controls. We concluded that treatment with systemic pesticide was effective in suppressing the spread rate of HLB in a newly planted orchard, and the effect was definitive when the virulent *D. citri* migrated from a nearby orchard (Fig. 9). Part of this study have been reported in Kobori

0 12 24 36 48 60 72 84

Dashed line, negative control (did not treat systemic pesticide); soiled line, treated with systemic pesticide (effective for 24*t*). Number of initial *H* trees in the simulated field was 256. No. of newly

Although these results generated valuable suggestions, there was a problem in our simulated conditions. In a number of replications, extinction of the vector insect occurred. In actual observation, even insects not found in a given season could be found in later observations. One cause of this may be the closed nature of our simulation system. In reality, vector insects could immigrate into the field. Currently, we are expanding the model to incorporate this possibility. In addition, the dispersal kernel was developed for calculations of single-field behavior. We must revise the kernel to simulate larger scenarios,

infected trees = *LP+IP+D* (excluding initial *D* trees). Modified from Kobori et al. (2011).

Fig. 9. Effects of systemic pesticide to suppress HLB in the field.

such as those involving multiple, separated fields.

*t*

Initial number of healthy trees in the field

et al. (2011c).

0

50

100

150

200

250

300

No. of newly infected trees

This chapter described the development of an individual-based modelling technique to simulate disease-spread dynamics. Our model was able to provide parameters for each individual citrus tree and the respective vector insect, and thereby to examine diseasespread dynamics in the simulation field by calculating the cumulative results of the individual behaviors. This model can be applied to many diseases vectored by insects, although we developed the model for HLB.

Our simulated results suggested that both delaying the transition from latent period (*LP*) to infectious period (*IP*), and removing infectious (*IP*) trees, suppressed the spread rate of HLB in an orchard. Additionally, the results of our a preliminary trail suggested that cooperative control of *D. citri* on the part of orchard owners in a given area*,* to reduce the number of virulent individuals, may be effective in suppressing the spread rate of HLB.
