Eucalyptus plantations in New Zealand are occupied by a number of exotic insect defoliators and have increasing risks of new pest incursions. Pest outbreaks causing significant defoliation can reduce tree growth and productivity. Integrated pest management (IPM) strategy is useful to reduce potential risk of insect outbreaks and minimise pesticide use that has negative impacts on the environment. However, IPM in forestry plantations in New Zealand is still in its infancy. An industry centred on the production of naturally durable wood products is being developed in dryland areas in New Zealand. One of the priority species in the emergent industry is Eucalyptus bosistoana, which is drought tolerant and can produce highly durable heartwood. For durable species, including E. bosistoana to be considered as a commercially valuable option for planting in the future, we need to understand the risk and impact of currently present insect defoliators on these species.

Understanding the population dynamics of key insect defoliators is essential to predict their outbreak potential. Hence, insect surveys were conducted for Paropsis charybdis, Opodiphthera eucalypti, Strepsicrates macropetana and Phylacteophaga froggatti over two growing seasons in a dryland E. bosistoana site. Additionally, an insect development assay was conducted in the laboratory to attain base temperatures and degree-day requirements (DD) of life stages of P. charybdis (the most important eucalypt insect pest in New Zealand) to construct a DD model to simulate its phenology. Results showed that the observation of one generation of P. charybdis was different from previous studies, likely due to the drought conditions at the site. One to two generations were observed for O. eucalypti, and multiple overlapping generations were observed for S. macropetana and Ph. froggatti. The model was most capable of predicting voltinism of P. charybdis under the scenario assuming longer DD requirement of median egg laying age and hibernation start by 20 March, or a scenario with assumptions of shorter DD requirement of median egg laying age, hibernation start by 20 March and later overwinter adult emergence date (late September). Prediction of appearance of life stages was not highly accurate, but models that assumed shorter DD requirement for the median egg laying age tended to be the most accurate.

To assess the impact of defoliation on the growth of young E. bosistoana in dryland area, a trial simulating different defoliation severity (moderate and severe defoliation) and timing (spring and late summer, and spring plus late summer) effects on the growth of E. bosistoana was conducted. Only spring moderate defoliation did not significantly reduce tree growth, while other defoliation treatments significantly reduced either diameter or height growth. Severity of defoliation had a negative relationship with tree growth, but there was no significant difference observed between moderate and severe defoliation treatments. Late summer defoliation had a larger impact than spring defoliation, and this was exacerbated by defoliation severity. Repeated defoliation had greater negative impact on tree growth relative to single defoliation events. These results imply that spring moderate defoliation may not require pest control.

With the objective to identify families of E. bosistoana that have higher/lower resistance or tolerance to insect defoliation, and the most suitable method for this purpose, tree health assessments were conducted on 14 E. bosistoana families and 1 E. globoidea family using four assessment methods (based on defoliation levels and pest loads) over two growing seasons in an E. bosistoana site. Significant variation in insect susceptibility and tolerance was found between E. bosistoana families to the examined pest species except O. eucalypti. Southern provenance families were found to be more insect tolerant. The single E. globoidea family and Family 125 (Bungonia provanence) of E. bosistoana were found to be relatively fast growing and resistant to examined pests. These families should be maintained in the breeding programme.

To assess the between species variation in susceptibility to Paropsisterna variicollis, tree heath assessments were also conducted on 11 durable eucalypt species at three dryland sites in the Hawke’s Bay region in the North Island. Significant between species variation in defoliation and pest loads of Pst. variicollis was observed. The most susceptible species tested were E. tricarpa and E. bosistoana, while the least susceptible species were E. macrorhyncha.

Implications of the thesis cover three aspects of IPM, including pest monitoring, defining control action thresholds and tree improvement (selective breeding) to reduce insect outbreaks. The findings from this thesis can be applied more broadly to the sustainable IPM of the developing New Zealand durable eucalypt industry and the wider plantation forestry industry.