Ecology and evolution of the invasive willow sawfly Nematus oligospilus Förster

Invasive organisms can have major impacts in their new environments, affecting biodiversity, community structure and ecosystem processes, and are one of the main causes of species extinction. Willows (Salix spp.) are among the worst environmental weeds in Australia, with important consequences for aquatic and terrestrial environments. Nematus oligospilus Förster (Hymenoptera: Tenthredinidae), the willow sawfly, has been introduced inadvertently into most regions of the Southern Hemisphere, where it established and quickly became invasive. It was first found in South America, then southern Africa, New Zealand and lastly, Australia. At high population densities, N. oligospilus larvae can cause extensive willow defoliation and even lead to tree death. This thesis assessed the ecology and evolution of N. oligospilus in its introduced range to understand the reasons behind its persistence, spread and potential long-term impacts. I focused on (a) factors that determine the pest status of N. oligospilus, (b) abiotic conditions required for establishment and persistence, (c) host-plant interactions, and (d) reproductive mode and population structure. Where willows are considered beneficial introductions (i.e. New Zealand), N. oligospilus is considered a pest, whereas where willows are considered noxious weeds, such as in Australia, N. oligospilus is deemed as potentially beneficial for willow management. A major field survey in New Zealand and Australia showed that N. oligospilus was widespread in both countries, occurring on all willow taxa, albeit with major differences in population densities. Highest densities occurred on Salix fragilis and its hybrid, Salix x rubens. Ecological niche modelling based on current sawfly distributions predicted that N. oligospilus has reached its full potential distribution in New Zealand, but that it had not yet reached its full range in Australia. Molecular analyses showed that N. oligospilus is a strict parthenogen in its introduced range. Furthermore, few genotypes were found in New Zealand, Australia and South Africa, with two major widespread genotypes that are environmental generalists or “superclones”. Fourteen other localized, closely related clones were detected. All clonal genotypes detected in the Southern Hemisphere are potentially derivable from the commonest one by mutation, and lineages differed in most cases by the smallest possible mutational unit in microsatellites. There was evidence of country-specific lineages derived from a local ancestor by mutation. Genotype patterns were consistent with sawflies dispersing themselves at local to regional scales, but with human assistance at larger scales. Host plant associations had strong effects on larval growth and adult oviposition, putatively related to fitness. Differences in host-plant interactions were found within clonal (microsatellite) lineages; however, it is unclear if these are due to inherited environmental effects or local adaptation. The willow sawfly represents a great opportunity to study an unimpeded invasion replicated across continents, particularly of a species that has switched from sexual to asexual reproduction during invasion. This research clearly indicates the high colonisation success and rates of spread of N. oligospilus in Australia, and that by combining field based, molecular and laboratory approaches, significant insights can be gained into the processes underlying invasion