Classical biocontrol systems are sometimes treated as an exercise in community assembly. As such, they include multiple species interactions. This thesis explores multi-species aspects in classical weed biocontrol, using thistles as a case study. The abundance, phenology and impact of three biocontrol agents were followed on their target host, Carduus nutans L. and are described, for the first time in New Zealand for two of them (Urophora solstitialis L. and Trichosirocalus horridus sensu (Panzer)). Composition in New Zealand of the recently revised Trichosirocalus weevil species complex was surveyed nation-wide. One species only was found, albeit exhibiting a wider host range than anticipated from the published revision. Interspecific interactions and individual and combined effect of multiple biocontrol agents on C. nutans were tested in cage setups; the effect on the weed population was then estimated by manipulations of an existing matrix population model for this weed in New Zealand. The potentially better seed predator (U. solstitialis) was outcompeted by the worse seed predator (Rhinocyllus conicus (Froehlich)) which has similar niche preference. Urophora solstitialis was also adversely impacted by the crown-root feeder (T. horridus). Trichosirocalus horridus affected C. nutans survival, even at the medium density used, and significantly reduced potential seed production by 33%; in field densities, T. horridus is likely to affect C. nutans even more. Urophora solstitialis was estimated to destroy about 28% of the remaining seed in the absence of the other agents, and about 17% in the presence of T. horridus. The estimated combined effect of T. horridus and U. solstitalis on C. nutans population growth rate was greater than the effect of either agent alone. In the face of growing weed invasions, multiple thistle species were used to test ‘multi-targeting’ as a novel approach to target groups of ‘sleeper weeds’. Both in a field experiment and in a field survey, the seed predator R. conicus was found to attack and damage some ‘non-target’ thistle species more in the presence of the target species (C. nutans) than in its absence; however, levels of attack on non-target species were always modest. The ultimate goal of biocontrol is to reduce weed populations. A field survey revealed that current population densities of multiple thistle species in Canterbury are not obviously lower than in the mid 1980s, when only R. conicus was present. This may be because successful biocontrol has reduced the management input required to maintain the same thistle density.

Classical biocontrol systems are sometimes treated as an exercise in community assembly. As such, they include multiple species interactions. This thesis explores multi-species aspects in classical weed biocontrol, using thistles as a case study. The abundance, phenology and impact of three biocontrol agents were followed on their target host, Carduus nutans L. and are described, for the first time in New Zealand for two of them (Urophora solstitialis L. and Trichosirocalus horridus sensu (Panzer)). Composition in New Zealand of the recently revised Trichosirocalus weevil species complex was surveyed nation-wide. One species only was found, albeit exhibiting a wider host range than anticipated from the published revision. Interspecific interactions and individual and combined effect of multiple biocontrol agents on C. nutans were tested in cage setups; the effect on the weed population was then estimated by manipulations of an existing matrix population model for this weed in New Zealand. The potentially better seed predator (U. solstitialis) was outcompeted by the worse seed predator (Rhinocyllus conicus (Froehlich)) which has similar niche preference. Urophora solstitialis was also adversely impacted by the crown-root feeder (T. horridus). Trichosirocalus horridus affected C. nutans survival, even at the medium density used, and significantly reduced potential seed production by 33%; in field densities, T. horridus is likely to affect C. nutans even more. Urophora solstitialis was estimated to destroy about 28% of the remaining seed in the absence of the other agents, and about 17% in the presence of T. horridus. The estimated combined effect of T. horridus and U. solstitalis on C. nutans population growth rate was greater than the effect of either agent alone. In the face of growing weed invasions, multiple thistle species were used to test ‘multi-targeting’ as a novel approach to target groups of ‘sleeper weeds’. Both in a field experiment and in a field survey, the seed predator R. conicus was found to attack and damage some ‘non-target’ thistle species more in the presence of the target species (C. nutans) than in its absence; however, levels of attack on non-target species were always modest. The ultimate goal of biocontrol is to reduce weed populations. A field survey revealed that current population densities of multiple thistle species in Canterbury are not obviously lower than in the mid 1980s, when only R. conicus was present. This may be because successful biocontrol has reduced the management input required to maintain the same thistle density.

Cirsium arvense (L.) Scop. (Californian, Canada, or creeping thistle) is an exotic perennial herb indigenous to Eurasia that successfully established in New Zealand (NZ) approximately 130 years ago. Presently, C. arvense is considered one of the worst invasive weeds in NZ arable and pastoral productions systems. The mechanism most commonly invoked to explain the apparent increased vigour of introduced weeds is release from natural enemies. The enemy release hypothesis (ERH) predicts that plants in an introduced range should experience reduced herbivory, particularly from specialists, and that release from this natural enemy pressure facilitates increased plant performance in the introduced range. In 2007 broad surveys were carried out in NZ and central Europe in order to quantify and compare growth characteristics of C. arvense in its native vs. introduced range. Additionally, permanent field plots were established in NZ and Europe where natural enemies were excluded with the use of insecticide and fungicide applications, and compared with controls (ambient natural enemy pressure). The impact of the specialist leaf-feeding beetle, Cassida rubiginosa Müller, which was recently released in NZ as a biological control agent against thistles, was also assessed. From the field surveys, significantly more endophagous herbivory was present in the native range compared to the introduced range, as predicted by the ERH. Endophagous herbivory in NZ was solely from the capitulum-feeding weevil, Rhinocyllus conicus (Frölich), and was only found in the North Island surveys. No stem mining attack was found anywhere in NZ. The proportion of shoots attacked by the specialised rust pathogen, Puccinia punctiformis (Str.) Röhl., was similar in both the native and introduced ranges. Interestingly, this has casted doubt on the idea that stem-mining vectors, such as Ceratapion onopordi Kirby, are important for transmission of the rust pathogen. Contrary to the ERH, there were no significant difference in plant performance between the native and introduced ranges, or differences could be explained by simple climatic factors. Climate tended to be more favourable for growth of C. arvense in NZ. In the permanent field plots in the native range, population growth of C. arvense was significantly greater where natural enemies were excluded, suggesting that insect herbivores and pathogens might have a regulating influence on the population growth of this plant. Furthermore, the probability of shoots transitioning to the reproductive growth stage was enhanced when insect herbivores were excluded, indicating that natural enemies might influence plant development. The biological control agent C. rubiginosa reduced the growth of C. arvense, although the impact of this herbivore was minimal in comparison to interspecific plant competition. Thus, although there is reduced specialist natural enemy pressure in NZ, the growth of C. arvense is not significantly different from in its native range. Nevertheless, there is some evidence that natural enemies in the native range might have a regulating influence on the population dynamics of the plant, and that the specialist herbivore, C. rubiginosa, can impact the plant in certain conditions.