Delairea odorata Lem., an asteraceous perennial vine indigenous to southern Africa, has become naturalised and invasive in many subtropical regions including California, South Australia and Hawaii. Biological control offers a potential long term solution to the management of this species in exotic locations. This study analysed aspects of the biology of D. odorata in its native environment to determine its suitability to classical biological control. To this end an examination of the reproductive biology and natural enemies of D. odorata was made. A study of the pyrrolizidine alkaloid profile was also conducted. Reproductive biology: Delairea odorata reproduces both sexually by seeds and asexually by stolons. The flowering season occurs over the autumn months from April to June. Results of the pollination trials indicate that D. odorata is a cross compatible species and an obligate outbreeder. There is no specialised pollination system and the predominant pollinators belong to the families Apidae, Syrphidae and Calliphoridae. Following pollination, numerous small achenes are produced. Laboratory trials indicate that these achenes germinate readily between 10 and 25 °C and, although germination occurs in both the light and dark, light clearly stimulates seed germination. Greenhouse trials conducted to determine the effect of light on growth and reproduction indicate that D. odorata is a shade tolerant species which shows plasticity in terms of growth form and deployment of biomass in response to changes in light intensity. Growth rate and allocation of biomass to vegetative and sexual reproduction are highest at an intermediate light level. However, greatest allocation of biomass is to stem growth regardless of light level. Natural enemies: Surveys for potential biological control agents against Delairea odorata were conducted in KwaZulu-Natal and several phytophagous species were associated with the plant. However, only one potentially suitable control agent was identified, a stem galling tephritid fly, Parafreutreta regalis Munro. Preliminary studies indicate this species to be fairly host-specific, a valuable asset if it is to be considered as a control agent. Furthermore, as D. odorata proliferates extensively by means of stem regeneration and elongation, galling of these growing points by P. regalis may limit stolon spread in exotic locations. Two species of parasitic wasp (Braconidae) were found to parasitise P. regalis pupae. If P. regalis is to be used as a control agent the likelihood of parasitisation in the new environment must be determined. Pyrrolizidine alkaloids: Host-specificity in insects is often dependent on host-plant chemistry (e.g. alkaloids or essential oils). Thus prior to any biological control programme it is important to determine if there are ecotypes of the host plant present. An investigation to determine the specificity of the pyrrolizidine alkaloid profile of D. odorata, occurring across KwaZulu-Natal, was made. The results indicate the presence of nine retronecine based pyrrolizidine alkaloids which occur in similar proportions in locally distributed plants. However, these alkaloid profiles differ considerably from those published for D. odorata occurring in California. This is an interesting and important result which indicates that chemotypes of D. odorata may exist, a factor which must be considered in the initiation of any biocontrol. If chemotypes of D. odorata are present this may affect the behaviour of natural control agents on the exotic plant populations. / Thesis (M.Sc.)-University of Natal, Pietermaritzburg, 2000.