Biological control of Scotch thistle (Cirsium vulgare)

Abstract

Cirsium vulgare is a common pasture weed in New Zealand and was introduced into this country from Europe where this species is of minor importance as a weed. The cytological status of the local population of C. vulgare was investigated, as two chromosome races have been reported from Europe. It was concluded that the 2n=68 race is present in this area. The biology, ecology, and fauna associated with this species were investigated. It was found that high numbers of viable seeds are produced during the late spring and summer months. These seeds are shed around the parent plant, with a smaller number being dispersed further away. Massive germination of seeds occurs when sufficient moisture is present, provided the seeds are also exposed to light. Temperature in this area of New Zealand is not thought to limit germination at any time of the year. A proportion of seed enters the soil seed bank where they remain viable for at least twelve months, and probably longer, provided they are buried at a sufficient depth. If these seeds are brought to the surface they will germinate. Small seedlings of C. vulgare are destroyed by slugs and they also tend to crowd each other at high densities. However, once a rosette has become established (diameter > 30 cm) it will reach maturity and set seed. The fauna associated with C. vulgare was collected and identified, with most effort being directed towards the rosettes and the flowers and seed heads. The probability of finding insects or pathogens that would quickly kill the mature plant before it set seed was not thought to be high. The fauna collected could be characterized as non-specific, non-damaging, and impoverished with respect to Diptera and Coleoptera in comparison with European and Asian fauna collected from C. vulgare. The population dynamics of this species in the Auckland area were modelled using a number of transition matrices. These matrices were based on data of survival probabilities between pre-selected size stages of C. vulgare, fecundity values, and estimates of entry of seed into the soil seed bank. Matrices for bare soil and pasture were constructed for a number of different sowing densities. The dominant latent root of these matrices (λ) gives the yearly rate of increase for the populations studied. λ values in the range of five to fifteen are realistic for this species on pasture. Manipulation of elements within the matrices was performed to investigate the effects of increasing mortality at the rosette and seed stages. Such manipulation indicated that both rosette and seed destruction in the order of 40%-60% were needed to reduce λ to one and hence to stabilize population growth, i.e. to control this species. Economic and environmental implications of weed control strategies are investigated and the feasibility of biological control examined. Some aspects of biological control of weeds are critically reviewed and suggestions for improvements are given. Lastly, a number of possible control agents for this species are detailed, together with an estimation of their usefulness in controlling C. vulgare. Recommendations include the introduction of two of these species into New Zealand and further study on a number of others. Because of the light requirement for germination of seeds of C. vulgare, the possibilities of cultural control through manipulation of grazing regimes is also possible, and further study on this aspect of control is also recommended. The respective roles of industry, government-sponsored research agencies, and universities in developing biologically sound weed control strategies are included in the final summary.