Scotch broom ('Cytisus scoparius, L.'), is an exotic leguminous shrub, native to Europe, which invades pastoral and woodland ecosystems and adjoining river systems in cool, high rainfall regions of southeastern Australia. Broom has invaded 10,000ha of eucalypt woodland in Barrington Tops National Park in New South Wales, and is having a major impact on the natural ecology of the sub-alpine environment. It is extremely competitive with the native flora, retarding their growth and in many areas blanketing the ground and preventing growth of many understorey species in open forest areas. A number of rare and endangered species are believed to be under threat from this invasion. An active program to manage the threat is being implemented by the New South Wales National Parks and Wildlife Service. The management issues relevant to this threat are addressed through the application of a detailed bio-economic model of broom management. The results show that intervention in the management of broom in this natural ecosystem is clearly economically justified, and increases in the existing budgets appear to be justified. A combination of control measures, rather than any single measure, is almost always justified. Attempts to eradicate the broom invasion appear to be undesirable, so containment is the preferable strategy. Further, funding bodies should give assurances of future budget levels. The method of analysis reported here should be applicable to other types of biological invasions.

Weeds have a wide variety of impacts on society, the environment and the economy. The economic impacts are usually losses and these can be measured as costs of control, decreases in yields, and reductions in economic surplus. In this paper, we attempt to estimate these economic effects of weeds in Australia. The impacts of weeds on agriculture were calculated through all three measures. But due to lack of data, the impacts of weeds on natural environments, other public land, and Indigenous land, could only be measured as the costs of control. Weeds reduce agricultural output, and so decrease farm income and increase the cost of food to consumers. The combined annual loss to farmers and consumers in 2001-02 was estimated to lie between $3442 m and $4420 m, and to average $3927 m. About 80 per cent of this annual loss falls on farmers and 20 per cent on consumers. The annual loss of $3927 m is one half of one percent of gross domestic product and 14 per cent of the value added by agriculture to the economy. These impacts in agriculture were calculated as the change in farm income and the cost of food between the current with-weeds situation and a without-weeds scenario. The difference is the maximum benefit that could be achieved by reducing the weed population, so it represents the size and national significance of the current problem. This kind of impact estimate, and information on how the loss is distributed across sectors of the economy and across industries, is a benchmark for policy and a starting point for the decision-making process. Weeds attract at least $116.4 m of government expenditure on control, surveillance and other management activities. Of this total, $19.6 m is the expenditure on natural environments by National Parks and Wildlife Services and the National Heritage Trust. Expenditure by four of the state and territory services has increased over the last four years. Of the total, $80.8 m is expenditure by other government agencies on control, inspection, research, and extension. Data on the distribution of weeds in natural environments were not available, so the value of the decrease in outputs in natural environments due to weeds could not be estimated.

Benefit-cost analysis is a method to measure the economic desirability of an activity, project or policy to the community as a whole. Issues in the application of the method to the management of weeds in natural environments are identified from a review of past applications. Opportunities are also identified from the review and the requirements for effective benefit-cost analysis. The valuation of benefits, and the treatment of risks, appears to be major areas for development - despite the advances in these areas over recent years. There is a need for production functions to relate weed spread to the outputs of natural environments, and the relevance of partial analyses appears to warrant examination.

Weeds have a wide variety of impacts on society, the environment and the economy. Some of the economic impacts are benefits but most are costs. Combellack (1987) valued the economic cost of weeds in 1981-82 to be $2096m. New methods of weed control and new techniques of farm management have since been developed, and new weeds have invaded. Therefore, the current cost of weed impacts cannot be readily compared to those of 1981-82. This paper attempts to estimate the economic costs of weeds in agriculture across Australia. But further, it offers an economic framework to help consider the problems that weeds create, and the generation and use of information to resolve those problems. A stochastic simulation model was developed to estimate the economic impact of weeds and to particularly account for variability in the cost estimates. The total annual economic loss to Australian agriculture ranged from $3400m to $4400m, with a mean loss of $3900m.

Scotch broom is an invasive weed in many subalpine ecosystems. It often has substantial negative effects on ecosystem structure and functions. Decisions on optimal management strategies require predictions of the rates and patterns of Scotch broom spread. This paper explores the environmental and management factors that influence the rate of spread of Scotch broom in Barrington Tops National Park. The NSW National Parks and Wildlife Service (NPWS) and the then neighbouring NSW State Forests prepared aerial maps showing Scotch broom infestation in the Park for 1989,1993 and 1999. These maps were used to generate data for the current analysis. Map reference points 1km apart along the southern boundary of the 1989 area of infestation were examined and 1993 and 1999 differences from the reference points were measured. Also measured were environmental factors including natural vegetation type, natural vegetation density, soil type, slope, altitude and the presence of private property or crown land. The incidence of natural disaster, feral animal activity and NPWS management activities were also included in the analysis.