To gauge ultimate success or failure

Agencies funding biological control of weeds require and deserve an assessment of the impact of their invested funds in reducing the target weed. In broader terms, critical evaluation of the impact of successful releases (See What successes and failures have there been?) provides justification for further investment in biological control of weeds.

To decide on future directions of the biological control program

On average it takes ten to twelve years for a successful biological control program to reduce populations of the target weed to levels below the noxious threshold (Lawton 1984). In the meantime it is impossible to know whether any progress is being made at all if no follow up studies are undertaken after the release of control agents. The more that is known about the interactions between control agent and target species, the easier it is to make the right decisions about the future directions of the project.

Both Samea multiplicalis and Cyrtobagous salviniae were released against Salvinia molesta in Australia. The relative ineffectiveness of the former compared with the latter was established by studies conducted in Australia and only C. salviniae was recommended for introduction to other countries (Room 1986), a considerable saving in time and effort.

The timing and sequence of release of agents can have an impact on the success of those agents. Studies on the weevils Neochetina eichhorniae and N. bruchi indicate that better control is achieved when both species are released together rather then individually (Harley 1990).

To develop integrated control strategies

Once the interactions between control agent and weed are understood, possibilities for integrated control strategies may become apparent.

The sesiid Carmenta mimosa is a stem-borer of the woody weed Mimosa pigra. Originally it was thought to be a promising agent and that it would inflict heavy damage on mature trees, perhaps severely enough to cause death. However, recent observations suggest it may, in the immediate future, be more useful as an agent to use in conjunction with fire to control seedling regrowth (G. S. Farrell, unpublished data).

Outbreaks of St John's wort, Hypericum perforatum, are successfully controlled by two chrysomelids (Chrysolina quadrigemina and C. hyperici) in Australia and California (DeBach 1964). Hypericum perforatum growing in shady habitats was found to escape from Chrysolina spp. and these habitats provide refugia from which the weed may repeatedly reinvade the rest of its habitat (Lawton 1984). Finding alternative methods to control H. perforatum in these shady habitats obviously is important to the overall control of the weed.

To develop the theory and practice of weed biological control

Better biological control can only come from knowing more about why projects succeed or fail. Prediction of success and, therefore, prioritisation of potential biological control agents are notoriously difficult. By and large, the critical dynamic processes that result in control of weeds by herbivores are not understood (Lawton 1984), and there is a paucity of data in this area.

Much of our understanding of insect-plant dynamics has come from, and is being driven by, biological control of both weeds and herbivorous insect pests. Studies on the spatial scale of weed control programs are beyond the scope of most individual or academic ecologists, and so offer a unique opportunity for the analysis of insect-plant dynamics in what are effectively large-scale ecological experiments.

[ Next ]