Richard Hobbs
CSIRO, Wildlife and Ecology

We need a mechanism to recognize, in the early stages, what species are going to be successful invaders. For example, Mimosa pigra sat around in a "lag phase" for 100 years, from 1880- 1980, and then its population exploded exponentially. Actually, it was increasing within certain areas for a couple of decades, but on the large scale, there was a sudden and dramatic leap. Do plant species build up "cumulative populations" during lag phases and then "triggers" allow for the explosion? Are lag phases a significant phenomenon or is it just the way we monitor and measure the growth that makes it look like an explosion, when really it was growing exponentially within the original patch all the while? This is partially a spatial question - is it growing exponentially within the one patch?

There seems to be consensus among this group that any time you see an exotic plant that has established a new population, and that species has a history of being a successful invader elsewhere, you should move to control or eradicate it before it becomes a major problem.

Australia has 222 species of noxious weeds and another 1500 other non-indigenous species to deal with. They can't respond to everything with a chainsaw. They need to monitor growth within the original patch, and if they see exponential growth, they can move to control it. Some feel they should control it sooner ("monitor it with a chainsaw"). What if the original patch was in a suboptimal environment and so may not have grown exponentially for that reason? What are the triggers for exponential growth - chance flood events, etc.?

The message to policymakers is that you need to monitor more carefully and manage invasions earlier. You should look for exponential growth within a population or the establishment of a second population elsewhere. Rapid deployment of resources is key to the ability to control an invasion.

Characteristics of an area that is susceptible to invasion:

1. In general, disturbed areas are more invasible than undisturbed areas. There is a difference between natural and human disturbances, as well as differences in degrees of disturbance. (For example, grazed land tends to be invaded much more easily than ungrazed land.) There are also synergisms among different kinds of disturbances that allow invasions to take hold.
2. Fragmentation of natural habitat makes an area vulnerable to invasion; edges of natural habitats are especially vulnerable.
   
3. Human activities modifying ecosystems cause most of the problems. There are also public perception problems - if people like a particular plant, they don't want to get rid of it.

In order to prioritize limited resources, we can characterize areas as high or low value, and as high or low disturbance, yielding four boxes. The strategy for each of these four boxes is different:

High value, low disturbance areas: protect and keep invaders out;
High value, high disturbance: try to control invaders;
Low value, low disturbance: do nothing;
Low value, high disturbance: let them go, we can't do everything with limited resources.

"Shocking, how easily we have lost a world that seemed so stable and safe."

Management actions required, in this order are:

• quarantine,
• prevention,
• detection and early control,
• eradication,
• integrated ecosystem management.

1. The current emphasis is on control measures implemented once a species has become a major problem. An integrated program of prevention, detection, early control and ecosystem management carried out at all stages of the invasion process is required.

2. Changes in emphasis from control to prevention, and from control of individual species to ecosystem management are required.

3. Research on plant invasions needs to be directed at elucidating the linkages between disturbances and invasions and at sorting out causal relationships. Research is also needed on the ecosystem impacts of invasions.

4. Much of the plant invasion problem stems from socioeconomic, rather than ecological, factors. Attempts to treat weed invasion problems will fail unless the underlying causes of the problem are identified and dealt with.

5. A rational framework for setting management objectives and priorities has to be adopted. This can be based on the relative value of different areas (in terms of conservation or production) and the relative likelihood of successful prevention or control. Weed problems are too numerous, too extensive, and too pervasive to allow dissipative use of resources through ad hoc decisions.

Plant invasions are a major threat to biodiversity worldwide. Unless conservation biologists, land managers, and policy makers respond to the problem now, we face a world dominated by a small aggressive fraction of the world's plant species.

Discussion Points

Epidemiology models for spread of disease can be used to look at spread and control strategies for exotic plants. Computer models/games can be used as tools to both understand and analyze these phenomena better and to educate people about the realities of invasions. Fire spread models are a good analogy also.

There is a potential fallacy of monitoring until you figure out what to do - the spread is tremendous while you wait for the results of further study, and it is also probably more cost effective to eradicate than to study.

It seems like many invasions are caused by serendipity. For example, the gypsy moth arrived at the North Carolina coast on July 4; since it was a holiday, proper inspection was delayed by two days, leading to an $800 million problem.