Detailed studies are needed to determine which agents have the greatest potential as candidates for introduction against the weed in the introduced range. They can aid in prioritising agents, especially if the list of potential candidates is extensive. Selection is largely on host specificity rather than a prediction of their effectiveness, though there will be a bias towards those agents which have a narrow host range and are very damaging.

Studies should provide information on seasonal abundance, distribution, biology and provide an indication of host range. The latter may be achieved through observation and by screening a few plants closely related to the weedy species.

If sufficient resources exist, field host-testing in the native range of the weed should be undertaken. Behavioural studies undertaken in natural conditions, minimise restrictions on the testing environment (e.g. cage trials) that reduce the number of stimuli available for agents to find their hosts (Cullen 1990). Such studies should be conducted throughout the year, and involve intensive searching of both the weed species and all other potential hosts along transects or in a defined area where potential agents occur. Other plants which are not present, but are required for host testing, may be placed/planted in the survey areas. The abundance and seasonality of these agents, and the damage they cause, is then determined for each plant species. Field host-testing can eliminate non-specific agents at an early stage, thus reducing unnecessary and time consuming laboratory studies. The effect of the biological control agent on the weed should be documented.

Other studies may include insecticidal exclusion experiments where similar plants of the target weed are divided into two groups under natural conditions. One group is sprayed with insecticide, excluding insect attack, while the other is untreated and therefore exposed to insect attack. The growth rates of each group are then compared to gauge the impact of insect herbivores.

[ Back ] [ Next ]