Below we provide a set of guidelines for selecting plants for a host test list. We use the weed Mimosa pigra (Mimosaceae) to illustrate each of these guidelines. The resulting test list is presented in a Table on the following page.

1. Start with the target species and add varieties of the target weed if any are present in the introduced range.

No varieties of M. pigra are known in Australia; the species is very uniform here.

2. Add species from the same genus.

There are no native or economically important species in the genus Mimosa in Australia or Southeast Asia but there are two introduced species which have weedy status, M. pudica and M. invisa. It is important to include these in the list as they may give an indication whether the agent is specific to the species M. pigra or to the genus Mimosa.

3. Add species from genera in the same tribe.

The genus Mimosa belongs to the tribe Mimosae. A representative species of all six other genera in this tribe was included on the list. All species in the genus Neptunia are added to the list for two reasons. First, observations in the native range of Mimosa pigra in Mexico have shown that the moth Neurostrota gunniella (Gracillariidae) is occasionally collected from Neptunia plena. Second, during host range testing of other agents for M. pigra, several agents showed some acceptance of Neptunia for oviposition and feeding.

4. Add species from genera in closely related tribes.

The closely related tribe Acaciae, includes the genus Acacia. This genus presents special problems as there are more than 800 native species present in Australia. In a case such as this, the question arises as to how many Acacia species should be tested and how does one decide which species should be tested. One approach is to select species from different subgenera according to whether the species occur in the actual or potential range of the weed. For example, Pedley (1978) and Maslin and Pedley (1988) divided the Australian acacias into three subgenera with seven sections in one subgenera. Species in one subgenus and one section of another subgenus are environmentally isolated from the potential range of M. pigra and are considered to be of less importance than Acacia species within the colonising range of M. pigra. The outcome of the selection process was to select a total of 20 species of Acacia representing all sections within the subgenera, and within the potential range of M. pigra in Australia. The number of species in each group was related to the size of that group, e.g. Juliforae and Phyllodineae are bigger groups that Botrycephalae.

Representative species from the Ingeae, the third tribe in the family were then added to the list.

5. Add species from representative genera in closely related families.

The families Fabaceae and Caesalpiniaceae are very closely related to Mimosaceae. These three families are considered by some to be three subfamilies within the same family. These families contain native species and a number of important introduced pasture and crop plants. Not all can be tested and it is preferable to test a species from each genus rather than several species from a few genera. Also, representatives from each of the tribes in these families were selected.

6. Add species that are chemically similar to the target weed if this information is known.

No evidence was found for including other species under this criterion.

7. Add a selection of plants which are economically important or of interest to conservation in the target country.

These species are usually crop or pasture species which are geographically associated with the target weed but are not closely related taxonomically. They are tested to allay the fears of farmers, conservationists and bureaucrats. It is desirable that the list does not exceed 10 species.

8. Add any host plant records of the biological control agent.

If the potential biological control agent has been collected from any other plant species, or is known from the literature to attack other species, then these species should be added to the list. These additions will normally apply only to specific agents and not to the basic list for the target weed. No examples of such additions to the list are available for M. pigra.

9. Add any host plant records of insects in the same genus as the biological control agent.

The Lepidoptera Ithome sp. (Cosmopterigidae) is being assessed as a potential agent for M. pigra. Its relative, Ithome lassula is known to attack Leucaena leucocephala. Therefore more species of Leucaena were added to the list.

[ Back ] [ Next ]