In both designs, a control plant consisting of the target weed, must be included. This tests whether the insects used were in a suitable condition for oviposition and feeding. It also provides
'baseline' data, or estimates of normal numbers of eggs laid and feeds inflicted by a given number of insects in a given period. Damage to other plant species can then be compared to these baselines.
Choice tests
The advantages of choice tests are that they are more natural (in
nature, insects are constantly faced with making choices). Choice tests
are also more efficient as several plant species can be processed
simultaneously. However, presence of the normal host may influence
behaviour on other species.
No-choice tests
No-choice tests are considered more rigorous than choice tests. An insect may not feed or lay eggs in a choice test because it preferred the normal host. Rejection in a no-choice test, however, indicates that the insect will not accept the plant for feeding or laying even though it has no alternative. No-choice tests may follow choice tests for those plant species accepted in choice tests. Alternatively, no-choice tests are done in addition to choice tests, even when no attack occurred in the choice test. A third option is to test some plants using choice tests and others using no-choice. In this case, the plants considered more at risk should be tested using the no-choice method.
No-choice tests may be serial (also called sequential) or parallel. In serial tests a group of insects is alternated between the target weed and test plants. In parallel tests, the group of insects is divided into two subgroups, one of which is placed on the target weed and the other on a test plant.
Starvation tests
These tests are no-choice adult and larvae feeding tests which are continued until the insects die. They ensure that insects will not begin feeding after extended periods of food deprivation.
New test designs
Tests are currently being developed that combine the advantages of choice and no-choice tests. One approach is to conduct choice tests but include a limited amount of host plant material. This ensures that the insects consume all the host plant material, or use all available oviposition sites, within the duration of the trial. The insects are then forced to look elsewhere and thus the test becomes similar to a no-choice test. Another approach is to alternate the normal host with a choice of test plant species. The time spent with the normal host gives information on the fecundity of the insects. The test plants are then assessed for their
acceptability in a choice design.
[ Back ] [ Next ]
Tim Heard
