Initially, only plants of economic importance were tested. Test lists
included many plants quite unrelated to the host weed, and which were most
unlikely to be attacked. In 1968, Harris and Zwolfer (1968) proposed that
testing should concentrate on plants related to the known host, and should
aim to determine the range of plants acceptable to the insect rather than
demonstrate that certain valued plants were immune to attack. This
approach was seen to be sensible, and test lists progressively changed to
focussed lists of plants botanically or chemically related to the host
weed.
The next major change was the growing concern over environmental
issues, and possible damage to native plants, and from the 1980s these
began to be included in test lists. In many cases, this has led to
conflict over whether feeding on native plants of no economic significance
is a sufficient reason to block release of a potentially valuable agent of
a major agricultural or environmental weed. Discussion on these issues
continues (McEvoy
1996; McFadyen
1998).
