Detailed host specificity tests must be undertaken to ascertain the host range of an insect species or plant pathogen. However, many biological control workers are loath to begin testing species which belong to families and orders which are considered unlikely to be host specific. For example, no mealybugs or scale insects have ever been deliberately introduced to control a non-cactus weed because of doubts about their host specificity. In contrast, Shorthouse (1990, citing Harris 1984) states that approximately 25% of the agents released in Canada by the end of 1983 were gall insects, not because they were thought to necessarily be very effective, but because of their assured high degree of host specificity.

The rationale for consideration of taxa to be host-specific or not has often been based more on precedence and personal bias than science. For example, prior to the well-documented devastation of leucaena by the host specific leucaena psyllid, non-galling psyllids were considered unlikely to have a narrow host plant range (R.E. McFadyen pers. comm. 1992). It is now recognised that most psyllids are very host specific and they are being increasingly used in weed biological control (D. Hollis pers. comm. 1991). As a second example, we would never have known the spectacular control of Salvinia molesta or Eichhornia crassipes if the status quo had been maintained and fish and snails continued to be trialed as biological control agents because, and I quote Wilson (1964), 'No insects have yet been used for the biological control of aquatic weeds, ... it may be that in the fresh-water environment the relatively small numbers of species of plants and phytophagous insects, and perhaps the domination of this environment by fish, have caused in aquatic phytophagous insects a level of host specialisation much lower than that occurs in the species-rich terrestrial environment'. The origin of this fallacy can be traced to the non-biological control theoretical literature (Brues 1946).

Plant pathogens are usually very host specific but have been used less frequently than insects in weed biological control because of a fear amongst some biologists and bureaucrats that they might 'host shift' and attack non-target plant species (Harris 1973).

[ Back ] [ Next ]