Should arthropod parasitoids and predators be subject to host range testing when used as biological control agents?

Testing of candidate biological control agents to estimate their likely field host ranges in the area of release has been part of weed biological control for several decades, with evolving techniques and goals. Similar efforts have been made less often for parasitoids and predators being introduced for arthropod biological control. Here, we review both techniques of host range testing and social objectives of such screening. We ask whether agents introduced for arthropod biological control should be subjected to host range testing before release, and if so, are methods used for estimating host ranges of herbivorous arthropods appropriate, or are different approaches needed. Current examples in which host range testing has been employed for arthropod biological control are reviewed. We conclude with recommendations concerning guiding principles about use of host range testing. We recommend modest expansion of host range testing for arthropod biological control for projects on continents. We recommend more extensive testing for projects of introduction onto islands. We note that introductions to islands could provide opportunities to gain experience in use of host range estimation for this class of organisms and to conduct post release evaluations of host ranges. We urge caution in efforts to mandate host range testing but simultaneously recommend consultation between biological control workers and insect conservationists. Copyright Kluwer Academic Publishers 1997