Environmental Engineering Reference
In-Depth Information
non-target eff ects in only 1.7% of all documented introductions, and reported
impacts are only relatively minor. However, there may have been impacts that
were not quantifi ed or reported, particularly in the very early history of biological
control (Lynch and h omas 2000).
Pest risk assessments and management tools have been developed and refi ned
over the past 100 years, almost exclusively by entomologists engaged in invasive
plant biological control. A main part of pest risk assessment, host specifi city
screening in invasive plant classical biological control, is founded on a centri-
fugal phylogenetic method of host-range testing, proposed in the 1970s (see
Wapshere 1975); this replaced an earlier methodology which was more focused
on the threat to crop plants in the release area rather than on genetic related-
ness. Subsequently, this was adopted for and adapted to the screening of plant
pathogens and further modifi ed for insect agents in order to reduce the chances
of rejection. Nevertheless, there are currently even more refi nements and intro-
spection of the agent-selection procedures in order to meet the more stringent
demands of the increasing risk-adverse societies. Hopefully, this should dispel
doubts and encourage greater support. Test requirements for fungal pathogens
of invasive plants diff er considerably from the relatively simplistic choice/no
choice tests conducted on potential insect agents, involving additional criteria
such as internal analyses of inoculated test plants (Evans 2000), in order to
better interpret host-pathogen relationships and the resistance mechanisms
deployed.
Host-specifi city screening of biological control agents of arthropod invasive
species has not been a feature of biological control projects against these targets
until quite recently. Indeed, the older, standard biological control texts, which deal
primarily with control of arthropod pests, make no mention of safety or risk assess-
ment (e.g. DeBach, 1974). h e numerous predators, parasites, and parasitoids
moved around the world for classical biological control of invasive species were
rarely, if ever, tested for specifi city: the assumption being that they were part of the
indigenous natural enemy guild of the target, and, therefore, inherently specifi c
and safe. And natural enemies, both arthropods and pathogens of arthropod inva-
sive species, are still sometimes moved between continents with little consideration
for safety or quarantine issues. h ere are also numerous instances where exotic
strains of entomopathogens, usually as bio-pesticide products, have been freely
exchanged for both laboratory and fi eld-based trials against invasive species. But
the situation is changing and there has been much research during the last decade
on how to assess the host-specifi city of potential insect and fungal agents of arthro-
pods. h e topic is more di cult than for agents of invasive plants because the main
groups of agents used for arthropods, parasitoids, and predators, have complicated
behaviours and ecology. For example, parasitoids respond to two trophic levels, the
host and the plants of the host. Nonetheless, protocols for testing have been sug-
gested which recommend criteria for drawing a list of test species (see, for example,
reviews in Bigler
et al
. 2006).
