species of exotic natural enemies have been
imported, mass-reared and released, result-
ing in successful control of many species of
pests (e.g. Greathead, 1995; Gurr and
Wratten, 2000; van Lenteren, 2000). As far as
we know, few problems have occurred con-
cerning negative effects of these releases (e.g.
Lynch and Thomas, 2000; Lockwood et al.,
2001; Lynch et al., 2001). The introduction of
natural enemies has largely been an empirical
activity and has depended on the knowledge
and insight of the biological control practi-
tioner. Recently, a more scientific approach of
evaluation of natural enemies before intro-
duction is advocated and applied, with the
aims: (i) to obtain insight into what consti-
tutes an efficient natural enemy; (ii) to reduce
research costs; and (iii) to limit risks of
imports (van Lenteren, 1993).
As each organism that is introduced into a
new region, be it for augmentative releases or
for permanent biological control, may become
established, extreme care should be exerted
during the evaluation phase. The more than
5000 introductions of exotic natural enemies
that have taken place during the past 100
years did not lead to environmental problems
when the procedure of selection, importation
and release was carefully applied (Greathead,
1995). In general, biological control workers
prefer to import specialist natural enemies,
attacking only one or a few related phy-
tophagous species. The few documented
problems caused by introduced organisms
concern attempts to control pests other than
insects, such as vertebrates and snails. The
introduction of a generalist vertebrate preda-
tor (mongoose, Herpestessp.) for rodent con-
trol in the 19th century resulted in economic
and ecological problems, as this opportunist
predator attacked all prey it could easily han-
dle and contributed to the extinction of some
endemic species. The introduction of other
large generalist predators such as the giant
toad (Bufo marinus) and predatory snails, had
unfortunate side-effects (Greathead, 1995),
and the liberation of such organisms has
never been recommended by researchers of
biological control (e.g. van Lenteren, 2000,
2001). To a large extent, only those organisms
are released that experience has shown to
have restricted host ranges and not to exhibit
switches in host ranges (Aeschlimann, 1995).
The use of biological control has
increased considerably during the past
decades, as it offers a sustainable, economi-
cal and environmentally attractive alterna-
tive to chemical pest control. Arthropod
biological practitioners are, however, con-
fronted with criticism from environmental-
ists because they are fearful that the released
natural enemies may attack: (i) beneficial
non-target organisms, such as pollinators or
other natural enemies; (ii) rare/endangered
insects, such as butterflies; and (iii) other
non-target organisms. This results in a
bizarre conflict situation: on the one hand,
biological control offers a very powerful
option for reducing the many potentially
negative environmental and human health
effects of chemical pest control, while, on the
other hand – though not found until now –
the arthropod natural enemies may influence
ecosystems in unwanted ways. The very
positive effects of releases of natural enemies
are in strong contrast with the effects of
other types of introductions. Many intended
introductions of animals and plants and
many more unintended introductions of pest
organisms have led to disasters worldwide
(for examples, see Hokkanen and Lynch,
1995; van Lenteren, 1995).
The potential negative effects of arthro-
pod biological control might be prevented
when, as in weed biological control pro-
grammes (e.g. Wapshere, 1974; Blossey, 1995;
Lonsdale et al., 2001), not only is the effect on
the target species determined, but also the
effect on indigenous non-target species (van
Lenteren, 1986, 1995; Blossey, 1995). Up until
now, indigenous non-target species testing as
part of a prerelease evaluation programme is
rarely applied (van Lenteren and Woets,
1988; Waage, 1997), but there are a few exam-
ples where non-target species testing has
been applied properly (e.g. Barratt et al.,
1999; Neuenschwander and Markham, 2001).
Another way to reduce the risks of release of
exotic natural enemies would be to first eval-
uate native natural enemies. Although this
seems a logical approach, the experience of
the last four decades shows that many exotic
natural enemies were imported and released
without preceding testing of native biocon-
trol agents (e.g. van Lenteren, 2000).192 J.C. van Lenteren et al.