correlation is found between laboratory
results and flight tests or field performance.
Van Lenteren and Tommasini (1999) mention
several attributes of field populations of nat-
ural enemies that might change once the lat-
ter are introduced into the laboratory under
mass-rearing conditions. Searching efficiency
and dispersal characteristics (e.g. flight
and/or walking behaviour) are among the
key attributes that might be influenced by
the mass-rearing conditions and might there-
fore negatively affect the field performance
of the natural enemy.
A list of criteria for product control of the
commercially available predatory mite
Phytoseiulus persimilis Athias-Henriot was
initially included in the protocols developed
by the International Organization for
Biological Control (IOBC) global working
group ‘Quality Control of Mass-reared
Arthropods’ (van Lenteren and Steinberg,
1991). The updated measure of fecundity for
P. persimilisdetermined in a laboratory test
should be a minimum of ten eggs per female
during a 5-day test (van Lenteren, 1998).
Penn et al.(1998) criticize the idea of labora-
tory tests, stating that, although it can be
helpful in determining predator fitness, test-
ing individual predatory mites is time-con-
suming and does not necessarily provide
adequate information on the performance of
the predators in the field. For instance, P. per-
similismites applied directly to ‘hot spots’
where prey is readily available do not
require the same attributes as predatory
mites distributed in a crop where prey is
more dispersed. Thus product-control stan-
dards for a natural enemy must allow flexi-
bility according to the conditions of its use
and the expected results (Penn et al., 1998).
Here we report on the initiative taken by
a commercial natural-enemy producer to
quantify the relationship between the earlier-
developed laboratory fecundity test and a
newly developed greenhouse cage test,
where dispersal of the natural enemy is pos-
sible. P. persimilis was chosen as a model
because of the key role it plays in many inte-
grated pest management (IPM)/biocontrol
programmes and hence its high economic
value in commercial augmentative biocon-
trol worldwide. The greenhouse cage test is
described in detail in this chapter. The details
of the laboratory fecundity test for P. persim-
iliscan be found in the last version of the
product-control guidelines established by
the IOBC/European Community (EC)
Concerted Action Group on Quality Control
(van Lenteren, 1998) and in Chapter 19.Materials and Methods
PlantsCucumber plants were grown in a ‘perlight’
medium in buckets under greenhouse condi-
tions. The plants reached up to 15 true leaves
and were trellised in a normal procedure.
They were kept completely separate from
one another, both to avoid migration of the
mites from one plant (= replicate) to another
(see below) and also to provide enough
space for the person to check the leaves and
move freely between the plants. Tanglefoot
glue was smeared on the trellising cords
between the plants as another measure to
ensure that no mites could migrate from one
plant to another.Infestation by spider mitesA piece of brown bean leaf was taken from
the greenhouse mass-production unit,
checked thoroughly for absence of predatory
mites and counted for presence of c. 25
mobile stages of the two-spotted spider mite
Tetranychus urticaeKoch. The infested leaf
was placed on a cucumber leaf at the top or
at the lower part of the plant (according to
the treatment, see below). Spider-mite infes-
tation was allowed to develop for a period of
48 h, at the end of which a colony consisting
of a few dozens of mobile stages as well as
eggs of spider mites was covering c. 50% of
the leaf underside.Introduction of the predatory mitesForty-eight hours after initial infestation, five
females of P. persimilis, randomly chosen
from a 7-day-old storage container (after har-226 S. Steinberg and H. Cain