1999), the use of antibiotics in artificial diets
does not always prevent the production of
entomophagous insects harbouring sym-
bionts belonging to alpha Proteobacteria
(Rickettsiaceae). The antibiotics used in
diets for preventing bacterial contamina-
tion, mainly by gamma Proteobacteria, are
not effective in removing endosymbiotic
Wolbachiafrom infected strains (Stouthamer,
1990). In fact, penicillin and streptomycin
are used for bacterial control and tetracy-
cline and rifampicin for Wolbachiaelimina-
tion. Pure female lines with a Wolbachia
infection could be an advantage for biologi-
cal control, because only females kill pest
insects. But Wolbachiainfections may mod-
ify the fecundities (see below and Chapter
8). The presence or absence of a symbiont
could be a key quality criterion for some
insects. Symbionts transferred from one
species to another could be used to improve
the quality of entomophages released for
biological control (Grenier et al., 1998), but
for a full discussion of this topic see also
Chapter 8.
In general, sex ratios of adult predators
obtained on artificial diets are not different
from those observed in prey-fed controls.
Fecundity and fertilityFecundity is influenced by female size (see
above) and also by the quality and quantity
of the adult food (Bernal et al., 1999). The
fecundity of Trichogrammafemales produced
on E. kuehniellaeggs is proposed as a quality
control parameter (Dutton et al., 1996). In
vitro-reared females of T. australicum pro-
duced significantly more progeny and para-
sitized more host eggs than the females
reared on natural or factitious hosts
(Nurindah et al., 1997). Wolbachiasymbionts
inducing thelytokous reproduction could
also modify the fecundity in T. pretiosum
even when grown in artificial diet (Grenier et
al., 2002). The fecundity of the tachinid fly E.
larvarum is slightly, but not significantly,
higher in individuals reared in artificial diets
than in vivo(Dindo et al., 1999). D. introita
females laid significantly fewer eggs when
reared on artificial diet than when reared on
live hosts (Greany and Carpenter, 1998).
Although several diets have been devel-
oped for predatory insects that have yielded
only slightly inferior to similar developmen-
tal success, most often the fecundity of
adults maintained on such diets is clearly
inferior to that on insect prey. Fecundity
appears to be less of an issue when the adult
stage does not feed at all, only does so to a
limited extent or has different feeding habits
from those of the larval stage (e.g. honeydew
feeders). For instance, artificial larval diets
for Chrysoperlaspp. have been reported to
yield adults with similar reproductive capac-
ities to those obtained on flour-moth eggs
(Hassan and Hagen, 1978; Cohen and Smith,
1998). However, adult lacewings can be
maintained on a much simpler diet than that
for larval feeding, usually consisting of
water, sugar and yeast. Only rarely have arti-
ficial diets yielded similar total fecundities
when the adult stage also has to be main-
tained on artificial food.
Arijs and De Clercq (2001) found that
females of the anthocorid O. laevigatusfed
different bovine meat and liver diets had a
similar fecundity to those fed E. kuehniella
eggs (120–200 eggs per female). Similar
results were reported for the mirid Dicyphus
tamaninii (Iriarte and Castañe, 2001). De
Clercq et al.(1998a) reported that the fecun-
dity of P. maculiventrison a meat diet (c. 550
eggs per female) was similar to that on T.
molitorlarvae (c. 460 eggs per female), but
lower than that on G. mellonellalarvae (c. 1000
eggs per female). The fecundity of other
arthropod predators reared on artificial diets
is reportedly half (Kennett and Hamai, 1980;
Abou-Awad et al., 1992; De Clercq and
Degheele, 1992) to less than one-tenth
(Kennett and Hamai, 1980; Chain-Ing et al.,
1993; Hattingh and Samways, 1993; Rojas et
al., 2000) of that of predators fed natural or
factitious food. Very few workers have calcu-
lated demographic statistics to evaluate an
insect diet. Wittmeyer and Coudron (2001)
estimated life–fertility-table parameters for P.
maculiventrisreared on artificial diet or on
Trichoplusia nilarvae. Net reproductive rate
(RO) and intrinsic rate of increase (rm) were
significantly lower, even when the predator
was only partially reared on the artificial
food.122 S. Grenier and P. De Clercq