The low vigour of the genetically engi-
neered sexing strain of medfly highlights an
important issue. Traditional captive-reared
colonies are derived from natural popula-
tions, but it has proved difficult to maintain
colonies that have acceptable field perfor-
mance (quality). The problem of maintaining
the quality of genetically engineered strains
is much greater because the beneficial
genetic changes are initially incorporated
into inbred, laboratory-adapted stocks.
Outcrossing the stocks into a higher-quality
genetic background can be quite compli-
cated. For example, in the medfly sexing
strain, one autosome carries a Y translocation
and its homologue carries two recessive
mutations. The integrity of these arrange-
ments must be maintained in any pro-
gramme of outcrossing.
Significant effort should be made to
ensure that genetically engineered strains
are put in the highest-quality genetic back-
grounds. Probably the best solution is to
build up a genetic reservoir of indepen-
dently derived isofemale lines from the
original engineered strain. These isofemale
lines can be maintained indefinitely and
periodically combined to reinitiate the rear-
ing colony.
Classical Biological Control
The methodologies discussed above for opti-
mizing the trade-off between quality and
quantity apply primarily when the continu-
ous augmentative release of a natural enemy
requires long-term captive rearing. However,
this trade-off is also important when the goal
is to establish a natural enemy for self-sus-
taining biological control. Domestication and
loss of genetic variability during the initial
rearing phase could contribute to the failure
of an introduction. Potential causes of failure
include inbreeding depression, lack of adap-
tation of released individuals to their new
natural environment and lack of ability of a
newly established population to adapt to
changes in their environment.
Inbreeding is the random loss of genetic
variation and results in a lack of individual
genetic heterozygosity. It can also result in
inbreeding depression. For example, the
extinction risk facing natural populations of
a northern European butterfly increases with
their level of inbreeding (Saccheri et al.,
1998). However, biological control agents
rarely exhibit significant inbreeding depres-
sion in the laboratory (Roush, 1990), but this
generality must be treated with some caution
since it is now recognized that inbreeding
depression can increase dramatically when
individuals are subjected to stressful condi-
tions (Bijlsma et al., 2000).
A lack of adaptation of released individu-
als to their new natural environment is likely
to have a profound influence on the proba-
bility of successful colonization (McDonald,
1976; Tauber and Tauber, 1986; Roderick,
1992). For this reason, even when laboratory
adaptation can be avoided, it is still impor-
tant to plan carefully how the founding pop-
ulation is to be established (e.g. geographical
origins; see González and Gilstrap, 1992).
Ideally, the released population should both
be preadapted and have the potential for fur-
ther adaptation to the environment of the
release site. As noted earlier, colony
improvement for such traits as insecticide
resistance can be important and may be
amenable to genetic engineering, but the
potential value of in situevolution of more
complex adaptations after release should not
be underestimated. Releasing genetically
depauperate stocks initiated from a handful
of wild-caught ancestors does not guarantee
failure, but it can be expected to minimize
the chance of success.
The importance of postrelease adaptation
is difficult to evaluate experimentally and
such evaluation has rarely been attempted
(Roderick, 1992). However, interest in the
role of genetic variability in the adaptation
and persistence of populations in the face of
environmental fluctuation has been stimu-
lated by the growth of conservation genetics
since the 1980s (Soulé, 1987; Nunney, 2000).
Many of the issues faced in classical biologi-
cal control have parallels in conservation
biology. Perhaps the most obvious is the
necessity to re-establish populations of
threatened species using individuals taken
from other areas (see Hedrick, 2001).Managing Captive Populations 83