However, the effect may be inferred from experiments designed to deter-
mine the effect of infection on activity levels, although one significant
problem with this approach is that human odour from the experimenter
was unlikely to have been excluded from these experiments and therefore
its potentially confounding influence cannot be ignored. In laboratory
experiments, Berryet al. (1988) showed thatAedes trivittatusinfected
withDirofilaria immitishad greatly increased flight activity.D. immitisis
the common heart-worm of dogs, cats, foxes and wolves in the USA,
where it occurs most frequently in the southern states.Aedes,Culexand
Anophelesspecies of mosquitoes are all competent vectors of this parasite
andA. trivittatushas been found to be naturally infected withD. immitis.
Within 24–36 h after taking an infected blood meal, larvae are found in
the Malpighian tubules of the mosquito vector. They reside there and
develop into the infective L3 stage, which then migrate to the haemocoel.
Eleven to 12 days postinfection, they are found in and among the fat
bodies in the lower half of the thorax, from where they finally migrate
to the mouth-parts. Berry et al. (1988) fed A. trivittatuswith either
uninfected blood or with blood infected withD. immitis3 days prior
to the start of the experiment and then monitored activity patterns for
15 days in acoustic chambers. During this time, mosquitoes would have
been likely to be conducting appetitive search behaviour. Flight activity
increased within and outside normal flight-activity periods as the para-
sitic infection developed and from 8 days postinfection was significantly
greater than in the uninfected controls. A greater number of flights of
shorter duration were made, a behavioural pattern that was particularly
marked in older mosquitoes with a heavier parasite burden. Flight activity
was significantly suppressed at specific times when parasites were
believed to be moulting or migrating to the mouth-parts. It is interesting
to note that the vector’s behaviour was not modified for 8 days after
infection and was dependent on the parasitic burden. Increased flight
activity as the parasite approached maturity could increase the vector’s
chance of encountering a host and thus improve the chances of the
parasite being successfully transmitted. Although there may be several
explanations to account for these observations, the increased activity is in
line with predictions made by Koellaet al. (1998a).
Although general activity levels cannot be specifically related to
ranging behaviour, it is interesting to note that Ryan (1984) observed that
Glossina longipalpisin the field and naturally infected with trypano-
somes were probably more active than uninfected flies.
In contrast, whenD. immitiswas present in a non-natural vector,
A. aegyptiflight activity was reduced (Berryet al., 1987). Similar loss
of flight activity was seen whenA. aegypti(a non-natural vector) was
infected withBrugia pahangi(Berryet al., 1986; Rowland and Lindsay,
1986) (see also review by Moore, 1993). Loss of flight activity was also
seen whenAnopheles stephensiwas infected with the rodent malaria
Plasmodium yoelii(Rowland and Boersma, 1988). AsA. stephensiis not a
natural vector ofP. yoelii, perhaps this illustrates that there is a conflictParasite Manipulation of Vector Behaviour 265