Coleman and Edman (1988) suggested that the number of landings on the
lesions was not significantly greater than on an uninfected mouse and that
the selection of landing sites was essentially random. The implication is
that, once the lesion has been encountered, the sandfly will choose to feed
at that site rather than select another site. However, lesions on infected
animals tend to occur at those sites that are preferentially probed in any
case, i.e. the exposed areas of skin on the ears, nose and paws.
It has been suggested (Davies, 1990; Molooet al., 2000; Taylor and
Hurd, 2001) that the epidemiological consequences of enhanced vector
feeding on the parasitized host may be increased biting and thus
increased transmission. However, once the parasite has developed within
the vector and is at an infective stage, further parasitic manipulations may
ensure enhanced host contact.Landing and probing by a parasitized vector
When the vector is infected with the parasite, the situation changes.
Laboratory studies have shown that, in A. aegypti mosquitoes, P.
gallinaciumsporozoites invade the salivary gland and reduce apyrase
secretion. Apyrase is an enzyme that inhibits the aggregation of platelets
and has been found in all haematophagous insects so far examined (Law
et al., 1992). Thus, confronted with the host blood-vessel wounds healing
more quickly because of reduced amounts of apyrase and the consequent
increased difficulty in feeding, the insect spends more time probing
(Rossignolet al., 1984, 1986; Ribeiroet al., 1985). There is no reduction in
saliva production because of infection and thus, as the malaria parasite is
transmitted in mosquito saliva during probing and feeding, transmission
is not reduced (Griffiths and Gordon, 1952) and, indeed, is likely to be
enhanced. Wekesaet al. (1992) also showed an increase in the number
of probes and in the likelihood of probing beginning whenAnopheles
mosquitoes in western Kenya were infected withP. falciparum.Ina
similar set of experiments,A. aegyptiinfected withP. gallinaceumprobe
for longer and more often (Rossignolet al., 1986). The effect of increased
probing is increased transmission (Kelly and Edman, 1992) and the ability
of a smaller number of vectors to maintain infection within a population
(Koellaet al., 1998a).
Recent studies have shown that parasitic modifications of mosquito
behaviour may be more closely related to the developmental stage of
the parasite than hitherto seen and that changes in vector behaviour may
be specifically timed to maximize sporozoite transmission.A. stephensi
infected withP. yoelii nigeriensissporozoites (the infective stage) were
significantly more persistent in their feeding attempts than uninfected
mosquitoes. However, when infected with oocysts (an uninfective,
developmental stage), the mosquitoes were significantly less persistent
than those that were uninfected and thereby possibly were able to reduce
dangerous host contact (Andersonet al., 1999).
Results of work conducted in the field reveal generally similar
changes in mosquito behaviour due to parasitic infection. Koella andParasite Manipulation of Vector Behaviour 271