Mechanical transmission is, in effect, passive and no opportunity
is created for the parasite or pathogen to interact with the vector. For
mechanical passage via contaminated mouth-parts to be successful, rapid
biting of successive new hosts must occur, because vector-transmitted
parasites are short-lived in the environment. This mode of transmission is
probably quite rare. Examples include parasites transmitted by fleas, such
as the myxoma virus, which, in Britain, is carried from rabbit to rabbit by
the fleaSpilopsyllus cuniculi. In addition, parasite density in the host’s
blood meal must be great for the tiny volumes of blood that remain on
most mouth-parts to contain sufficient parasites to infect a new host.
In this respect, tabanid flies, with their spongy mouth-parts, which soak
up blood, could be expected to be capable of mechanical transmission.
Tabanus fuscicostatustransmitsTrypanosoma evansi, the causative agent
of surra in horses and camels, in this manner.
In contrast, if the vector also acts as a ‘host’ to parasite life stages,
complex interactions between arthropod and parasite are likely to evolve.
Evolutionary pressure will drive parasites to enhance their transmission
success. In the case of vector transmission, it is reasonable to assume that
the basic case reproduction number,R 0 (the number of new infections
that arise from a single current infection), in a defined population of
susceptibles (Macdonald, 1957) will be increased if more vector–host
contact occurs and if transfer between vector and host becomes more
efficient. This is particularly so as transmission dynamics of vector-borne
diseases are very different from those of directly transmitted diseases,
becauseR 0 can be an order of magnitude greater. The basic equations
used to modelR 0 for vector-transmitted diseases incorporate the vectorial
capacity (C) of the bloodsucking insect population. Vectorial capacity is
the daily rate at which future inoculations arise from a current infective
case (Garrett-Jones, 1964). Thus, in these cases,R 0 =C/r(whereris the
rate of recovery from infectiousness) (Garrett-Jones and Shidrawi, 1969).
An estimate ofCis derived from eight components (Dye, 1990); however,
for each of these components assumptions are made that do not always
hold (Dye, 1990; Rogers and Packer, 1993). For example, host selection by
vectors is not always random and may be influenced by the disease status
of the host (Day and Edman, 1983) and mosquito biting behaviour changes
when infected with malaria (Andersonet al., 1999). It is almost inevitable
that adaptive changes that enhance parasite transmission will cause
changes to vector life-history traits, because the interests of the parasite
are intricately linked with that of the vector via blood feeding (Hurdet al.,
1995; Klowden, 1995); thusCwill be underestimated.
Blood feeding is essential to the fitness of the vector, as it provides the
female with a proteinaceous meal for egg production, and it is essential to
the parasite, as host contact provides an opportunity for transmission. But
blood feeding can be risky, due to host defensive behaviour (Day and
Edman, 1984a; Randolphet al., 1992). Increased host contact increases the
chances of vector mortality. Thus many vectors face trade-offs in terms
of feeding strategies between carbohydrate (nectar) meals, which allow260 J.G.C. Hamilton and H. Hurd