House Screening - Biological and Environmental Control of Disease Vectors - page 125

this behaviour and therefore infl uence the
ei cacy of house screening are the degrees of
endophagy (indoor feeding), anthropophily
(feeding on humans), fl ight height, behaviour on
encountering a vertical surface and the timing
of host seeking. Species that feed late at night
and preferentially on humans must be adapted
to entering houses to fi nd a blood meal. It is
against such species that eave treatments,
ceilings, door and window screening and
curtains should be most ef ective.

common. Ogoma et al . (2009) found 59% of
sampled houses in Dar es Salaam, Tanzania had
open eaves. Yet in north-west Burkina Faso
almost all the houses are built with closed eaves
(Yé et al ., 2006). In The Gambia eaves are
usually closed if the roof is made from metal
sheeting in order to reduce the likelihood of the
roof being carried of by gusts of wind that lift
the overhanging section (Lindsay and Snow,
1988; Kirby et al ., 2008a). This also seems likely
to be the case in Sri Lanka (Gamage-Mendis et
al ., 1991). In this way, metal roofi ng is protective
against vector house entry largely by association
with closed eaves. In contrast, metal roofs in
part of the Kenyan highlands are associated
with open eaves and therefore have a higher OR
for malaria transmission than thatched roofs
with closed eaves (Ernst et al ., 2006).
To stop house entry via this route the eave
gap can be completely closed of (blocked) or
screened with insecticide-treated or untreated
material (Fig. 7.2), sometimes referred to as 'eave
curtains'. Closing the eave gap can be done with
sand, rubble and cement (Kirby et al ., 2008b) or
other appropriate fi ller and mortar for gaps
smaller than 20-30 cm. However, for larger
spaces, a dif erent approach such as hanging
cloth or netting to cover the gap might be more
appropriate. The uniformity of the gap and the
materials surrounding the gap are also important
as they must provide attachment points for
any eave treatment. In southern Mozambique,
houses commonly have large (>15 cm) openings
or 'gables' at the ends of the building with much
smaller 'eave' gaps to the sides (Kampango et al .,
2012). Closing the larger gaps is obviously the
priority for such houses; in fact, covering the
gables only with ZeroVector®, a woven-shade
cloth, was as ef ective as covering the openings
on all four sides (Kampango et al ., 2012). Closing
the eaves with rubble and mortar achieved a 40-
66% reduction in indoor An . gambiae s.l. densities
in a range of studies conducted in The Gambia
(Lindsay and Snow, 1988; Lindsay et al ., 2003;
Njie et al ., 2009), while in other studies fi tting
traps, mesh screening or shade cloths over the
eaves in experimental huts and local housing
also showed reductions of up to 85% in An .
gambiae s.s. densities (Smith et al ., 1972; Ogoma
et al ., 2010; Kampango et al ., 2012). Similar
reductions were also achieved for Mansonia spp.
(Njie et al ., 2009; Ogoma et al ., 2010), vectors of

7.4.1 Eave treatments

Olfactory cues, chiefl y carbon dioxide and body
odour volatiles, guide vectors towards hosts
(Bertram and McGregor, 1956; Costantini et al .,
1999) (see Logan et al ., Chapter 6, this volume).
When hosts are indoors, either animals in
shelters and stables or more particularly
humans in houses, those attractive odours are
contained and concentrated. The air laden with
odour particles will fl ow predominantly from the
house through the eave gap creating an odour
plume that vectors can follow to the house. The
degree of attractiveness of this odour and, to a
lesser extent, behaviour on encountering a
vertical surface are important determinants of
eave entry into houses for any given vector
species. Highly anthropophilic species such as
An . gambiae s.s. will be more attracted to odours
emanating from human habitations. An. gambiae
typically seeks a blood meal at night, fl ies at a
low or intermediate height (Gillies and Wilkes,
1976; Snow, 1979) and then fl ies upwards when
encountering a wall to follow the odour plume
through the eave gap. It thus becomes funnelled
indoors by the overhanging roof. Less anthro-
pophilic and/or more opportunistic mosquito
species may feed on humans outdoors but fail to
feed once humans retreat indoors, possibly
because they typically encounter their preferred
hosts at a dif erent vertical level and are not as
adapted to entering houses as An . gambiae
(Snow, 1979, 1987; Lee et al ., 2006).
Eaves in rural tropical housing are usually
open, though this does vary culturally and
geographically. In a rural area of The Gambia,
75% of all houses and 91% of thatched roof
houses had open eaves (Lindsay and Snow,
1988). Even in urban areas open eaves can be