healthmedicinet

Study sites and period of capture

Entomological surveys were conducted in two communes: Kiangara and Marondry within
the Ankazobe district (Fig. 1). The site of Andranovelona II (Site I, S 18°23?25.2??/EO 47°01?00.0??) was sampled
in April and May 2013 whereas the sites of Bemasoandro (Site A, 17°S 57?26; 47°EO
02?60), Kianjasoa (Site B, 17°S 58?25; 47°EO 01?80), Ambohimiadana (Site C, 17°S 58?59;
47° EO 03?26), Ambohimanjaka (Site D, 17°S 59?27; 47°EO 02?07), Morafeno (Site E,
18°S 24?13; 47°EO 03?03), Miarinarivo Sud (Site F, 18°S 26?05; 47°EO 00?02), Voninahitrinitany
(Site G, 18°S 31?50; 47°EO 01?42), and Tsarahonenana (Site H, 18°S 33?30; 47°EO 01?49)
were sampled in March 2014 (Fig. 1). In April 2013, human landing catches (HLC), Centers for Disease Control and Prevention
simple light traps (CDC LT), collection of mosquitoes resting indoors (MRI) and collection
of mosquito resting in stables by oral aspirator were carried out. In May 2013 and
March 2014, collection of mosquitoes resting in stables and HLC were carried out,
respectively.

Fig. 1. Study sites

Human landing catches

HLC were performed over two consecutive nights from the local population from 18:00
to 08:00 h in four different houses with two adult volunteers per house: one located
inside and another one outside 11]. Mosquitoes coming to bite the collectors were detected using a flashlight, collected
with glass tubes and placed in the collecting bags every hour. The four houses were
chosen randomly in the village with no repetition. The capturers took one tablet of
doxycycline over 5 days as prophylaxis against infection with the malaria parasite.
The HLC were approved by the local health authority.

Centers for Disease Control and Prevention simple light traps

CDC LT is a system that incorporates a mini-light source attracting mosquitoes, which
are drawn in through the top of the trap and forced downwards by the fan into the
collection bag-net. Light traps were powered by 6 V batteries. Six CDC LT were set
before sunset (18:00 h) and off after sunrise (06:00 h). The CDC LT was set up outdoors
over two successive nights.

Collection of mosquitoes resting indoors

A spraying of non-remanent insecticide was performed in five houses chosen randomly
where no HLC and CDC LT were made. Doors and windows were closed and eaves, which
allow mosquitoes to escape from houses, were covered during the spraying. After 15 min,
all dead and paralyzed mosquitoes fell onto sheets placed on the floor and over the
beds and furniture before the spraying. All mosquitoes were collected and identified.
Those identified as anophelines were preserved for further analysis. MRI was carried
out between 06:00 and 08:00 h over 2 days.

Collection of mosquito resting in stables

On the first and second day, two entomologists went into stables and caught mosquitoes
with oral aspirators. Two stables were sampled during two mornings between 07:00 and
09:00 h.

Identification

Mosquito identification was performed with the aid of a binocular microscope following
the morphological keys 12] and (Fontenille, personal communication). Identifications were carried out in the
field. Legs or wings of mosquitoes from the An. gambiae complex were used for PCR identification 13]. The amplification was done under the following conditions: 5 min at 94 °C followed
by 30 cycles of 1 min at 94 °C, 50 s at 50 °C and 50 s at 72 °C, with a final elongation
step (5 min at 72 °C). The sizes of the fragments obtained were, respectively, 315,
390 and 464 base pairs for An. arabiensis, An. gambiae s.s. and An. merus.

Estimation of the entomological indices

Detection of Plasmodium in mosquitoes was carried out with head and thorax of all female anopheline species
by ELISA CSP screening. Any sample positive following ELISA CSP screening were tested
in monospecific ELISA CSP (for identification of P. falciparum, Plasmodium vivax) 14] and in PCR (for identification of P. falciparum, Plasmodium malariae, P. vivax, or Plasmodium ovale) 15]. To avoid false positive, lysates were not heated as previously described 16], 17] but all specimens were confirmed in PCR. To avoid contamination during the grinding,
each specimen was put in a 1.5-ml Epperdorff with 3 ml of steel beads. Then, they
were ground in Tissulyser (Tissulyser II, Qiagen
^®
).

Human biting rate (HBR) was estimated for An. funestus, An. mascarensis, An. arabiensis, and An. coustani. For the sample positive to Plasmodium, the entomological infection rate (EIR) and the sporozoite indices were estimated.
HBR is the number of bite for a given vector per person per night (bpn). The EIR is
the product of the HBR, the number of bites per person per month by vector mosquitoes
and the fraction of vector mosquitoes that are infectious (the sporozoite rate). The
sporozoite index indicates the proportion of individual positive with Plasmodium among the total individuals caught for one species. These indices were calculated
based on the HLC in April 2013 and March 2014.

Blood meal analyses

Blood meal analyses were performed on blood fed females caught in stables in May 2013.
Direct ELISA as previously described using antihost (IgG) conjugate against human,
cow, chicken, and pig proteins were carried out 18].