Indoor residual spraying (IRS) and long-lasting insecticidal nets (LLINs) are the key malaria vector control interventions in Ethiopia. The success of these interventions rely on their efficacy to repel or kill indoor feeding and resting mosquitoes. This study was undertaken to monitor human-biting patterns of Anopheles species in south-central Ethiopia. Human-biting patterns of anophelines were monitored for 40 nights in three houses using human landing catches (HLC) both indoors and outdoors between July and November 2014, in Edo Kontola village, south-central Ethiopia. This time coincides with the major malaria transmission season in Ethiopia, which is usually between September and November. Adult mosquitoes were collected from 19:00 to 06:00 h and identified to species. Comparisons of HLC data were done using incidence rate ratio (IRR) calculated by negative binomial regression. The nocturnal biting activities of each Anopheles species was expressed as mean number of mosquitoes landing per person per hour. To assess malaria infections in Anopheles mosquitoes the presence of Plasmodium falciparum and P. vivax circumsporozoite proteins (CSP) were determined by enzyme-linked immunosorbent assay (ELISA). Altogether 3,408 adult female anophelines were collected, 2,610 (76.6 %) outdoors and 798 (23.4 %) indoors. Anopheles zeimanni was the predominant species (66.5 %) followed by An. arabiensis (24.8 %), An. pharoensis (6.8 %) and An. funestus (s.l.) (1.8 %). The overall mean anopheline density was 3.3 times higher outdoors than indoors (65.3 vs 19.9/person/night, IRR: 3.3, 95 % CI: 1.1–5.1, P = 0.001). The mean density of An. zeimanni, An. pharoensis and An. funestus (s.l.) collected outdoors was significantly higher than indoors for each species (P < 0.05). However, the mean An. arabiensis density outdoors was similar to that indoors (11.8 vs 9.4/person/night, IRR: 1.3, 95 % CI: 0.8–1.9, P = 0.335). The mean hourly human-biting density of An. arabiensis was greater outdoors than indoors and peaked between 21:00 and 22:00 h. However, An. arabiensis parous population showed high indoor man biting activities during bedtimes (22:00 to 05:00 h) when the local people were indoor and potentially protected by IRS and LLINs. All mosquito samples tested for CSP antigen were found negative to malaria parasites. Results show much greater mosquito human-biting activities occurring outdoors than indoors and during early parts of the night, implying higher outdoor malaria transmission potential in the area. However, high bedtime (22:00 to 05:00 h) indoor biting activities of parous An. arabiensis suggest high potential intervention impact of IRS and LLINs on indoor malaria transmission.

Background: Development strategies in Ethiopia have largely focused on the expansion of irrigated agriculture in the last decade to reduce poverty and promote economic growth. However, such irrigation schemes can worsen the socio-economic state by aggravating the problem of mosquito-borne diseases. In this study, the effect of agro-ecosystem practices on malaria prevalence and the risk of malaria transmission by the primary vector mosquito, Anopheles arabiensis, in Ethiopia were investigated. Methods: In three villages in western Ethiopia practising large-scale sugarcane irrigation, traditional smallholder irrigation and non-irrigated farming, cross-sectional parasitological surveys were conducted during the short rains, after the long rains and during the dry season. Entomological surveys were undertaken monthly (February 2010-January 2011) in each village using light traps, pyrethrum spray collections and artificial pit shelters. Results: Malaria prevalence and the risk of transmission by An. arabiensis assessed by the average human biting rate, mean sporozoite rate and estimated annual entomological inoculation rate were significantly higher in the irrigated sugarcane agro-ecosystem compared to the traditionally irrigated and non-irrigated agro-ecosystems. The average human biting rate was significantly elevated by two-fold, while the mean sporozoite rate was 2.5-fold higher, and the annual entomological inoculation rate was 4.6 to 5.7-fold higher in the irrigated sugarcane compared to the traditional and non-irrigated agro-ecosystems. Active irrigation clearly affected malaria prevalence by increasing the abundance of host seeking Anopheles mosquitoes year-round and thus increasing the risk of infective bites. The year-round presence of sporozoite-infected vectors due to irrigation practices was found to strengthen the coupling between rainfall and risk of malaria transmission, both on- and off-season. Conclusion: This study demonstrates the negative impact of large-scale irrigation expansion on malaria transmission by increasing the abundance of mosquito vectors and indicates the need for effective vector monitoring and control strategies in the implementation of irrigation projects.

Anopheles arabiensis is a dominant vector of malaria in sub-Saharan Africa, which feeds indoors and outdoors on human and other vertebrate hosts, making it a difficult species to control with existing control methods. Novel methods that reduce human-vector interactions are, therefore, required to improve the impact of vector control programmes. Investigating the mechanisms underlying the host discrimination process in An. arabiensis could provide valuable knowledge leading to the development of novel control technologies. In this study, a host census and blood meal analysis were conducted to determine the host selection behaviour of An. arabiensis. Since mosquitoes select and discriminate among hosts primarily using olfaction, the volatile headspace of the preferred non-human host and non-host species, were collected. Using combined gas chromatography and electroantennographic detection analysis followed by combined gas chromatography and mass spectrometry, the bioactive compounds in the headspace collections were identified. The efficiency of the identified non-host compounds to repel host-seeking malaria mosquitoes was tested under field conditions. The host census and blood meal analyses demonstrated that An. arabiensis strongly prefers human blood when host seeking indoors, while it randomly feeds on cattle, goats and sheep when found outdoors. However, An. arabiensis avoids chickens despite their relatively high abundance, indicating that chickens are a non-host species for this vector. Eleven bioactive compounds were found in the headspace of the non-host species. Six of these were species-specific, out of which four were identified using combined gas chromatography and mass spectrometry. When tested in the field, the chicken-specific compounds, isobutyl butyrate, naphthalene, hexadecane and trans-limonene oxide, and the generic host compounds, limonene, cis-limonene oxide and β-myrcene, significantly reduced trap catches within the house compared to a negative control. A significant reduction in trap catch was also observed when suspending a caged chicken next to the trap. Non-host volatiles repel host-seeking An. arabiensis and thus play a significant role in host discrimination. As such, this study demonstrates that non-host volatiles can provide protection to humans at risk of mosquito-vectored diseases in combination with established control programmes.

Maize cultivation contributes to the prevalence of malaria mosquitoes and exacerbates malaria transmission in sub-Saharan Africa. The pollen from maize serves as an important larval food source for Anopheles mosquitoes, and females that are able to detect breeding sites where maize pollen is abundant may provide their offspring with selective advantages. Anopheles mosquitoes are hypothesized to locate, discriminate among, and select such sites using olfactory cues, and that synthetic volatile blends can mimic these olfactory-guided behaviours. Two-port olfactometer and two-choice oviposition assays were used to assess the attraction and oviposition preference of gravid Anopheles arabiensis to the headspace of the pollen from two maize cultivars (BH-660 and ZM-521). Bioactive compounds were identified using combined gas chromatography and electroantennographic detection from the headspace of the cultivar found to be most attractive (BH-660). Synthetic blends of the volatile compounds were then assessed for attraction and oviposition preference of gravid An. arabiensis, as above. Here the collected headspace volatiles from the pollen of two maize cultivars was shown to differentially attract and stimulate oviposition in gravid An. arabiensis. Furthermore, a five-component synthetic maize pollen odour blend was identified, which elicited the full oviposition behavioural repertoire of the gravid mosquitoes. The cues identified from maize pollen provide important substrates for the development of novel control measures that modulate gravid female behaviour. Such measures are irrespective of indoor or outdoor feeding and resting patterns, thus providing a much-needed addition to the arsenal of tools that currently target indoor biting mosquitoes.

Blood-feeding behavior studies are important for estimating the efficiency of pathogen transmission and assessing the relative human disease risk. However, in Ethiopia and other parts of East Africa there are large remaining gaps in identifying the feeding habits of Phlebotomus orientalis, the vector of Leishmania donovani. The aim of the study was to determine the blood feeding patterns of P. orientalis in Tahtay Adiyabo district, northern Ethiopia. For bloodmeal analysis, sandflies were collected from three different villages of Tahtay Adiyabo district using CDC light traps, sticky traps, and pyrethrum spray catches. Bloodmeal of engorged female sandflies was identified using cytochrome (cyt) b-PCR and reverse-line blotting (RLB) and enzyme linked immunosorbent assay (ELISA) assays. Most (637/641) of the females analyzed were P. orientalis. Successful identification of the host from bloodmeals was achieved in 83.03 and 92.1 % using cyt b PCR-RLB and ELISA, respectively. Bloodmeal analysis of P. orientalis females revealed that they have a range of hosts with predominant preference to bovines followed by donkey, human, goat, sheep, dog, and camel. Results obtained from bloodmeal analyses demonstrate that the feeding preference of P. orientalis is mainly zoophilic, which could vary depending on the availability of hosts.

Host choice and feeding success of sand fly vectors of visceral leishmaniasis (VL) are important factors in understanding the epidemiology and for developing efficient control strategies. The aim of the present study was to determine the host preference of Phlebotomus orientalis in the VL focus of Tahtay Adiyabo district, northern Ethiopia. Two separate experiments were conducted testing attraction of P. orientalis to humans, domestic animals, and small wild animals. The host choice of P. orientalis and other sand fly species was assessed using tent traps baited with seven different animals (human, cow, sheep, goat, donkey, dog and chicken) and a blank control. Baited traps were rotated every night in a Latin square design for two consecutive full rounds totaling 16 trap-nights. The second set of experiments tested attraction to small wild animals including; ground squirrel (Xerus rutilus), hare (Lepus sp.), gerbil (Tatera robusta) and spiny mouse (Acomys cahirinus). Animals were caged in standard rodent traps or cylindrical wire-mesh cages. The bait animals were placed in agricultural field and the attracted sand flies were collected using unlit CDC traps for 10 trapping nights. Sand fly specimens collected from each of the experiments were identified to species level and counted. Significant difference (P < 0.05) was observed in the attraction and feeding rate of P. orientalis to different baits. In the first experiment, cow-baited tent traps attracted the highest mean number of P. orientalis (mean = 510 flies). The engorgement rate of P. orientalis on donkey was the highest followed by cow, and much lower on goat, sheep, dog and chicken. In the case of smaller wild animals, more numbers of P. orientalis females were attracted to squirrels followed by hares, gerbils and the spiny rat. However, the engorgement rates for P. orientalis in the smaller animals were very low (1.08%) compared with larger domestic animals (30.53%). The tendency of female P. orientalis to engorge in large numbers on certain species of domestic as well as wild animals strongly indicated that the species is primarily zoophilic in its host preference with feeding habits that may vary depending on the availability of hosts.