Methods of sampling malaria vectors and their reliability in estimating entomological indices in Africa

Abstract

In efforts to intensify malaria control through vector control and hasten the progress towards elimination, the impact of control interventions needs to be evaluated. This requires sampling vector population using appropriate trapping methods. The aim of this article is to critically review methods of sampling malaria vectors and their reliability in estimating entomological indicators of malaria transmission in Africa. The standard methods are human landing catch (HLC), pyrethrum spray catch, and pit shelter for sampling host-seeking, indoor resting, and outdoor resting malaria vectors, respectively. However, these methods also have drawbacks such as exposure of collectors to infective mosquito bites, sampling bias, and feasibility issue. Centers for Disease Control and Prevention (CDC) light traps placed beside human-occupied bed nets have been used as an alternative to the HLC for sampling host-seeking malaria vectors. Efforts have been made to evaluate the CDC light traps against HLC to generate a conversion factor in order to use them as a proxy estimator of human biting rate and entomological inoculation rates in Africa. However, a reproducible conversion factor was not found, indicating that the trapping efficiency of the CDC light traps varies between different geographical locations. Several other alternative traps have also been developed and evaluated in different settings but most of them require further standardization. Among these, human-baited double net trap/CDC light trap combination and mosquito electrocuting trap have the potential to replace the HLC for routine malaria vector surveillance. Further research is needed to optimize the alternative sampling methods and/or develop new surveillance tools based on vector behavior.

Background

Malaria remains to be one of the most serious vector-borne infectious diseases, affecting hundreds of millions of people in Africa. In 2021, an estimated 247 million malaria cases and 619,000 malaria-related deaths were reported globally, with about 95% of the cases and 96% of the deaths estimated to have occurred in Africa (WHO 2022). The disease is caused by 5 Plasmodium species: P. falciparum, P. vivax, P. ovale, P. malaria, and P. knowlesi. Of these, P. falciparum is the predominant species and is responsible for most of the deaths in sub-Saharan Africa (SSA) (WHO 2020). It is transmitted by bites of infected female Anopheles mosquitoes. There are over 144 species of Anopheles mosquitoes in SSA, of which about 20 are known to transmit malaria to humans (Irish et al. 2020). Of these, Anopheles gambiae, An. coluzzii, An. arabiensis, An. funestus, and An. stephensi are the most efficient vector species (Sinka et al. 2010, 2020).

In the past decades, massive scale up of malaria control interventions has resulted in a substantial decline of malaria in SSA, averting an estimated 663 million cases between 2000 and 2015 (O’Meara et al. 2010, WHO 2015). Between 2000 and 2020, an estimated 1.7 billion malaria cases have been averted globally, with over 10.6 million lives estimated to have been saved over the past 2 decades (WHO 2020, 2021). Most of the averted cases (82%) and deaths (95%) were from the SSA. Vector control is one of the key elements in achieving the remarkable reduction, with long-lasting insecticidal nets (LLINs) and indoor residual spraying (IRS) contributing significantly (WHO 2015). However, such progress has stalled in the past 6 yr, and increases in malaria incidence and prevalence were observed in several endemic countries (WHO 2018a, 2019, 2020) due to several factors including the spread of insecticide resistance in malaria vectors (Ranson and Lissenden 2016), persistent malaria transmission by outdoor and early biting vectors (Sherrard-Smith et al. 2019), drug resistance in malaria parasites (Lu et al. 2017, Conrad and Rosenthal 2019, Uwimana et al. 2020), and spread of the invasive urban malaria vector, An. stephensi, in Africa (Takken and Lindsay 2019). This suggests the need to regularly evaluate the effectiveness of the current malaria control interventions to identify gaps and design complementary vector control interventions in order to further reduce the disease burden and hasten the progress towards elimination.

Evaluating the effectiveness of vector control interventions requires assessing the impact of the interventions on various entomological indices (entomological indicators of malaria transmission) including vector species composition, distribution, density, biting behavior, feeding behavior (blood meal indices), resting behavior, human biting rate (HBR), sporozoite rate, and entomological inoculation rate (EIR) (WHO 2013, 2018b). The success of the control interventions can be measured in terms of reduction in EIR i.e., malaria transmission intensity (Lines et al. 1991). The EIR, the number of infective mosquito bites received by a person per unit time, is determined by multiplying the HBR and sporozoite rate (Lines et al. 1991). Both HBR and sporozoite rate could vary in time and space depending on the mosquito vector density and behavior (Kenea et al. 2016, Burkot et al. 2018).

Monitoring the entomological indicators of malaria transmission requires sampling mosquito vectors using appropriate trapping methods to obtain representative samples of the vector populations (Service 1977). The trapping methods to be used may differ depending on the type of the entomological indices to be measured, i.e., mosquito biting behavior, blood meal sources, resting habits, or malaria transmission intensity (Service 1977). The mosquito vector species may exist as indoor host-seeking, indoor resting, outdoor host-seeking, and outdoor resting fractions, each fraction requiring different sampling methods and tools (WHO 1975, 2013).

Studies have shown that entomological indices such as HBI and EIRs vary widely across different African settings. For instance, the HBI of An. arbaiensis was reported to be <10% in some African settings (Degefa et al. 2017) and over 80% in some other settings (Fornadel et al. 2010b, Kibret et al. 2017). The EIRs range from <1 to over 800 infective bites per person per year in the region (Degefa et al. 2017, Epopa et al. 2019, Kwi et al. 2022, Fondjo et al. 2023, Koffi et al. 2023). While this variation could be due to differences in the receptivity of the different African settings to malaria, such variation could also occur due to differences in vector sampling methods and approaches used to estimate the HBI and EIRs (Massebo et al. 2013b, Degefa et al. 2017). The main aim of this article is to critically review methods of sampling adult malaria vectors and their reliability in estimating entomological indicators of malaria transmission in Africa.

Methods

Electronic search of relevant literatures was made on PubMed and Google Scholar Databases using different keywords such as “malaria vector sampling methods, vector sampling techniques, mosquito traps, adult mosquito collection methods in Africa, comparison of adult mosquito sampling methods in Africa, blood meal indices of African malaria vectors, and entomological inoculation rates of malaria vectors in Africa”. Moreover, the names of different traps used for sampling resting and host-seeking malaria vectors in Africa were used as keywords to retrieve the relevant literatures. Abstracts of the searched literatures were reviewed, and if found relevant, their full-text contents were assessed.

This review study focuses on methods of sampling indoor and outdoor resting and host-seeking malaria vectors to assess the reliability of the sampling methods in estimating entomological indices in Africa. Other methods of sampling malaria vectors at different life stages such as different mosquito larval and pupal sampling tools, and emergence traps for collection of adult mosquitoes emerging from breeding sites were not included in this work.

Results and Discussion

Methods of Sampling Indoor and Outdoor Resting Malaria Vectors

Sampling indoor and outdoor resting mosquito vectors is a pre-requisite to get information about the most common resting places (indoor or outdoor) and to determine the effect of control interventions on indoor and outdoor resting vector density and species composition (WHO 2013). Moreover, samples of both indoor and outdoor mosquito collections are crucial to determine vector feeding behavior and host preferences by estimating human blood index (HBI) and blood meal indices of other vertebrate hosts through host blood meal analysis (WHO 2013).

Pyrethrum spray catch.

Pyrethrum spray catch (PSC) involves using a pyrethrin space spray to knockdown mosquitoes resting indoors and collecting them on white sheets spread on the floor (WHO 2013). PSC is considered to be a gold standard method for estimating indoor resting mosquito density. Moreover, PSC is an ideal method to obtain engorged anopheline mosquitoes for monitoring vector feeding behavior through host blood meal analysis (Githeko et al. 1994b, Animut et al. 2013, Massebo et al. 2013a, Ndenga et al. 2016). The method can also be employed for indirect estimation of HBR and EIR as follows (WHO 2013):

However, since PSC is used only for indoor resting mosquito collection, it misses fractions of mosquitoes that leave houses immediately after feeding due to the excito-repellent effect of LLINs and IRS (Durnez and Coosemans 2013). Moreover, PSC is less sensitive in settings where mosquitoes mostly exhibit exophagic and exophilic behavior (Mahande et al. 2007). In western Kenya for example, a substantial proportion of indoor biting An. gambiae s.l., 19.4% of which were blood fed, were captured while exiting houses, before PSC was conducted in the morning (Degefa et al. 2017, 2019). In this case, PSC could underestimate the actual HBR thereby underestimating the actual EIRs. PSC may also result in a false impression of the effectiveness of malaria vector control measures by underestimating vector density when vector populations are exophagic and exophilic. In Guinea-Bissau for instance, reliance on indoor resting collection methods such as PSC underestimated the presence of An. arabiensis population in the country (Gordicho et al. 2014). This suggests the need to complement PSC with other surveillance tools such as exit trap that could trap fractions of the mosquito vectors that exit after feeding in order to quantify the actual HBR of the vectors. Moreover, it is reasonable to pair PSC and exit trap with another trap that could capture the outdoor resting fractions of the vectors in order to have unbiased estimation of blood meal indices and HBR of mosquito vectors.

Another limitation of PSC is that it may create inconvenience to residents of sprayed houses as the collection procedure involves temporary removal of house properties such as furniture, cookware, food, water, people, and animals in the early morning just before spraying the houses (WHO 2013). Moreover, PSC may cause discomfort to occupants as some people may have allergy to the sprayed insecticides.

Hand collection.

Hand collection involves the use of mouth aspirators or sucking tubes to sample mosquitoes resting indoors and/or from their natural outdoor resting sites. Hand collection provides important information about mosquito vector resting density and seasonal changes in their density. It also provides live specimens for bioassay testing to determine insecticide susceptibility level of malaria vectors (WHO 2013). The limitation of the hand collection is that it is time-consuming and it unlikely captures all resting mosquitoes, hence it is not appropriate trapping method to be used for a routine monitoring of vector density as it may not indicate the actual mosquito density (WHO 2013). Furthermore, mosquito collection using manual aspirators requires systematic and attentive work that is highly reliant on the skill and motivation of the collectors (WHO 2013).

Electronic aspirators.

In efforts to minimize the level of skill needed to use manual aspirators, several battery-powered aspirators have been developed and tested in different settings. The most frequently used electronic mosquito aspirators are the backpack aspirator, developed in 1990s by the Centers for Disease Control and Prevention (CDC) (Clark et al. 1994), and a prokopack aspirator devised by Vazquez-Prokopec et al. (2009). The aspirators can be used for both indoor and outdoor resting mosquito collection (Maia et al. 2011). In Tanzania, prokopack aspirator was found to be more efficient than the manual aspirator, yielding about 1.5 times higher mosquito density compared to the manual aspiration (Charlwood et al. 2018). According to another study done in southern Tanzania, prokopack and backpack aspirators showed a similar performance, although the prokopack aspirator was more consistent when used by different collectors (Maia et al. 2011). In coastal Kenya where mosquito density was low, prokopack aspirator yielded a similar density of An. gambiae s.l. and An. funestus to that of PSC (Onyango et al. 2013).

However, both aspirators do have limitations. Backpack aspirator is relatively heavy; thus, it may not be suitable to use for routine surveillance of malaria vectors (Maia et al. 2011). Both aspirators rely on batteries, and may not be feasible to use them in rural African environments where there is no electricity for charging the batteries (Vazquez-Prokopec et al. 2009, Maia et al. 2011).

Exit traps.

Some mosquito species such as An. arabiensis exhibit a tendency to enter houses in the evening or at night to bite, and then leave the houses immediately after feeding without resting indoors (Mboera 2005, Pates and Curtis 2005, Tirados et al. 2006, Killeen et al. 2016). This fraction of mosquito vectors, together with those that rest indoors but eventually leave the houses to lay eggs, can be monitored by setting exit traps on windows (WHO 1995). Mosquitoes are captured by the exit traps as they escape, thus allowing malaria vector density to be monitored. Data obtained from the exit trap collections are important to get information regarding the exophilic versus endophilic resting behavior, and physiologic status of the vector population. This method may also be used to test the behavioral avoidance responses of mosquito vectors to different insecticides sprayed on the wall surface of the houses or insecticides used to impregnate bed nets (Quinones et al. 1997).

Exit trap has been reported to be useful for monitoring malaria vector density and resting behavior in African countries such as Southern Africa (Mouatcho et al. 2007), Equatorial Guinea (Sharp et al. 2007, Ridl et al. 2008), Kenya (Wong et al. 2013) and Ethiopia (Abraham et al. 2017). In Kenya, exit trap yielded a significantly higher density of An. gambiae s.l. and An. funestus compared to pit shelter but lower density compared to PSC (Table 1). Because the exit trap catches endophagic vectors while exiting houses, but this fraction of mosquitoes are unlikely captured by the gold standard PSC, it plausible to use both exit trap and PSC in routine vector surveillance to determine the actual HBR and EIRs, using the following modified equations:

The limitation of the exit trap is that its trapping efficiency might be affected by variations in house design and structures (Govella et al. 2011, Sikaala et al. 2013). Exit trap showed poor sampling sensitivity in African settings where most houses had open eaves and without ceiling (Govella et al. 2011), as mosquitoes could find several alternative openings to exit.

Pit shelters.

Traditionally, mechanical aspiration of mosquitoes from man-made pit shelters has been used as a standard method for sampling outdoor resting fractions of mosquito vectors (WHO 1995). To use this method, rectangular pits should be dug into the ground 1.5–2 m deep with 1.2–1.5 m length and 1 m width (WHO 1995). The site chosen to dig the pits should be well shaded by trees or the pits should be shaded by artificial framework thatched with locally available reeds. Small cavities are then hollowed out to a depth of about 30 cm in each vertical side of the pit at 50–60 and 90–100 cm from the bottom of the pit. Mosquitoes are collected from the small cavities and the wall surface of the pit using aspirator.

Pit shelter has the potential to provide focused sites for mosquito collections and yields a representative mosquito samples that can be used for different immunological and molecular assays (Kweka and Mahande 2009, WHO 2013). This method was proved to be an effective means of sampling outdoor resting population of An. arabiensis for blood meal analysis in different parts of Ethiopia (Ameneshewa and Service 1996, Tirados et al. 2006, Massebo et al. 2015), Eritrea (Shililu et al. 2004), Kenya (Degefa et al. 2017), and Tanzania (Ijumba et al. 2002, Kweka and Mahande 2009). However, sampling inside pits is difficult to standardize. The procedure requires skill to capture all resting mosquitoes, as mosquitoes may escape when they are disturbed during the collection process. It is also difficult to maintain pit shelters during the wet season as the pits could be muddy or filled with rainwater. Such pits constructed in residential compounds may also be risky to kids and domestic animals especially when the pits are filled with rain water. Moreover, harmful animals such as snakes may also be found in the pits and if encountered, they may cause a risk to mosquito collectors.

Resting boxes.

Resting boxes have been used to sample mosquitoes since it was first observed that mosquitoes tend to congregate in dark places (Crans 1989). It was assumed that resting boxes could provide unbiased samples of both endophilic and exophilic mosquito populations when the traps are placed both indoors and outdoors. However, the number of adults resting in the resting boxes depends on the availability of alternative resting sites (Service 1977); hence, it may not be as productive as the traditional outdoor trapping methods such as pit shelter. A recent study conducted in Burkina Faso showed that resting boxes yielded a positive correlation with pit shelters in sampling An. gambiae s.l. Nevertheless, the daily performance of the resting boxes was 5 times lower in terms of mosquito density per trap (Pombi et al. 2014).

Clay pots.

Clay pots have also been developed and evaluated for outdoor resting mosquito collection (Odiere et al. 2007). In western Kenya, clay pots have been successfully used to collect outdoor resting female and male An. arabiensis and An. gambiae (Odiere et al. 2007, Machani et al. 2020). The advantage of the clay pots is that they are small in size and transportable, so that they could be deployed in large numbers and in many sites. However, recovering mosquitoes resting within the pots still needs aspiration by mosquito collectors (Odiere et al. 2007), which may lead to sampling bias due to variation in skill among the collectors. Moreover, mosquitoes could escape before collection when the pots are disturbed by animals or children playing around the pots.

Sticky pots.

In an attempt to overcome the limitation of clay pots, a sticky pot, a sticky variant of the clay pot has been developed and evaluated recently (Degefa et al. 2019). In a sticky pot, the interior surface of the clay pot is lined with waterproof black paper which is pre-coated with Tangle-Trap sticky substance. The addition of this sticky substance allows for mosquitoes that rest within the pot to be continuously trapped, rather than only observing a fraction of mosquitoes that happen to be resting at the time of collection in a standard clay pot. Sticky pots are made using locally available pots, and hence they are cost effective to use them for routine surveillance of outdoor resting malaria vectors. In western Kenya, a sticky pot captured 1.6 times as many An. gambiae s.l. as a clay pot (Degefa et al. 2019). This shows that the sticky pots could be a useful and complementary tool for routine surveillance of outdoor resting African malaria vectors.

However, the sticky pot yielded low vector density when compared with pit shelter although it captured higher mosquito density compared to clay pot (Table 1). In Kenya, a sticky pot yielded 4 times lower density of An. arabiensis compared to pit shelter while clay pot yielded 6 times lower density of the same species compared to the pit shelter (Degefa et al. 2019). This shows the need to deploy 4 sticky pots or 6 clay pots per compound as a replacement for a pit shelter in order to use them for routine surveillance of outdoor resting malaria vectors.

Methods of Sampling Indoor and Outdoor Host-Seeking Malaria Vectors

Human landing catch.

The HLC involves a volunteer person (usually male) exposing his legs and collecting mosquitoes using an aspirator when they land on his legs, before they commence biting (Mboera 2005). This is considered to be the most reliable and direct method for estimating human exposure to mosquito bites and obtaining samples of human-biting mosquitoes (Lines et al. 1991, Service 1993, Davis et al. 1995, Mboera 2005), and is therefore it is widely accepted as a gold standard method for determining EIRs (Service 1993). Since mosquitoes are captured while attempting to bite human (Lines et al. 1991, Davis et al. 1995, Mboera 2005), the number of mosquitoes caught per person/night can be considered to reasonably represent the HBR, and the sample of mosquitoes obtained to have similar distribution of age, physiological status and infection status as those which feed on people at that time and place. Moreover, HLC can be conducted both indoors and outdoors, and hence it could provide important information about when and where people are exposed to infectious vector bites, as well as the degree of exophagy of the vector populations. Such information on indoor and outdoor biting behavior of the mosquitoes have implications for malaria epidemiology, both in terms of host-vector contact and the choice of effective malaria vector control measures (Pates and Curtis 2005).

Nonetheless, HLCs have several shortcomings. It is a difficult, uncomfortable, tiresome, and labor-intensive technique, requiring such an intense supervision that it is hard to sustain on large scale. Close supervision is required because the collector needs not only to remain awake, but also to constantly cautious for the data to be trustworthy (Mboera 2005). Moreover, there may be significant differences between biting rates experienced by different mosquito collectors due to variation in attractiveness to the mosquitoes (Lindsay et al. 1993) and difference in mosquito catching skill (WHO 1995, Mboera 2005). A greater concern arises from the fact that this method may increase the risk of exposure of the collectors to different mosquito-borne infections (Mboera 2005), which is difficult to vindicate on ethical grounds.

CDC light traps.

In an attempt to search for an alternative trap to HLC, different designs of light traps were developed and their reliability in estimating EIRs has been evaluated under different settings. Of the various designs, CDC miniature light traps are the most frequently used alternative method for sampling host-seeking malaria vectors. The traps are battery powered with a motorized fan, light bulb, and a collection cup. Mosquitoes attracted to the traps by host odor and light, are drawn in at the top, and forced downward by the fan into the collection cup, from which they cannot escape. In the first evaluation, it was noted that the trapping efficiency of the CDC light traps has increased when the traps were placed close to hosts (Odetoyinbo 1969), and subsequent experiments proved that its sampling efficiency has improved significantly by setting the trap beside human hosts protected by either untreated or insecticide-treated bed nets (Garrett-Jones et al. 1975, Magbity et al. 2002). Since then, CDC light traps have been used by setting indoors beside human-occupied bed nets for monitoring vector density and for estimation of HBR, sporozoite rates, and EIRs (Lines et al. 1991, Mbogo et al. 1993, Githeko et al. 1994a, Davis et al. 1995, Drakeley et al. 2003, Mathenge et al. 2004).

Several studies have evaluated the trapping efficiency of the CDC light traps against the gold standard HLC to find a conversion factor (CF) that may be used to infer HBR from the number of mosquito vectors caught by the light traps, but different studies reached on different conclusions (Lines et al. 1991, Fornadel et al. 2010a, Kenea et al. 2017). In many studies conducted in different African countries, the CDC light traps yielded significantly lower vector density compared to HLC (Lines et al. 1991, Le Goff et al. 1993, Githeko et al. 1994a, Govella et al. 2011, Kenea et al. 2017), but positive correlations were reported between the 2 trapping methods in most of the studies (Lines et al. 1991, Githeko et al. 1994a, Kenea et al. 2017). In other studies, the light traps captured significantly higher mosquito density (Davis et al. 1995, Costantini et al. 1998, Mathenge et al. 2004, Fornadel et al. 2010a) (Table 2). In both cases, the HBR and EIR can be estimated from the CDC light trap collections by considering the CF as follows:

However, CDC light traps have also several limitations. The conversion factors that have been suggested for CDC light traps versus HLC vary between different countries and even within the same country in different geographical locations (Table 2), leading to confusion on which conversion factor to use for estimation of HBR from CDC light traps. In some studies, the trapping efficiency of CDC light traps was found to be density dependent, and its trapping efficiency was shown to be affected by seasonal variations, weather conditions, and trap positions (Mbogo et al. 1993, Service 1993, Mboera et al. 1998, Overgaard et al. 2012). Moreover, some studies have documented higher sporozoite rates for mosquitoes captured by the CDC light traps as compared to that of HLC (Mbogo et al. 1993, Mboera 2005), and this may lead to an overestimation of EIRs. Furthermore, CDC light traps have been reported to be less effective and unreliable for sampling outdoor host-seeking malaria vector populations in most studies (Service 1993, Costantini et al. 1998, Overgaard et al. 2012, Kenea et al. 2017). In addition, CDC light traps may not be effective for sampling urban malaria vectors such as An. stephensi at least for 2 reasons: (i) this invasive vector species is mostly exophagic; hence, indoor CDC light trap could miss substantial proportion of An. Stephensi and (ii) if CDC light traps are set outdoors in such urban settings, the function of the light within the trap is likely compounded by outdoor electricity light, leading the mosquitoes to fly to different light sources.

Mbita trap.

The Mbita trap was designed primarily for sampling unfed host-seeking mosquitoes, based on observations of their host-seeking behavior around human-occupied bed nets. It is conically shaped, resembling a bed net made of cotton cloth with its circular upper part consisting of a netting funnel with a small inner aperture kept open by a small metal ring. These structural features allow the entrance of mosquitoes but limit their exit (Mathenge et al. 2002). The Mbita trap does not expose volunteers to mosquito bites, allows them to sleep throughout the night, and it does not require skill and electrical power (Mathenge et al. 2002). Initial evaluation of the Mbita trap in Kenya showed the trap to be relatively sensitive and provided catches that were proportional to those obtained by HLC (Mathenge et al. 2002, 2004). However, other studies have reported poor performance for this trap (Laganier et al. 2003, Braimah et al. 2005, Mathenge et al. 2005, Okumu et al. 2008) (Table 3). In Madagascar for instance, the Mbita trap yielded mean mosquito density of 1.0 per trap-night while HLC collected on average 15 mosquitoes per person-night in the same villages (Laganier et al. 2003). In western Kenya, the Mbita trap caught about half of the number of An. gambiae s.l. caught in the HLC (Mathenge et al. 2005).

Tent traps.

Several designs of tent traps have been developed and evaluated for outdoor biting malaria vector surveillance. These include Ifakara tent traps (Govella et al. 2009, 2011) and Furvela tent trap (Charlwood et al. 2017). The Ifakara tent trap is a human-baited trap made of rectangular boxes containing 6 funnel-like entrances for mosquitoes and inner small apertures tilted to an angle so that mosquitoes have to fly upward to enter the trap after which cannot escape once they enter the trap (Govella et al. 2009). Although tent traps have been shown to possess the potential for monitoring African malaria vectors, they do have major limitations. The use of Ifakara tent traps, for example, may raise major ethical concerns due to risk of operators’ exposure to mosquito bites during the collection process (Govella et al. 2009). Moreover, there is a doubt whether the tent traps best estimate indoor biting or outdoor biting mosquito densities (Govella et al. 2011).

Odor-baited traps.

Host odors play a major role in attracting host-seeking mosquitoes (Takken and Knols 1999). Carbon dioxide (CO₂), human sweat, and skin residues such as ammonia and L-lactic acid are known to attract host-seeking malaria vectors (Takken and Knols 1999, Healy and Copland 2000), and hence they can be used as a strategy to attract and sample mosquito vectors. Several designs of traps such as mosquito magnet-x (MM-X) trap (Njiru et al. 2006, Schmied et al. 2008), BG-Sentinel (BGS) trap (Kröckel et al. 2006, Batista et al. 2017), BG-Malaria (BGM) trap (Batista et al. 2017), Suna trap (Hiscox et al. 2014, Mburu et al. 2019), host decoy trap (Hawkes et al. 2017, Abong’o et al. 2018) and odor-baited entry traps (Costantini et al. 1993, Duchemin et al. 2001) have been developed for sampling host-seeking mosquitoes by incorporating such chemical attractants.

The MM-X trap uses different attractants, CO₂, and counterflow technology to capture mosquitoes (Kline 1999). The trap has a potential to attract and catch host-seeking mosquitoes (Njiru et al. 2006, Schmied et al. 2008). However, it was not evaluated and optimized against the gold standard HLC in Africa. Outside Africa, the MM-X trap caught a significantly lower number of Anopheles mosquitoes compared to the HLC (Jeyaprakasam et al. 2021).

The BGS (BioGents HmGb, Regensburg, Germany) is a simple suction trap that uses upward-directed air currents and visual cues to attract mosquitoes. It has a dispenser system, BG-Lure, which releases artificial human skin odor (Kröckel et al. 2006). The BGM trap is a modification of BGS trap, hung upside down at 40 cm above the ground, and has an electrical fan which produces an upward suction that captures mosquitoes approaching the trap (Batista et al. 2017). A study done in Tanzania showed that both BGM and BGS traps caught significantly lower number of An. gambiae s.l. than HLC, but the BGM yielded a higher density of An. funestus compared to the HLC (Batista et al. 2017) (Table 3).

Suna trap is an odor-baited trap that has been developed to collect host-seeking mosquitoes both indoors and outdoors (Hiscox et al. 2014). To attract mosquitoes, it uses a synthetic blend of chemicals found on human skin (Mukabana et al. 2012) and CO₂ produced through a process of yeast and molasses fermentation (Mweresa et al. 2014). In Malawi, Suna trap caught a similar number of Anopheles mosquitoes as the HLC both indoors and outdoors (Mburu et al. 2019). Suna trap does not require human labor once it is set in the evening as it can collect mosquitoes throughout the night (Hiscox et al. 2014). However, it yielded lower mosquito density compared to the HLC in another study (Verhulst et al. 2015), suggesting the need to evaluate the trap in different African settings and standardize to use it for routine malaria vector surveillance.

Human-odor-baited CDC light trap.

Human-odor-baited CDC light trap (HBLT) is a modification of CDC light traps developed for the surveillance of outdoor host-seeking mosquito vectors (Degefa et al. 2020). It consists of a CDC light trap baited with human-odor pumped from a sleeping room (Degefa et al. 2020). In Kenya and Ethiopia, HBLT captured 2–3 times as many malaria vectors as the conventional CDC light trap but significantly lower mosquito density compared to HLC (Table 3). The limitation of the HBLT is that it requires connecting pipe from a sleeping room to outdoor mosquito-catching station through a hole made on either window or wall of the room. Rooms with cement-plastered wall and without window may not be appropriate to set the HBLT (Degefa et al. 2020). Hence, it needs further modification to easily dispense human odor.

Bed net traps.

Various designs of bed net traps have been developed and used to catch outdoor host-seeking mosquitoes, using either humans or animals as bait. The usual procedure involves a man sleeping under a mosquito net that is either raised a few centimeters from the ground or has 1 or 2 panels rolled back or horizontal slits to provide entrance for host-seeking mosquitoes (Service 1977). The person acting as bait can be enclosed within a fully protective inner net to prevent him from being bitten by mosquitoes. Mosquitoes trapped within the net can be collected either by the person acting as bait or by another person at intervals throughout the night.

Human-baited double net traps have been shown to have good trapping efficiency when compared with HLC in Asia (Tangena et al. 2015, Gao et al. 2018). In Lao PDR for instance, human-baited double trap collected a similar number of Anopheles mosquitoes as an outdoor HLC (Tangena et al. 2015). However, they have been found to be insensitive for sampling malaria vectors across different epidemiological settings in Africa (Service 1977, Le Goff et al. 1997). For example, double net traps collected 4 and 7.5 times lower number of Anopheles mosquitoes than the HLC in Nigeria and Cameroon, respectively (Service 1963, Le Goff et al. 1997).

In some studies, 2 persons were used to conduct a double net trap, i.e., 1 individual acting as a bait and the other as collector (Gao et al. 2018), and such approach makes the double net trap almost as labor-intensive as conducting the HLC. In another circumstance when 1 person is used both as bait and collector (Tangena et al. 2015), there might be a possibility of exposure to infectious mosquito bites during the collection process suggesting that these trapping methods need further modification in order to use them for routine malaria vector surveillance.

Human-baited double net/CDC light trap combination.

Human-baited double net/CDC light trap combination (HDNT) is an improved modification of the previously designed double net trap with an integrated CDC light trap (Tangena et al. 2015, Degefa et al. 2020). In this outdoor trapping method, mosquitoes attracted to the human bait are collected by setting a CDC light trap between the 2 nets. Field evaluations done in western Kenya and Ethiopia showed that this trap caught 6 times as many malaria vectors as the regular CDC light trap, and similar number of mosquitoes collected by the traditional HLC method (Degefa et al. 2020). Moreover, the HDNT showed a strong positive correlation with HLC, suggesting that it could represent an alternative tool to HLC for surveillance of outdoor biting malaria vectors in Africa.

Mosquito electrocuting trap.

Mosquito electrocuting trap (MET) consists of four 30 cm × 30 cm grid panels made of wooden frames that can be assembled into a square trapping box with the bottom and top open (Maliti et al. 2015). Stainless steel wires are embedded to run from the top to bottom of each frame at a spacing of 5 mm. A volunteer person sits on a stool with his lower legs positioned inside the trapping box. Adjacent embedded wires are differentially charged as negative or positive, such that mosquitoes approaching human bait will be shocked on contact with both wires. Knockdown mosquitoes due to the electric shock can easily be collected afterwards (Maliti et al. 2015, Meza et al. 2019, Sanou et al. 2019).

The MET is an exposure-free and promising alternative tool to HLC for surveillance of outdoor host-seeking malaria vectors (Maliti et al. 2015). The trap has showed positive correlation with the gold standard HLC (Sanou et al. 2019). In southeastern Tanzania, MET trap yielded higher density of An. arabiensis and An. funestus than HLC (Meza et al. 2019). However, it yielded significantly lower mosquito density compared to the HLC in some settings (Maliti et al. 2015, Sanou et al. 2019) (Table 3). Moreover, MET use may raise ethical concerns due to the possible risk of human contact with the electric grid (Maliti et al. 2015).

Conclusion and the Way Forward

In efforts to intensify malaria control through vector control and accelerate the progress towards elimination, the impact of the control measures needs to be evaluated regularly. This requires sampling vector populations using appropriate trapping methods to measure various entomological indices. The standard methods are HLC, PSC, and pit shelter for sampling host-seeking, indoor resting, and outdoor resting malaria vectors, respectively. However, these methods also have drawbacks such as exposure of collectors to infective mosquito bites, sampling bias, and feasibility issue. CDC light traps placed beside human-occupied bed nets have been used as an alternative to the HLC for sampling host-seeking malaria vectors. Efforts have been made to evaluate the CDC light traps against HLC to generate a conversion factor in order to use them as a proxy estimator of HBR and EIRs in Africa. However, a reproducible conversion factor was not found, indicating that the trapping efficiency of the CDC light traps varies between different geographical locations. Moreover, CDC light trap is not effective for surveillance of outdoor host-seeking malaria vectors. Several other alternative traps such as electronic aspirators, resting box, clay pot, sticky pot, tent traps, Mbita trap, different designs of odor-baited traps, bed net traps, HDNT, and MET have also been developed and evaluated in different settings but most of them need further standardization and/or modifications to improve their performance and/or to guarantee the safety of collectors. Among the alternative methods, HDNT and MET have the potential to replace the HLC for routine surveillance of host-seeking malaria vectors. For better estimation of host blood meal indices and HBR, PSC should be complemented with exit traps to capture both endophagic vectors, which are endophilic and endophagic vectors, which are exophilic. Moreover, indoor vector sampling methods should be paired with outdoor trapping methods during routine malaria vector surveillance to obtain a representative sample of both resting and host-seeking vector populations for unbiased estimation of the entomological indices, and to accurately evaluate the likely success of the current vector control measures. Further research is needed to optimize the alternative sampling methods and/or develop new surveillance tools for vector species such as An. stephensi.

Funding

This work was supported by grant from the National Institutes of Health (U19 AI129326).

Author Contributions

Teshome Degefa (Conceptualization [Lead], Data curation [Lead], Formal analysis [Lead], Investigation [Lead], Methodology [Lead], Project administration [Lead], Resources [Lead], Software [Lead], Validation [Lead], Visualization [Lead], Writing—original draft [Lead], Writing—review & editing [Lead]), Delenasaw Yewhalaw (Methodology [Equal], Writing—review & editing [Equal]), and Guiyun Yan (Methodology [Equal], Writing—review & editing [Equal])

References