New tools for Aedes control: mass trapping

Abstract

Aedes aegypti, the main vector of dengue, chikungunya, and Zika viruses uses artificial containers around homes to undergo immature development, making household-level detection and control extremely difficult in large urban areas. Mass trapping is an emerging methodology to control container-Aedes species such as Aedes aegypti and Aedes albopictus because effective traps for adult stages of these mosquitoes were developed recently. There are three main approaches to mass-trapping these mosquitoes: 1) Pull (attract/kill), 2) push (repel) - pull (attract/kill), and 3) pull (attract/contaminate/infect) - push (fly away). Effective mass-trapping depends on trap quality (capture efficiency, sturdiness, frequency of servicing), trap density and areal coverage, community involvement, and safety. Recent studies showed that Ae. aegypti populations can be sustainably controlled by mass trapping, although more area-wide studies showing effectiveness at preventing disease are needed for all trapping systems. Cost-effectiveness studies are needed for all emerging Aedes control approaches.

Introduction

Aedes aegypti is the main mosquito vector of dengue, chikungunya, and Zika viruses around the tropical and subtropical world [1]. There are several reasons why this mosquito is such an important vector, including its domesticity and high preference for biting people [2]. This mosquito species has access to shelter inside homes, human blood to produce eggs, and indoor/outdoor container aquatic habitats that are needed for immature development. Habitats that overlap between Ae. aegypti and people facilitate epidemics of arbovirus in urbanized areas. A main challenge to controlling arboviral epidemics consists of keeping the abundance of this mosquito species below levels that would prevent or control the transmission of arboviruses (mosquito density threshold) [3].

Defining and validating mosquito density thresholds partially depends on the tools used to assess mosquito abundance (e.g., immature surveys, counts of adult mosquitoes landing on humans, ovitraps, adult mosquito traps) and the stage of the mosquito (e.g., eggs, larvae, pupae, adults) [4]. Mosquito density thresholds provide well-defined targets for Mosquito Control Programs. The relatively recent development of new and effective mosquito traps targeting Ae. aegypti adults provides opportunities to improve its surveillance and control. A recent review of evidence recommended mass trapping of gravid females of Ae. aegypti using newer generation, larger traps that compete with naturally occurring aquatic habitats [5].

Trapping systems

A trap is defined by WHO [6] as a “Structure or device unto which vectors enter and/or make contact with, which ultimately results in their capture, death and/or sterilization”. Currently, there are three main control approaches using trap devices: 1) pull (attract/kill) (e.g., ovitraps, adult mosquito traps, attractive toxic sugar baits), 2) push (repel) - pull (attract/kill) (e.g., use of spatial repellent devices and adult mosquito traps) , and 3) pull (attract/contaminate with a control agent) - push (fly away to disseminate a control agent) (e.g., autodissemination devices impregnated with an insect growth regulator product that is dispersed by ovipositing females in aquatic habitats nearby).

Pull (attract/kill).

1. Ovitraps.

These are small dark containers (e.g., black, red) made from different materials (e.g., glass, metal, plastic, rubber) containing water (e.g., 0.25 – 2 lt) or water with decomposing organic material (e.g., hay, leaves, yeast), and a substrate to collect mosquito eggs (e.g., wooded tongue depressor, germination paper, cloth). Ovitraps were produced to detect the presence of Ae. aegypti during the eradication campaign of this mosquito in the Americas [7]. Ovitraps target eggs from ovipositing gravid female mosquitoes, so that mass ovitrapping intends to reduce the fecundity of the Aedes population. A principal limitation of small ovitraps as an area-wide mass trapping strategy is the need for servicing them frequently to replace water, attractants, and oviposition substrates, but also to prevent hatched larvae from becoming adults. The latter issue has been addressed by adding larvicides that do not repel ovipositing females (Bacillus thuringiensis var. israelensis, spinosad, novaluron, S-methoprene, yeast interfering RNA) [8,9]. Chan et al. [10] developed an autocidal ovitrap to control Ae. aegypti in Singapore that prevented the emergence of adult mosquitoes developing inside the ovitrap by mechanical means (asphyxiation). The authors recommended the use of autocidal ovitraps along with source reduction to maximize control. There are no recent publications that address controlling Aedes species using ovitraps and earlier studies on mass trapping were reviewed by Johnson et al. [5]. Currently, insecticidal traps are not considered for mass trapping Aedes species.

2. Gravid-female adult traps.

Like ovitraps, gravid traps are dark, although some other colors such as terracotta and dark blue resulted in similar attraction to Ae. aegypti in field studies in Puerto Rico [8]. Gravid traps are usually bigger than ovitraps (e.g., 2 - 10 1) and contain water with decomposing organic material to attract gravid females looking for a place to lay eggs. Adult mosquitoes are retained on a sticky glue board [11,12], killed with a residual insecticide [13], or impaired to fly with canola oil [14]. It is important to monitor and control the gravid adult female mosquito population because those females must have had a blood meal to produce eggs, and therefore, it is the most likely stage of the mosquito to be infected with arboviruses. Additionally, monitoring the presence of arboviruses in gravid Ae. aegypti mosquitoes has been useful to detect local transmission of chikungunya, dengue, and Zika viruses [15,16].

There are a several recent studies showing significant effectiveness of mass trapping with gravid traps at reducing: Ae. aegypti populations [17,18,19], prevalence of arborvirus in mosquitoes [15], and risk of human infections with chikungunya [50%; 20] and dengue viruses [36%; 18]. Non insecticidal gravid traps have been used to propose Ae. aegypti density thresholds in Puerto Rico (e.g., < 3 females Ae. aegypti/trap/week) to prevent chikungunya and Zika transmission [15,21]. Mass trapping has produced relatively stable, below thresholds densities in small communities for several years and in a medium-size city, as shown by a cluster randomized step-wedge intervention during the 2016 Zika epidemic in Puerto Rico [17]. Modeling has shown that a steady reduction of 70-80% of the female Ae. aegypti population, such as that observed in these studies does not require removing 70-80% of mosquitoes. The reason for such population reduction, other than directly eliminating a fraction of the gravid females, is a significant reduction of the average longevity of the mosquito female population [22]. So far, there are no reports that local control of gravid females of Ae. aegypti by non-insecticidal mass trapping for several years had caused the evolution of trap’s avoidance behavior by Ae. aegypti. A recent investigation tested the hypothesis that long-term mass trapping of gravid Ae. aegypti would eliminate insecticide resistance in populations that were initially resistant to commercially available domestic insecticides. It was thought that insecticide susceptibility could be restored (loss of resistant genes) from processes happening in small populations such as genetic drift, bottle neck events, or lack of adaptive selection if residents discontinued the use of insecticides [23]. The results showed that mass trapping did not restore insecticide susceptibility and provided evidence of high Ae. aegypti migration from nearby resistant populations. These results highlight the importance of scale on dynamic processes involving factors (e.g., migration) that can only be assessed by large area-wide studies of mosquito population control.

Several factors contribute to the effectiveness of mass trapping for controlling Ae. aegypti and perhaps, other container mosquitoes: trap efficiency (capacity to attract and retain, kill, or contaminate nearby mosquitoes), effective number of traps per house or area (e.g., three traps/home), good areal coverage (e.g., > 60-80% of houses/area with traps), long periods without trap servicing, timely trap maintenance, acceptance from the community, an efficient system to collect real time data in the field using computer applications in cell phones or tablets to monitor quality control, and use of Geographical Information Systems (GIS) to keep track of traps’ location and condition. Additionally, Johnson et al. [5] recommended involving residents in trap servicing for sustainability [24], avoiding the use of insecticides against landing female mosquitoes given current widespread insecticide resistance in Ae. aegypti and using organic larvicides to prevent the production of mosquitoes in unattended traps. Traps requiring frequent maintenance (e.g., < 2-3 mo.) will have high staffing costs. Community involvement could lower cost of mass trapping if traps can be fabricated locally and are durable and easy to maintain. A general need for all emerging new tools for the control of Ae. aegypti that applies to mass trapping, is cost-benefit studies to understand if they are affordable and if they have a significant impact on reducing or preventing arbovirus transmission and disease, particularly in lower-income settings.

3. Attractive toxic sugar baits (ATSBs).

Adults of container Aedes sp. attracted to a source of sugar can be killed using ingested toxic compounds (e.g., boric acid, fipronil, Ribonucleic Acid interference or RNAi) [25,26]. Baits can be applied directly to vegetation and other surfaces that are frequented by adult Aedes mosquitoes or be contained in a bait station. Floral attractants and fruit juices have been explored as attractants for Ae. aegypti. Frikig et al. [27] reported that fruit juices were more effective than water at luring males of Ae. aegypti, although none of these lures were effective at attracting male mosquitoes to an insecticidal adult trap. Sippy et al. [28] used a dry ATSB consisting of black/white foam disks sprayed with sugar and boric acid to attract and kill indoor Ae. aegypti females. Reported cumulative mortality of adult Ae. aegypti released in experimental houses with dry ATSB in Machala, Ecuador varied between 77-100% after 48 h exposure. Dry ATSBs stations that do not rely on chemical attractants or liquid sugar/toxic solutions seem to be advantageous in terms maintenance, price, and persistence.

Sugar utilization of container Aedes species in urbanized areas, and therefore the potential usefulness of ATSBs, seems to vary widely depending on location and degree of urbanization [29,30,31]. The use of sugar from plants by Ae. aegypti was investigated in two neighborhoods with different degrees of urbanization and availability of sugar sources in Mali. They found 39-40% sugar feeding in sugar-poor sites and 60-65% in sugar-rich sites [32]. They also showed effective control of the local Ae. aegypti population during 50 days after spraying the vegetation with ATSB with micro-encapsulated garlic oil. Revay et al. [33] found higher reductions of the Ae. albopictus populations in tire sites in Florida by foliar applications of ASTB and eugenol than by using bait stations. More research is required to develop efficient bait stations that would restrict access only to target species. Like other trapping approaches, there is a need to conduct area-wide, cluster randomized trials with entomological and epidemiological outcomes using the ASTB methodology.

Push (repel) - pull (attract/kill).

The basis for this mosquito control approach is to repel mosquitoes away from an area (push) and trap and eliminate pushed-away mosquitoes (pull). The expected benefits from this and other mass-trapping approaches are reducing both biting rate and mosquito abundance to prevent or control exposure to vector-borne pathogens. Recent studies explored the use of various types of traps and the human landing technique to assess protection of a push-pull semi-field evaluation using transfluthrin as the spatial repellent for Ae. aegypti [34,35]. Some relevant findings were that the push (repellent) aspect of the system had higher efficacy (protection against bites) than the combined push-pull aspect (repellent + traps) of the trials. They also found that trap efficiency markedly decreased in the presence of humans, which may affect how these types of trials are performed or evaluated with mosquito traps. Area-wide studies on controlling Ae. aegypti using the push-pull approach have not been reported so far.

Pull (attract/contaminate/infect) - push (fly away).

Devices attracting container-Aedes female mosquitoes are treated with an insect growth regulator (IGR) that adheres to female mosquitoes when they land on the contaminated surface, so that they transfer the chemical to other containers while ovipositing, to suppress immature mosquito development [36]. Thus, the purpose of auto-disseminating devices is not to readily kill the female mosquito but to use it to amplify suppression of mosquito production in nearby containers with water. Commonly, auto-dissemination devices are similar to ovitraps that attract gravid females. One study used an electromechanical trap to disseminate pyriproxyfen for the control of Ae. aegypti in Madeira, Portugal [37]. The most used IGR is pyriproxyfen, a juvenile hormone analog that interferes at extremely low dosages with immature development and metamorphosis, preventing the emergence of adult mosquitoes [36]. Accumulation of IGR in containers in time by repeated visits of contaminated female mosquitoes increases the effectiveness of mosquito suppression [38]. A recent review on the use of this technique concluded that it is a promising new approach and a valuable addition to the vector control toolbox [39]. Observed Aedes mortality in small field studies ranged 50-92% [38,40,41]. Neighborhood-level studies have shown mixed results varying from relatively ineffective Aedes population suppression [42,43] to highly effective control as measured by immature emergence suppression [40], reduced oviposition or adult mosquito abatement [44]. An additional effect of the exposure of Aedes to pyriproxyfen is reduced fertility and fecundity [45]. Autodissemination of pyriproxyfen has been combined in one device with spores of the entomopathogenic fungus Beauveria bassiana, to reduce the vectorial capacity and slowly kill the adult mosquito [46]. A field trial conducted at the neighborhood level in Florida, USA using these devices showed significant reductions in eggs and larvae in sentinel ovitraps, and a borderline, non-significant reduction of adult mosquitoes captured in electromechanical traps [47].

There are several aspects of this approach that may need more research, such as whether females visiting the traps could still bite people and transmit pathogens as opposed to directly killing the visiting females. A positive expected result of this approach is that contaminated females would be able to reach cryptic aquatic habitats that are not possible to locate or treat using other means, such as source reduction or larvicides. Because this approach depends on the number of contaminated adult mosquitoes visiting the devices, as the mosquito population goes down it is expected to lose effectiveness and perhaps reach a mosquito density threshold below which further suppression is not possible. It would be important to understand if such a theoretical mosquito density threshold is sufficient to prevent local outbreaks of arboviruses or if a combination of this approach along with other mosquito control measures can reduce the mosquito population to safe levels.

Conclusions

There is a growing number of studies aimed at controlling container-Aedes species by mass-trapping using a diversity of approaches to lure, capture and eliminate, or to use mosquitoes as means of disseminating control agents to abate its own population. A few mass-trapping non-randomized studies have shown that the density of Ae. aegypti can be kept at very low levels for several years and that people living in areas with mass trapping had significantly lower prevalence of arboviral antibodies [20]. Several novel mass-trapping approaches have shown promising results in laboratory and small field studies but there is a need to conduct larger studies, ideally using robust experimental designs with entomological and epidemiological outcomes. Similarly, studies on the cost-effectiveness of mass-trapping are needed to evaluate their feasibility and effectiveness. These needs also apply to other emerging Ae. aegypti control tools.

Table.

Examples of recent trials on trapping systems for the control of container Aedes spp.

Trapping system	Species	Type of study	Location	Type of mosquito control	Results	Reference
Pull (attract/kill)	Aedes aegypti	Cluster randomized step-wedge, implemented during a Zika epidemic.	Most residential and commercial building in Caguas city, Puerto Rico.	Mass-trapping with 3 sticky gravid traps/home in 60-80% of homes, and limited community education, larviciding and source reduction.	Achieved steady control below mosquito density threshold (2-3 female Ae. aegypti/trap/week. No Zika virus present in Ae. aegypti when 60-80% of the houses were treated.	[17]
		Controlled before – after intervention.	34 treatment sites in Singapore.	One sticky gravid trap per 20 households in apartment buildings.	36% reduction in dengue cases.	[18]
		Cluster randomized crossover.	2 low and 2 middle-income communities in Hildalgo and Cameron counties, Texas.	Mass-trapping with sticky gravid traps.	77% reduction of mosquitoes when trap coverage was 2.7 traps/house.	[19]
		Semi-field trials in experimental houses.	Machala, Ecuador.	Dry attractive toxic sugar bait.	54-98% and 77.3-100% mortality of adult Ae. aegypti when exposed for 24 and 48 h, respectively. Most mortality occurred within 48 h.	[28]
Push (repel) - pull (attract/kill)	Ae. aegypti	Semi-field randomized block design.	Bagamoyo, Tanzania.	Push: 2 Freestanding transfluthrin passive emanator (FTPE) and Pull: 1 Electromechanical trap / block.	Protection efficacy: 61.2% FTPE alone, 2.1% Trap alone, and 64.5% FTPE + Trap.	[34]
Pull (attract/contaminate/infect) - push (fly away)	Ae. albopictus	Controlled before – after intervention.	3-4 mosquito hot-spot city-blocks in control and interventions areas, Mercer county, New Jersey.	26-28 Autodissemination stations / treatment block.	No significant reductions in egg or adult populations.	[42]
	Ae. aegypti / Ae. albopictus	Before – after intervention..	3 dengue hotspot high-rise buildings in Selangor, Malaysia	356 – 552 autodissemination stations.	Ovitrap prevalence was 44.8% before intervention and 53.4% after intervention.	[43]
	Ae. aegypti / Culex quinquefasciatus	Cluster-randomized controlled trial.	2 neighborhoods in Federal District, Brazil.	1 autodissemination station per 10 houses. Routine ultra-low volume spraying of insecticides in both neighborhoods.	Reductions of 60.0% in Ae. aegypti and 55.5% is adult Cx. quinquefasciatus mosquitoes. No measurable effect on Ae. aegypti eggs.	[44]
	Ae. aegypti	Comparison of Autodissemination devices with integrated vector management (source reduction, larviciding, adulticiding).	2 suburban neighborhoods in Manatee County, Florida.	15 autodissemination devices/ha and entomopathogenic fungus.	Significantly fewer eggs and larvae were observed in the sites with autodissemination stations.	[47]