Field Trials to Evaluate the Effectiveness of the Biogents®-Sweetscent Lure in Combination with Several Commercial Mosquito Traps and to Assess the Effectiveness of the Biogents-Mosquitaire Trap with and Without Carbon Dioxide

The growing spread of the invasive mosquito species Aedes aegypti (L.) and Ae. albopictus (Skuse) in the United States increases the complexity of mosquito control in the country. In contrast to several other mosquito species that are found in the home environment, including Culex quinquefasciatus Say, one of the vectors of West Nile virus, the container-inhabiting tiger mosquitoes are day active. Adulticidal operations at daytime are not feasible, and reaching all breeding sites in larviciding operations is practically impossible.

Maybe partially owing to the lack of a perceived effect of area-wide applied mosquito control measures, the lack of dedicated mosquito control programs, and because of concern about contracting mosquito-borne diseases, such as West Nile, dengue fever, chikungunya, and Zika, many American home owners use mosquito traps to reduce mosquito populations in their houses and backyards. Many mosquito traps with different attraction mechanisms and price categories are commercially available, but few scientific studies have been conducted to evaluate their effectiveness (e.g., Brown et al. 2008, Jackson et al. 2012). Furthermore, no studies have evaluated whether adding an attractant to commercially available traps can impact the capture of Asian tiger mosquitoes.

The Biogents® BG-Sentinel trap has been used by researchers for over a decade to monitor dengue vectors and is generally recognized as one of the best mosquito traps, especially for yellow fever and Asian tiger mosquitoes. This is reflected by the fact that the Centers for Disease Control and Prevention (CDC) recommends the use of the BG-Sentinel for Ae. aegypti and Ae. albopictus surveillance (CDC 2016) and that large-scale research projects, such as the World Mosquito Program (Eliminate Dengue no date) that operates in 12 countries, rely on BG-Sentinels for monitoring (Schmidt et al. 2017). The BG-Sentinel also catches high numbers of Cx. quinquefasciatus, especially when operated with CO₂ (Ázara et al. 2013). The trap uses the BG-Lure, an artificial human skin scent that contains lactic acid. The BG-Sweetscent was reformulated for home use and also contains lactic acid. It is packed in a sachet instead of a cartridge, has a shorter time span of effectiveness (up to 2 months, instead of up to 5 months), and is as efficient for catching Ae. albopictus as the BG-Lure (Akaratovic et al. 2017). The BG-Sentinel can additionally be operated with CO₂ to increase catch rates and species spectrum (Farajollahi et al. 2009), but even without carbon dioxide, it can catch more Ae. aegypti and Ae. albopictus than CO₂-baited encephalitis virus surveillance (EVS) traps (BioQuip®, Rancho Dominguez, CA) and CDC traps (Williams et al. 2006, Meeraus et al. 2008).

The BG-Mosquitaire trap uses the same mosquito attraction and collecting mechanisms as the BG-Sentinel, but instead of being lightweight and collapsible for scientific or public health purposes, it was designed to be more robust and visually pleasing, for use in fixed positions throughout the whole mosquito season in backyards, restaurants, hotels, and similar locations. The BG-Mosquitaire has been used in scientific studies before, including for monitoring of invasive mosquito species (Medlock et al. 2017) and studies of mosquito fauna (Beleri et al. 2017) and host preferences (Schönenberger et al. 2016), but an evaluation of the BG-Mosquitaire with and without CO₂ in comparison to the well-known BG-Sentinel and other commercial mosquito traps has not been published before.

The aim of the present study was to investigate whether Ae. albopictus and Ae. aegypti catch rates of commercially available mosquito traps could be increased when they are operated with the BG-Sweetscent lure, to demonstrate that the BG-Mosquitaire is as effective for catching Ae. aegypti and Cx. quinquefasciatus as the BG-Sentinel, and to compare the BG-Mosquitaire without CO₂ with the CO₂-baited BG-Mosquitaire, Mosquito Magnet, and SkeeterVac mosquito traps.

The efficacy of the BG-Sweetscent (Biogents AG, Regensburg, Germany) was tested in 2 field experiments, one in Gainesville, FL, and the other in Lake Charles, LA.

In Gainesville, 5 different traps were evaluated with and without BG-Sweetscent in July and August 2016: MosClean UV LED (ultraviolet light-emitting; Seoul Viosys Co., Ltd., Seoul, South Korea), Dynatrap DT2000XL (Dynamic Solutions Worldwide LLC, Milwaukee, WI), Bite Shield Protector (Koolatron, Ontario, Canada), Flowtron Galaxie PV 75 (Flowtron®, Malden, MA), and SkeeterVac Bite-Guard SVE6211 (Blue Rhino®, Winston Salem, NC). All of these electric traps use heat, UV light, and TiO₂ (to produce minimal amounts of CO₂ through a photocatalytic reaction) to attract mosquitoes. Insects are pulled into the traps by a suction fan and retained in a catch bag or chamber, where they die by dehydration. The SkeeterVac additionally uses a black-and-white sticky TacTrap as a supplemental capture system for attracted mosquitoes.

Five locations (distance between locations >50 m) were selected for the experiment, and 2 traps of the same type (1 with and 1 without BG-Sweetscent) were tested in 2 positions (minimum distance between positions = 10 m) at the same location. The traps' positions at each location were switched after 24 h, and after 4 days (e.g., after 4 repetitions on the same location), traps changed to another location, so that each trap configuration was tested 20 times in a modified Latin Square design.

In Lake Charles, 3 trap types were tested with and without BG-Sweetscent from July to September 2016 (peak population time for Ae. albopictus) in older city neighborhoods with mature trees and vegetation: Dynatrap DT2000XL, SkeeterVac Bite-Guard SVE6211, and Black Flag® BZ-40 (Spectrum Brands, Inc., Middleton, WI). The Black Flag BZ-40 is a bug zapper, which attracts mosquitoes by UV light and kills them by electrocution. The Black Flag trap was modified with an under tray to collect insects that were electrocuted. The traps were evaluated in a 6 × 6 Latin Square trial where trap positions were switched every 24 h.

The comparison of the BG-Sentinel and the BG-Mosquitaire took place in Clovis, CA, and in New Orleans, LA. In Clovis, 2 residential locations were selected, and at each location a BG-Sentinel version 1 (BG-Sentinel-1) with BG-lure cartridge was compared with a BG-Mosquitaire with BG-Sweetscent (all attractants and traps from Biogents AG). The traps were additionally baited with CO₂ from dry ice and operated for 20 h every 7 days for a total of 10 wk (end of July until beginning of September 2017). Traps at the same location were visually separated and had a minimum distance of 5 m between each other, and their positions were switched weekly.

In New Orleans, the BG-Sentinel version 2 (BG-Sentinel-2, Biogents AG) with BG-lure cartridge was compared with the BG-Mosquitaire with BG-Sweetscent at 2 routine surveillance sites: The New Orleans Police Department horse stables and a local backyard, both with lush vegetation. The traps had a minimum distance of 10 m between each other and were operated at fixed positions without CO₂ for 24 h once per week from September to November 2017.

This experiment was conducted in Lake Charles in August and September 2014 in older neighborhoods with mature vegetation. The BG-Mosquitaire without CO₂ and with BG-Sweetscent was compared with a CO₂- and BG-Sweetscent–baited BG-Mosquitaire, a SkeeterVac SV3100 (Blue Rhino, Winston Salem, NC) with BG-Sweetscent, and a Mosquito Magnet Patriot with Lurex3 (Woodstream Corp., Littiz, PA) in a 4 × 4 Latin Square trial. The SkeeterVac SV3100 and the Mosquito Magnet Patriot use a catalytic converter to generate electricity and CO₂ from propane. The electricity is used to run a fan that expels heat, moisture, and CO₂, as well as the smell of an optional lure, and to pull mosquitoes into the traps, where they are retained in a catch bag. The SkeeterVac SV3100 additionally uses a black-and-white sticky TacTrap as a supplemental capture surface. Of the 2 BG-Mosquitaire traps, 1 was baited with CO₂ from dry ice. Trap positions were changed every 24 h.

Statistical and exploratory data analyses were performed using R studio version 1.1.453 (Rstudio Team 2016), based on R version 3.3.2 (R Core Team 2016). Data from all experiments were analyzed, using generalized linear mixed models (GLMMs) and the libraries lme4 (Bates et al. 2015) and MASS (Venables and Ripley 2002). The fixed main effect was the variable trap, which included all trap configurations that were evaluated in each experiment. Owing to experimental design, the random factors position and date were chosen a priori in order to account for spatial and temporal correlation. The dependent count variable was either the number of female and male Ae. albopictus, Ae. aegypti, Cx. quinquefasciatus, or the total number of all Culicidae collected in trap t and on date d. Poisson models were first fitted, and when these were overdispersed, negative binomial models with a log-link were adjusted. Models' adequation was assessed through diagnostic residual plots and through evaluation of overdispersion. Tukey multiple comparisons of means of the GLMMs was performed using the emmeans package (Lenth 2018).

In Florida, Ae. albopictus females and males and Ae. aegypti females were collected by all traps, while only 3 out of 10 traps collected at least 1 Ae. aegypti male during the 20 trapping days (Table 1). All evaluated traps collected more total Ae. albopictus when they were baited with the BG-Sweetscent lure (Table 1; Fig. 1). The Poisson GLMM estimates that the SkeeterVac Bite-Guard collected on average 2 times more (P = 0.35), the MosClean 3.3 times more (P < 0.01), the Flowtron 2.4 times more (P = 0.98), the Dynatrap 2.4 times more (P = 0.01), and the Bite Shield 4.2 times more (P < 0.001) (Table 2). With the exception of the Flowtron, all traps also collected more Ae. aegypti when operated with the BG-Sweetscent lure (Table 1). No formal statistical analysis is presented on Ae. aegypti data, since residual plots of negative binomial GLMMs indicated major problems, probably due to the high number of zero observations (157 out of 197). In terms of total mosquitoes, the Bite Shield Protector, Dynatrap DT2000XL, and the MosClean UV LED collected more specimens when operated with BG-Sweetscent (Table 1); the differences, however, were insignificant (P > 0.9).

In Louisiana, all 3 evaluated traps collected more total Ae. albopictus when operated with the BG-Sweetscent lure (Fig. 1; Table 1). The negative binomial GLMM indicates that the Dynatrap collected 1.9 times more (P = 0.16), the SkeeterVac Bite-Guard 2.3 times more (P = 0.011), and the Black Flag collected 3.2 times more (P = 0.17) (Table 2).

In the BG-Sentinel version 1 versus BG-Mosquitaire trial in California, only Ae. aegypti were collected. Both CO₂-baited traps collected almost the same number of male Ae. aegypti, while the number of females collected was slightly higher in the BG-Mosquitaire (Table 1). Overall, the BG-Mosquitaire collected 1.2 times more Ae. aegypti (P = 0.175) than the BG-Sentinel (Table 2; Fig. 2A).

In New Orleans, where both traps were operated without CO₂, mainly Cx. quinquefasciatus and Ae. aegypti were collected (Table 3). The BG-Mosquitaire collected 1.2 times more Ae. aegypti (P = 0.62) than the BG-Sentinel (Fig. 2B), and the BG-Sentinel collected 1.1 times more Cx. quinquefasciatus (P = 0.84) than the BG-Mosquitaire (Table 2).

In this trial, the BG-Mosquitaire with CO₂ collected by far the highest number of Ae. albopictus, Cx. quinquefasciatus, and total mosquitoes (Table 4). For total Ae. albopictus (Fig. 3A), the CO₂-baited BG-Mosquitaire collected 11.9 times more (P < 0.0001) than the SkeeterVac Propane, 6.8 times more (P < 0.0001) than the Mosquito Magnet Patriot, and 4.7 times more (P < 0.0001) than the BG-Mosquitaire without CO₂ (Table 2). The BG-Mosquitaire without CO₂ collected 2.6 times more total Ae. albopictus than the SkeeterVac (P = 0.011), and the catch rates of the Mosquito Magnet and the BG-Mosquitaire were similar (P = 0.55). The Mosquito Magnet collected 1.7 times more than the SkeeterVac (P = 0.225). For total Cx. quinquefasciatus (Fig. 3B) and total Culicidae, the BG-Mosquitaire with CO₂ also collected significantly more specimens (3.6–17.8 times more, P < 0.001) than the 3 other evaluated traps, and the BG-Mosquitaire without CO₂ collected significantly more than The SkeeterVac and the Mosquito Magnet (1.8–4.4-fold; P < 0.05). The maximum number of species collected per 24 h trapping period was 3, 4, 7, and 12 for the Mosquito Magnet, SkeeterVac, BG-Mosquitaire, and BG-Mosquitaire + CO₂, respectively.

In this manuscript we demonstrated that the BG-Sweetscent lure increases the tiger mosquito catch rates of several commercially available mosquito traps. We also showed that the BG-Mosquitaire is as effective for catching Ae. aegypti and Cx. quinquefasciatus as the BG-Sentinel trap, a recognized gold-standard monitoring device for dengue vectors. Furthermore, the BG-Mosquitaire without CO₂ was proved to be as efficient for catching Ae. albopictus as the CO₂-baited Mosquito Magnet Patriot and significantly better than the CO₂-baited SkeeterVac SV3100. Adding CO₂ to the BG-Mosquitaire significantly increased its Ae. albopictus and other mosquito catch rates and increased the number of collected species.

The BG-Lure, which is essentially the same attractant as the BG-Sweetscent, was previously shown to increase the Ae. albopictus catch rates of BG-Sentinel (Pombi et al. 2014, Roiz et al. 2016) and CDC light traps (Urquhart et al. 2016), especially when used together with CO₂. The performance of the BG-Lure in homeowner mosquito traps, however, was not previously evaluated. Most of the traps that were evaluated in the Sweetscent efficacy trials collected low numbers of dengue vectors, but the Bite Shield Protector, Dynatrap DT2000XL, MosClean UV LED, and SkeeterVac Bite-Guard SVE6211 trapped considerable numbers of total mosquitoes (mainly Uranotaenia spp., Cx. erraticus, and Anopheles crucians). With the exception of the SkeeterVac Bite-Guard in 1 of 2 experiments, adding the BG-Sweetscent lure to these traps increased their total mosquito and dengue vector catch rates. Since all of the six traps that were evaluated in the BG-Sweetscent efficacy experiments use UV light as an attractant, they could be expected to have a higher number of insects other than mosquitoes in the catch. Studies by Surgeoner and Helson (1978), Nasci et al. (1983), and Frick and Tallamy (1996) on UV electrocution traps found only a small proportion of trapped specimens (<10%) as blood-sucking insects. It might be exactly this high amount of by-catch nontarget insects that makes trap owners satisfied, because what they perceive is that the trap catches a lot of “bugs.”

The BG-Mosquitaire was compared with the BG-Sentinel in 2 trials: In California, we used the BG-Sentinel-1, and both traps were operated with CO₂, while in Louisiana we used BG-Sentinel-2 and did not add CO₂ to the traps. The BG-Sentinel-1 was previously compared with version 2, and there was no difference between them regarding their Ae. albopictus catch rates (Arimoto et al. 2015, Akaratovic et al. 2017). In both present trials, the BG-Mosquitaire collected only slightly more Ae. aegypti than the BG-Sentinel, and in Louisiana, the BG-Sentinel collected slightly more Cx. quinquefasciatus, indicating that both traps are equally efficient for trapping these mosquito species.

In the present study, the catch rates and the species spectrum of the BG-Mosquitaire were substantially increased through addition of CO₂, in agreement with previous results for the BG-Sentinel (Farajollahi et al. 2009). The Ae. albopictus catch rates of the BG-Mosquitaire without CO₂ were still slightly higher than those of the Mosquito Magnet Patriot and significantly higher than those of the SkeeterVac SV3100, which are both operated with propane-generated CO₂. When considering all collected mosquitoes, the BG-Mosquitaire without CO₂ achieved significantly higher catch rates than the other 2 traps, and the species spectrum collected per 24 h was also higher. The Mosquitaire without CO₂, therefore, clearly outperformed 2 CO₂-baited traps. In agreement with our results, previous studies showed that the BG-Sentinel without CO₂ outperformed CO₂-baited EVS and CDC traps. Rochlin et al. (2016), however, found that the Mosquito Magnet baited with BG-Lure and R-octenol collected significantly more Ae. albopictus than a BG-Sentinel without CO₂ but with BG-Lure and R-octenol in Suffolk County, NY. This strong discrepancy between the results of our study might be due to the different Mosquito Magnet models that were used. We used the Mosquito Magnet Patriot, while Rochlin and coauthors used the Liberty and Executive models. Another big difference is that Rochlin and colleagues used R-octenol in addition to the BG-Lure. R-octenol is not a good attractant for Ae. albopictus in BG-Sentinel traps (Unlu et al. 2016) and the potential repellent effect of R-octenol might be stronger in the absence of CO₂ as an additional lure. Since the use of CO₂ is expensive and can be logistically challenging, it is of great importance to have a trap that catches high numbers of mosquitoes without it. The BG-Mosquitaire can be operated with CO₂, but unlike the Mosquito Magnet and the Skeetervac SV3100, it does not have to. In this way, it is much more flexible, and the question of adding CO₂ or not can always be adapted according to many factors, such as mosquito density and CO₂ availability.

We conclude that the BG-Sweetscent lure can be used in a wide spectrum of commercially available mosquito traps in order to increase their Ae. albopictus catch rates. Addition of the BG-Sweetscent is therefore recommended for improving trap performance, especially when aiming for increased dengue vector collections. We also conclude that the BG-Mosquitaire is as efficient as the BG-Sentinel, which is already recognized by researchers as one of the most effective mosquito traps for dengue vectors. Therefore, a trap of scientifically proved effectiveness is commercially available for private use. The BG-Mosquitaire can also be recommended for scientific and public health purposes, especially when the intent is to use the trap in a fixed position.

Biogents AG provided all traps and lures that were used in the experiments and paid a scholarship to a student who was involved in the field work and mosquito identification in Lake Charles.