Efficacy and Spatial Extent of Yard-Scale Control of Aedes (Stegomyia) albopictus (Diptera: Culicidae) Using Barrier Sprays and Larval Habitat Management

Abstract

The Asian tiger mosquito, Aedes (Stegomyia) albopictus (Skuse), is a peridomestic, container-ovipositing mosquito commonly found throughout the southeastern United States. In the United States, Ae. albopictus is typically considered a nuisance pest; however, it is capable of transmitting multiple pathogens. Ae. albopictus is an important pest species and the target of numerous mosquito control efforts in the United States. Here, we evaluate the effectiveness and spatial extent of Ae. albopictus population reduction using a bifenthrin (AI Bifen IT, 7.9%) barrier spray and larval habitat management (LHM) in a temperate, suburban setting. Sixteen pairs of adjoining neighbors were randomly assigned to treatment groups with one neighbor receiving a treatment and the other monitored for evidence of a spillover effect of the treatments. Ae. albopictus populations in both yards were monitored for 33 d, with treatments occurring on the eighth day. Barrier sprays, both alone and combined with LHM, resulted in a significant reduction in Ae. albopictus abundance posttreatment. While LHM alone did not result in a significant reduction over the entire posttreatment period, Ae. albopictus populations were observed to be in decline during this period. No treatments were observed to have any reduction in efficacy 25 d posttreatment, with treatments involving LHM having a significantly increased efficacy. Yards neighboring treated yards were also observed to have reduced population sizes posttreatment, but these differences were rarely significant. These results provide insights into the population dynamics of Ae. albopictus following two common treatments and will be useful for integrated pest management plans.

Aedes (Stegomyia) albopictus (Skuse) is a peridomestic container-ovipositing mosquito commonly found throughout the world. After its initial introduction in Texas in 1985 (Sprenger and Wuithiranyagool 1986), it quickly spread throughout the southeastern United States and now has a range stretching from the Atlantic coast in the east to Texas in the west and from the southern tip of Florida up to New York in the north, along with areas of recent introduction in California on the west coast (Hahn et al. 2017). Its preference for ovipoisiting in artificial containers, along with its propensity for biting humans, makes it one of the most common nuisance mosquitoes in the United States. Unfortunately, controlling Ae. albopictus using traditional methods such as ultra-low volume applications has proven difficult, suggesting a need for alternative control strategies (Roiz et al. 2018, Achee et al. 2019). Further, its ability to transmit dengue, Zika, and chikungunya viruses make it a potential public health threat, and the likely vector in the case of introduction of these pathogens into the United States (Gostin and Hodge 2016, Messina et al. 2016, Moreno-Madriñán et al. 2018).

Traditionally, mosquito control in the United States has been performed by mosquito abatement districts, with the aim of mitigating community-wide nuisance and risk of disease associated with mosquitoes. Abatement districts commonly deploy vehicle-based, ultra-low volume (ULV) application in order to treat large areas with insecticides. These ULV applications are often coupled with source reduction efforts to remove large oviposition sites, such as tire piles. However, recent budgetary cuts, including in North Carolina in 2011, have resulted in underfunded and underprepared mosquito control districts (Del Rosario et al. 2014).

Within North Carolina, 55% of mosquito control programs self-reported a barely functional budget (Del Rosario et al. 2014) and a recent survey of vector control organizations nationwide showed that 84% needed improvement in at least one of five core capacities: routine mosquito surveillance, surveillance-based treatments, larviciding and/or adulticiding, routine vector control activities, and pesticide resistance testing (NACCHO 2017). These reports, combined with concerns about the effectiveness of ULV applications for reducing Aedes populations, and Aedes-vectored disease (Bonds 2012, Wilson et al. 2015, Bowman et al. 2016, Faraji and Unlu 2016, Roiz et al. 2018), suggest mosquito abatement districts may be ill-equipped to respond to vector-borne disease outbreaks and highlight a need for new plans for mosquito-borne disease response. The private mosquito control industry, in contrast, has grown significantly over the past decade (Specialty Consultants 2017). While this growth has been mostly driven by a high willingness to pay for mosquito nuisance control (Dickinson and Paskewitz 2012), fear of potential outbreaks of dengue and Zika virus in the United States has likely also increased enrollment in their services. Unlike mosquito abatement districts, these companies deploy yard-scale mosquito treatments, the efficacy and spatial scale of which has not been fully quantified. It is possible that, if shown effective, these yard-scale treatments could provide a useful tool for precision mosquito control in the event of a mosquito-borne disease outbreak. In light of this, it is important to quantify the short-term effectiveness of yard-scale treatments and their effect on the dynamics of nearby mosquito population.

The yard-scale treatments employed by private mosquito control companies typically target nuisance mosquitoes, including Ae. albopictus, with treatments involving the use of a combination of an adulticidal barrier spray, most commonly a pyrethroid, and larval habitat management (LHM), tip-and-toss combined with a long-acting larvicide, applied regularly during mosquito season. The success of these companies and the growth of the industry provide anecdotal evidence of their effectiveness. However, empirical studies of the effects of barrier sprays (Muzari et al. 2014, Fulcher et al. 2015, Stoops et al. 2019), especially when applied to individual yards (Trout et al. 2007, Hurst et al. 2012, VanDusen et al. 2016, Richards et al. 2017a), have focused on the long-term effects of the treatments inside the treated area, and not on the short-term effects or the spillover effects in the neighboring yard, which would likely play a major role in disease dynamics. Similarly, studies of the effect of LHM, or larval source reduction, on Ae. albopictus have typically focused on neighborhood or citywide efforts (Richards et al. 2008; Unlu et al. 2011, 2013; Fonseca et al. 2013; Faraji and Unlu 2016), with mixed results and further investigation about its effectiveness is needed (Faraji and Unlu 2016).

Here, we report on the efficacy of two standard mosquito control techniques, LHM, through source reduction, and an adulticidal barrier spray, when applied to individual yards in the Wake County, NC area. Mosquito abundance from adjacent yards was also monitored to determine if there was a reduction outside of the treated area.

Methods

Participant Recruitment

Households were recruited in Wake County, NC (35°47′24.00″N, −78°39′0.00″W) beginning in May 2018 using a combination of recruitment fliers placed at local community centers and recruitment e-mails sent through neighborhood listservs. Recruitment was limited to pairs of neighboring houses that had not received any professional mosquito treatment that year (since January 2018) and were willing to participate. Participating houses were enrolled on a first-come basis and adult residents for each house were met in person for the informed consent process. All participants received treatments free of charge, and participants not receiving a barrier spray treatment were offered a free treatment following the end of the study. Designation of treatment and neighboring house for each pair was determined at time of informed consent. Typically, the house belonging to the primary contact for the pair, whoever initiated contact, was assigned to treatment unless residents preferred otherwise.

Ethical Clearance

Informed consent was obtained from adult residents of all participating houses before the beginning of the study. All participants were informed about their rights and all risks associated with their inclusion in the study. Ethical approval was obtained from the North Carolina State University Institutional Review Board (Approved, NCSU Protocol # 12800, 14 May 2018).

Study Design

We employed a split-split-plot design with sites, consisting of 16 house pairs, randomly assigned to a LHM by barrier spray (2 × 2) balanced factorial design and trap locations designated within each site. Each house pair was assigned to receive either LHM alone, the barrier spray alone, a combination of both LHM and barrier spray, or no treatment, with four replicates of each treatment. Three sampling locations were determined within each site (see Fig. 1). Each location within a site was sampled concurrently every fourth day over a 33-d period.

Fig. 1.

Trap locations within the treated and neighboring yards. The center trap was placed near the center of the treatment yard. The side trap was then set in the treatment yard, approximately 5 m from the edge of the neighboring yard, and the neighbor trap was placed approximately 15 m into the neighboring yard to form a transect.

After being assigned a treatment, each site was assigned to one of four groups for sampling, such that each group contained exactly one site assigned to each of the treatment levels. Sampling was conducted on a rotating basis, with all members of a group sampled on the same days, e.g., group 1 was sampled on days 1, 5, 9, etc. and group 2 on days 2, 6, 10, etc. Sampling consisted of nine trap days at each location occurring on every fourth day, lasting 24 h, and occurred over a period of 36 d, including the pretreatment period, from 6 August 2018 to 10 September 2018 due to rotating between groups.

During each sampling day, three BG-Sentinel 2 (Biogents, Regensburg, Germany) traps were set for 24 h along a transect, with traps placed in the center and side (approximately 5 m from the property line) of the treatment yard and adjacent neighboring yard (approximately 15 m from the property line), all traps were baited using BG lures (Fig. 1). Trap bags were collected at the end of the sampling period and collections frozen until sorted, with numbers of female Ae. albopictus, male Ae. albopictus, and other mosquitoes recorded for each trap day. Due to low count numbers of other mosquito species in collections, only female Ae. albopictus counts are reported here.

Insecticide Application

Insecticide applications occurred the day before the third round of sampling at each site, e.g., day 8 for group 1 and day 9 for group 2 (13 August to 16 August 2018), denoted day 0 (gray dashed bar in Fig. 1), and were performed between the hours of 1400 and 1700. Treatments were carried out by a trained, licensed applicator from a local mosquito control company. For houses receiving LHM or combined treatment, the yard was initially surveyed for larval habitat by the applicator and primary researcher. All containers with standing water were emptied and removed if possible and the larvicide, Altosid (AI 1.5% (S)-Methoprene, Zoecon, Schaumburg, IL), was applied to any standing bodies of water that could not be removed and did not contain fish, e.g., birdbaths. In yards that received the barrier spray or combined treatment, the applicator applied Bifen IT (AI 7.9% bifenthrin, Control Solutions, Pasadena, TX) as a barrier spray around the property and to any resting habitat, e.g., dense foliage, with care taken to avoid any flowering or fruiting plants and any ponds, consistent with the EPA/FIFRA pesticide label. Barrier sprays were performed throughout the front, back, and side of the yard and to any fences using a Stihl SR450 mist blower with the applicator walking approximately 3–4 km/h, applying approximately 1 gallon of mixed product, containing approximately 1 oz Bifen IT, per 1,000 ft². For houses assigned to the combined treatment group, LHM was performed before application of the barrier spray.

Statistical Model

To evaluate the reduction in the female Ae. albopictus population due to the treatments, we modeled the number of female Ae. albopictus in the trap as a Poisson random variable using a generalized linear mixed model with a log-link function (Bolker et al. 2009, Zuur et al. 2010). All models included a random intercept for the study site, with correlation between traps at the same site, the trapping date, and an observation level random effect to account for overdispersion (Harrison 2014). Fixed effects were modeled as a LHM by barrier spray by location by days since treatment factorial design, with day taken as a covariate for comparisons between treatments and locations and as a factor for comparisons between days. The average pretreatment mosquito count was then used as a per location offset. All models were fit in R (R Development Core Team 2019) using the lme4 package (Bates et al. 2015), with means and contrasts calculated using the emmeans package (Lenth 2019).

The percent reduction in mosquito population due to treatment is found as the contrast between the treatment and control group ((1 − contrast) * 100%). Multiple comparisons against a control, either the untreated yards or the first day post-control, were performed using a Dunnett adjustment. Pairwise multiple comparisons were conducted using a Tukey adjustment, while comparisons of individual days against the average of the days used a Bonferroni method (Longnecker and Ott 2015). Where not specifically noted, we test the contrast been a treatment and the untreated group using the appropriate adjustment. Confidence intervals for all estimates are given at the 95% level and contrasts are considered significant at the α = 0.05 level.

Results

Aedes albopictus counts for 429 trap days were recorded (3 d, one of which was pretreatment, were lost due to trap failures) (Fig. 2). There was no significant overall trend with regard to mosquito abundance for the control group (95% CI for slope [−0.0311, 0.0010]) over the study period, suggesting that mosquito populations were not declining throughout the study period. However, counts in the posttreatment period were significantly lower than the pretreatment period for the control group (Supp Table 1 [online only]). Mean counts for each group are reported in Table 1.

Fig. 2.

Mean number of female Ae. albopictus for each treatment and location combination. Treatment occurred on day 0 (vertical dashed line). The average counts suggest an effect of the combined and barrier spray treatments.

Table 1.

Mean count of Ae. albopictus on given day after treatment, aggregated for each trap location, yard, and treatment pair within each treatment group

		Pretreatment	Day
			1	5	9	13	17	21	25
Combined	Total	10.08	5.33	4.58	2.75	3.50	3.91	3.25	2.17
	Treated	10.625	6.25	4.75	3.25	3.75	4.50	3.38	1.75
	Center	12.12	8.00	8.25	5.00	4.25	7.00	4.00	1.25
	Side	9.13	4.50	1.25	1.50	3.25	2.00	2.75	2.25
	Neighbor	9.00	3.50	4.25	1.75	3.00	2.75	3.00	3.00
Barrier spray	Total	20.63	7.00	5.92	6.75	6.67	5.58	6.83	5.08
	Treated	15.25	5.00	4.25	4.38	4.75	4.13	7.25	3.50
	Center	13.63	1.50	2.50	3.75	2.75	3.75	4.00	2.50
	Side	16.88	8.50	6.00	5.00	6.75	4.50	10.50	4.50
	Neighbor	31.38	11.00	9.25	11.50	10.50	8.50	6.00	8.25
LHM	Total	14.17	12.75	5.75	4.33	7.25	7.25	6.17	3.33
	Treated	13.88	10.88	6.00	4.50	8.13	7.75	5.38	3.13
	Center	9.38	8.50	5.25	4.50	6.50	5.75	6.50	3.00
	Side	18.38	13.25	6.75	4.50	9.75	9.75	4.25	3.25
	Neighbor	22.00	16.50	5.25	4.00	5.50	6.25	7.75	3.75
Control	Total	14.17	16.67	12.5	8.45	13.17	6.67	13.64	8.42
	Treated	13.75	12.00	11.25	6.86	13.00	6.13	13.86	9.00
	Center	11.13	10.75	6.75	4.00	6.25	7.50	5.33	4.50
	Side	16.38	13.25	15.75	9.00	19.75	4.75	20.25	13.5
	Neighbor	15.14	26.00	15.00	11.25	13.50	7.75	13.25	7.25

Treatment Effects

Overall, we found that all treatment groups exhibited a decrease in the mosquito count during the entire posttreatment period. Compared to the pretreatment mean, pairs of houses in the LHM treatment group had a 63.9% (47.5%, 75.1%) mean reduction in the abundance of female Ae. albopictus, the barrier spray treatment group had a reduction of 77.1% (66.8%, 84.2%), and the combined treatment group had a reduction of 74.6% (62.8%, 84.2%) (Supp Table 1 [online only]). The application of LHM resulted in a 40.9% (P = 0.0866) mean reduction in the abundance of female Ae. albopictus across the pair of yards over the posttreatment period, the barrier spray resulted in a 62.5% (P = 1.87 × 10⁻⁴) mean reduction, and combining the treatments resulted in a 58.5% (P = 0.00123) mean reduction.

Inside the treated yards, LHM resulted in a 33.3% (P = 0.457) mean reduction in female Ae. albopictus abundance over the posttreatment period, the barrier spray in a 62.2% (P = 0.00808) mean reduction, and the combined treatment in a 59.5% (P = 0.0176) mean reduction. In the untreated neighboring yard, LHM resulted in a 54.6% (P = 0.209) mean reduction over the posttreatment period, the barrier spray in a 62.5% (P = 0.0796) mean reduction, and the combined treatment in a 58.6% (P = 0.143) mean reduction (Table 2).

Table 2.

Percent reduction in adult female Ae. albopictus population due to treatment and P-values

		Pair	Treated	Neighbor
Combined	Posttreatment	58.5 (0.0012)*	59.5 (0.0176)*	58.6 (0.143)
	Day 1		37.6 (0.534)	72.4 (0.0865).
	Day 5		58.9 (0.109)	58.1 (0.330)
	Day 9		46.9 (0.376)	82.8 (0.133)
	Day 13		63.1 (0.0539).	65.3 (0.209)
	Day 17		40.0 (0.527)	35.1 (0.805)
	Day 21		69.2 (0.0223)*	60.6 (0.304)
	Day 25		81.2 (0.00233)*	28.0 (0.882)
Barrier spray	Posttreatment	62.5 (1.88 × 10−4)*	62.2 (0.00808)*	62.5 (0.0796)
	Day 1		60.7 (0.0717)	68.9 (0.0921)
	Day 5		65.4 (0.0351)*	67.3 (0.120)
	Day 9		47.0 (0.351)	37.0 (0.730)
	Day 13		73.4 (0.00524)*	64.8 (0.1626)
	Day 17		53.7 (0.205)	49.3 (0.514)
	Day 21		57.9 (0.102)	77.9 (0.0275)*
	Day 25		67.3 (0.0288)*	53.9 (0.412)
LHM	Posttreatment	40.9 (0.0866).	33.3 (0.457)	54.6 (0.209)
	Day 1		−04.3 (0.996)	35.1 (0.754)
	Day 5		25.5 (0.795)	63.8 (0.202)
	Day 9		11.0 (0.972)	53.9 (0.431)
	Day 13		35.4 (0.576)	66.3 (0.168)
	Day 17		00.2 (1.00)	38.4 (0.741)
	Day 21		60.8 (0.0857).	59.2 (0.304)
	Day 25		63.7 (0.0668).	53.5 (0.461)

Percent reduction is found as the contrast between the reduction seen in the treated and control yards for each treatment and time period. Differences that are significant at the α = 0.1 (.) and α = 0.05 (*) are denoted.

Edge and Spillover Effects

In general, Ae. albopictus counts were higher at the side and neighbor trap locations for the barrier spray group and lower for treatments that included LHM. However, we found no significant difference in the reduction of Ae. albopictus counts between the traps at the center, side, and neighboring locations, regardless of treatment (Supp Table 2 [online only]).

Temporal Trends

The day immediately following treatment application, reduction due to treatment in the mean female Ae. albopictus abundance in treated yards was not significant for any treatment, with the largest reduction, 60.7% (P = 0.0717), occurring when only the barrier spray was applied. Similarly, the reduction was not significant in the neighboring yards, despite the barrier spray resulting in a mean reduction of 68.9% (P = 0.0921) and the combined treatment resulting in a 72.4% (P = 0.0865) mean reduction.

In treated yards, the greatest reduction due to treatment was seen 17 d posttreatment when a barrier spray was applied, and 25 d posttreatment for LHM and combined treatments. Similar results were seen for neighboring yards, with the exception of when the combined treatment was implemented for which the greatest reduction occurred 9 d posttreatment (Fig. 3). However, the reduction in abundance due to treatment was not significantly lower than the mean posttreatment reduction for any treatment in either the treated or neighboring houses (Supp Table 3 [online only]).

Fig. 3.

Estimated percent reduction in female Ae. albopictus count due to treatment (dot) and 95% confidence interval (bars) on individual days in the treated and neighbor yards. A value of 0 (dashed line) represents no difference from no treatment group. Only a few individual days differ significantly from the no treatment group. There is no overall trend showing a loss of effectiveness over 25 d, and treated yards that received LHM had reductions on day 25 that were significantly larger than the average reduction.

To determine if the Ae. albopictus population rebounded posttreatment, we examined four models for the change in treatment effectiveness over time. Based on AIC (Longnecker and Ott 2015), overall trends were best described using a linear fit for the time since treatment, compared to quadratic fits (ΔAIC = 14.1), cubic fits (ΔAIC = 28), or treating time as a factor (ΔAIC = 62). Within treated yards, estimates for trends in the Ae. albopictus reduction due to all treatments were positive, suggesting that treatments became increasingly effective during the posttreatment period. However, this trend was only significant for LHM (P = 0.0336). Estimates of the trends for reduction due to the barrier spray and LHM in the neighboring yard were negative, but neither was significant (Supp Table 4 [online only]).

Contrasts between the effects 25 d posttreatment and the effects 1 d posttreatment also fail to show any evidence of a rebound in the Ae. albopictus population. Twenty-five days posttreatment, treatment with LHM resulted in Ae. albopictus populations being reduced by 70.6% (P = 0.0123) of its size 1 d posttreatment. The combined treatment and barrier spray also saw an increased reduction 25 d posttreatment compared to after 1 d, but neither was significant. Aedes albopictus populations in neighboring yards were lower 25 d posttreatment than 1 d posttreatment when the focal house was treated with LHM, showing a significant additional reduction 25 d posttreatment (73.2% reduction, P = 0.0361) (Supp Table 5 [online only]). These results, combined with the overall trends provide no evidence that there is a loss in effectiveness of treatments 25 d posttreatment, and that there is significant evidence that Ae. albopictus populations in yards treated with, and neighboring to yards treated with, LHM are still declining 25 d posttreatment.

Discussion

The results of our study concur with previous studies that the yard-scale application of barrier sprays using the pyrethroid, bifenthrin, in conjunction with LHM can successfully suppress Ae. albopictus populations in yards in the temperate United States (Trout et al. 2007, VanDusen et al. 2016, Richards et al. 2017b) and our estimates of the overall effect of bifenthrin barrier sprays are in the range of previous studies. In addition to these estimates, the design of this study allowed for the decoupling of the effects of LHM and barrier spray, and trapping repeatedly at short intervals allowed the estimation of how the effectiveness of treatments changed over the 25 d following treatment. Our study showed that bifenthrin barrier sprays quickly reduced the Ae. albopictus population by 60% and continued having an effect at 25 d, with little change in effectiveness over the 25 d posttreatment, while LHM effectiveness increased over the study period, having no effect immediately following treatment and only began to have an effect after about 21 d. While we did not find any evidence of an increased effect with combined treatments, the difference in timing of the effects suggests that there is an added benefit of using both treatments. In addition to measuring the effectiveness inside of treated yards, we were able to measure the effect of these treatments in untreated neighboring yards. Estimates in the untreated neighboring yards were on the same scale as in treated yards, suggestive of a spillover effect of the treatments, but the differences were not significant. These estimates suggest that it may not be necessary to treat every yard to successfully reduce Ae. albopictus counts across a neighborhood, significantly reducing the risk posed by untreated houses.

While our results show that bifenthrin barrier sprays and LHM have an effect beyond 25 d posttreatment and may reduce Aedes populations beyond the perimeter of the treated yard, future studies need to maintain surveillance more than 25 d posttreatment and should monitor the mosquito populations further than 15 m beyond the treated area. We also failed to see a significant decrease in numbers in the untreated neighboring yards, despite estimates of 50–60% reduction due, in part, to large variability in the results, something future studies should take into account. Previous studies of barrier sprays have also shown that the timing, with respect to the mosquito season, of treatments has an impact on the effectiveness of the control (Williams et al. 2019), with increased effectiveness later in the season, suggesting that the effectiveness of the barrier sprays would be lessened earlier in the season, something that should be accounted for.

Our results show that it is possible to locally reduce Ae. albopictus populations using these treatments for significant periods of time. Applicators could deploy yard-scale treatments, such as barrier sprays and LHM, to target ‘hot-spot’ yards containing large mosquito populations, which has been shown to be effective (Unlu et al. 2016), a strategy that could prove more effective than traditional neighborhood- and city-scale ULV applications. In addition, the estimated effects of the treatment on untreated neighboring yards, while not significant, suggest that an economically optimal strategy for deploying treatments may exist for temporarily reducing the mosquito population in a neighborhood or town in response to mosquito-borne disease outbreaks, e.g., spraying some proportion of yards. However, the efficacy and cost-effectiveness of such a plan would be highly dependent on the costs and accuracy of determining yards with consistently larger Ae. albopictus populations.

Control of Ae. albopictus populations using strategically deployed yard-scale treatments could also help prevent the emergence and spread of insecticide resistance in local populations when it is not necessary to suppress the entire population. Targeting specific yards and other habitats with large mosquito populations could mean that applicators are able to apply a higher concentration than what is possible from vehicle-based ULV applications. Fewer mosquitoes will be exposed to a sublethal dose, while leaving other areas untreated could serve as a natural refuge for susceptible populations. This would serve as a high-dose/refuge approach, similar to what is suggested for resistance management in crop pests (Gould 1998, Tabashnik et al. 2004, Gould et al. 2018), and possibly creating a reservoir for genes beneficial for resistance management (Maino et al. 2019).

Yard-scale control of mosquitoes with a combination of a barrier spray and LHM successfully suppressed mosquito populations in treated and adjacently neighboring yards in our study. However, while we show an estimated reduction of 59.5% in the 25 d posttreatment, it is unclear if this would be sufficient to satisfy private customers. The use of LHM alone, with a longer effective period and an estimated 40.9% reduction over the initial 25 d posttreatment, may prove preferable for long-term reduction of nuisance due to Ae. albopictus. Barrier sprays provided an immediate reduction of the Ae. albopictus, an important addition for applications performed commercially or in response to a Aedes-borne disease outbreak. However, whether the additional reduction provided by barrier sprays would be sufficient for stopping an outbreak of an Aedes-vectored disease, e.g., dengue, would depend on the vectorial capacity of the local population (Fouet and Kamdem 2019). Yard-scale applications of barrier sprays could prove preferable to ULV spraying when used as part of a well-designed integrated vector management program as it allows for targeted ‘hot-spot’ treatments. For instance, targeted barrier sprays could be used to supplement ongoing neighborhood-wide LHM, which has previously been shown to reduce Ae. albopictus populations (Fonseca et al. 2013), in response to a disease outbreak, or as part of ongoing activities (Roiz et al. 2018). While the results presented above show the potential of yard-scale treatments to reduce Aedes populations, much more information is needed to be able to optimally deploy yard-scale targeted controls as part of an integrated management plan. Key among this is the scale and magnitude of the spatial heterogeneity of the Ae. albopictus population, techniques to quickly and efficiently identify ‘hot-spots’, and a framework for determining optimal treatment patterns (Baldacchino et al. 2015, Fouet and Kamdem 2019).

Supplementary Data

Supplementary data are available at Journal of Medical Entomology online.