Mosquitoes pose health risks to human populations by serving as vectors of diseases. Mosquito control organizations are responsible for inspecting and controlling vector populations to reduce the risk of infection of these diseases. Current sampling methods are effective for numerous types of mosquito habitat, but not conducive for sampling small overhead habitat such as roof gutters or tree holes. We have developed and tested a tool called the Mosquito GutterSnipe to sample these overhead habitats. Volumetric and larval capacity testing of the tool prototype demonstrated comparable sampling integrity to standard mosquito dipping methods. The GutterSnipe can be employed as a reliable way to sample previously overlooked mosquito habitat. Its current model is cost effective and easy to produce for mosquito control organizations and easy to use for inspectors.
Mosquitoes employ a diverse selection of breeding habitats, both naturally occurring and man-made. Habitat types ranging from cattail wetlands and tree holes to standing water in catch basins and waste containers all offer ideal reproductive conditions for numerous species of mosquitoes (Fincke et al. 1997, Muturi et al. 2007). While all habitats are researched and controlled in consideration of mosquito-borne disease prevention, focal habitat type will vary by area based on other surrounding environmental factors. These factors help to direct funding, resources, and time allocated toward higher risk habitat types for vector population control (Nasci et al. 2001, Evans et al. 2018, Ng et al. 2019). This does not mean that smaller or less researched mosquito habitats cannot produce disease-bearing mosquito vectors or overwhelming populations of mosquitoes (Yanoviak 2001a, 2001b).
Some man-made structures are successful mimics to some of the less investigated mosquito habitats, and many require further study for proper health risk assessments. One such habitat includes roof gutters on buildings or shelters (Sota et al.1994, Saleza et al. 2011). Acknowledged but often overlooked, roof gutters can become dammed with leaves, sticks, and other debris or runoff from roofs and lead to failed drainage in wet conditions (Troyo et al. 2008, Parker and Allan 2019). The standing water can serve as habitat if roof gutters remain unchecked, allowing various species of mosquitoes to use it for reproduction, such as Aedes aegypti (L.), Ae. albopictus (Skuse), Limatus durhamii Theobald, Ae. notoscriptus (Skuse), Ae. sierrensis (Ludlow), and Culex spp. (Goddard et al. 2002, Montgomery and Ritchie 2002, Calderón-Arguedas et al. 2009). As a result of increased habitat and increased reproductive potential, risk for the development of mosquito-borne diseases also increases (Goddard et al. 2002, Turell et al. 2005, Hemingway et al. 2006).
The importance of surveying gutter habitats is understood, but appropriate sampling methods for this unique location and form of the habitat are often challenging to perform. In standard mosquito inspection practice, specialized long-handled 400-ml ladles (dippers) and pipets are the most common sampling tools (Andis et al.1983 Claborn et al. 2018, Dinh and Noval 2018, Ostrum and Mutebi 2019). Observational inspections are another inspection method (Montgomery and Ritchie 2002, Calderón-Arguedas et al. 2009, Saleeza et al. 2011, Haas-Stapleton et al. 2019). Combinations of the listed techniques have been employed when studying mosquito gutter habitat, but no uniform methodology that is safe, effective, and time efficient has been defined at present (Slaff et al. 1983, Rajarethinam et al. 2020). To address this, we developed and tested a tool prototype designed to allow mosquito inspectors to access roof gutters and other overhead mosquito habitats that would typically be out of an inspector's reach: Mosquito GutterSnipe.
The GutterSnipe is a dowel-mounted siphon operated by a simple lever that uses a modified baster as the means of sample collection. It comprises mainly poplar dowels, metal fasteners, bolts, and screw eyes, and colored mason line. The tool's body measures 31.75 cm (12.5 in.) in length, 22.86 cm (9 in.) in height, and 10.16 cm (4 in.) in width in a non-space-filling volume with the plunger depressed. When a baster is inserted into the frontal screw eye, it can extend down to 19.05 cm (7.5 in.) below the bottom surface of the body. When assembled, the GutterSnipe is 0.63 kg (1.39 lb). The handle may vary in length, but all handles must have circular ends with a 1.91 cm (0.75 in.) diam. Strong adhesives are required to modify the baster by depressing the bulb in on itself and cementing the crease. Adhesives may also be used to help cement connections or fasteners and prevent loosening of the plunger (Fig. 1; see also https://youtu.be/OQMQxjgXMpg).
The GutterSnipe prototype was designed to be cost effective, easy to modify, and easy to repair and maintain. Materials for building the prototype can be acquired at hardware stores. In many cases, more than enough materials are included with the packaged source materials, further increasing the cost-effective production of the GutterSnipe (Table 1).
The GutterSnipe was tested in both laboratory and field settings. Volumetric tests were performed in the field to compare sampling methods and determine uptake volume, whereas larvae sample tests were performed in a laboratory setting. Volume potential was tested by taking samples from wetland habitat using only the baster part of the tool in hand and immediately measuring the contents, using the GutterSnipe and immediately measuring the contents, and using the GutterSnipe and waiting 5 sec before measuring the sample contents. Respectively, these categories of measurements served as volumetric control, maximum potential, and simulated-use results. Additionally, volumetric tests were conducted at 3 separate field sites in Fridley, Minnesota, and were categorized by qualitative environmental factors, particularly visual water clarity. Sites 1, 2, and 3 were selected based on respective decreasing water clarity.
Larval sampling tests were conducted by drawing samples from premeasured Culex pipiens L.–burdened containers and counting the numbers of larvae drawn up by the GutterSnipe. Three containers of larvae-infested water were tested: 325 ml with a depth of 8.16 cm, 1.81 liter with a depth of 1.9 cm, and 1.93 liter with a depth of 12.5 cm; respectively; these containers represented sampling control, wet habitat, and flooded habitat. All larval test conditions were conducted with a stock larval count of 488 mosquito larvae. Findings from volumetric and larval testing were analyzed statistically using analyses of variance (ANOVA) conducted in the program R (R Core Team 2020).
A total of 270 dips were measured when assessing the volumetric potential of the GutterSnipe. These were subdivided by site, sample type, or both (90 dips, 90 dips, or 30 dips respectively). Using ANOVA tests, mean dip volumes were compared by site (F = 1.105, P = 0.333) and showed no significance between any of the 3 sites sampled. Volumes were also compared by sample type (F = 49.75, P < 0.001) and showed significance between the tested sampling methods. Further analysis with a Tukey HSD post hoc test showed that the control, no wait, and 5-sec wait sampling methods were all significantly different from one another (P < 0.001).
A total of 300 dips were taken when assessing the larval sampling potential of the GutterSnipe. These were categorized by the container from which the samples were siphoned (control, wet, and flooded habitat simulations). Average larval sample sizes were compared by simulated habitat type (F = 20.89, P < 0.001). Further analysis with a TukeyHSD post hoc test showed that samples collected from the flooded habitat showed statistically significant differences from both the control (P < 0.001) and wet habitats (P < 0.001). However, a comparison of the control to the wet habitat did not demonstrate a statistically significant difference (P = 0.57).
Both the volumetric potential and larval sampling capability results were compared to determine an overall sampling rate estimate. Based on the average 5-sec-wait volume (simulating field use of the GutterSnipe; 28.51 ml), the overall larval sampling average (9.01 siphoned larvae), and common habitat sampling procedure, the GutterSnipe has an average calculated sampling rate of 3.16 mosquitoes/dip.
The GutterSnipe allows mosquito inspectors to collect samples easily and inspect them with greater scrutiny than other available sampling methods. It provides an easier and safer alternative to climbing up to the overhead mosquito habitat or attempting to bring it down and collect samples. Additionally, it is a less invasive manner of data collection, especially on or near privately owned properties (Focks 2003).
The tool exhibited an average maximum potential volume of 41.20 ml of water drawn up by suction, but this maximum volume is not commonly achieved with normal use of the tool. Mosquito dippers are expected to collect a maximum volume of approximately 350 ml of water but will realistically gather less volume in field samples (Andis et al. 1983, Dinh and Novak 2018, Harbison et al. 2018). This indicates that, like mosquito dippers, the GutterSnipe has high volumetric potential but in realistic use will not reach this value due to intricacies of sampling techniques. It is important to consider for the effect on the overall sampling potential of the GutterSnipe.
Standard mosquito dippers perform best in moderately deep water such that the cup can be completely submerged in the body with the base parallel to the water's surface. In shallower bodies of water, rotation of the cup and use of additional pressure to retrieve water samples is often required and can impact the sampling rate in some cases. Additionally, the cup's large size and shape can make it difficult to acquire samples from small sections of habitat or areas with small entry points (Focks 2003). The Mosquito GutterSnipe is ideal for reaching confined mosquito habitat with its narrow baster and suction action. In a laboratory setting, larval samples from simulated flooded habitat were significantly different from both the control and wet habitat simulations. However, there was no statistical difference between the control and wet habitats, indicating the GutterSnipe performs more effectively in small shallow habitat types. In testing, the GutterSnipe retrieved the least number of larvae from the flooded habitat (4.54 mosquitoes/dip), which was significantly less than the averages for the control and wet environments (9.42 mosquitoes/dip and 8.41 mosquitoes/dip, respectively). The GutterSnipe is less effective in conditions where a standard dipper would be used. Instead, it is better for sampling confined habitat types holding smaller volumes of water.
Larval sampling with a dipper seeks to obtain a sampling rate greater than or equal to an established risk threshold (Focks 2003). Mosquito dippers are expected to find a minimum average of 2 mosquito larvae per dip in appropriate habitat. The estimated sampling rate for field use of the GutterSnipe was able to meet this threshold expectation in a laboratory test setting calculated at 3.16 mosquitoes retrieved per dip. In a field application, it is likely this value will be closer to the established 2 mosquitoes/dip rate considering environmental factors such as water cleanliness and volume.
It should be noted that the Mosquito GutterSnipe is amenable to further modifications and improvements. Potential modifications include telescopic poles in place of the static dowel handle, a grip at the end of the pull string, simple pulleys to help guide the pull string, use of static or telescoping dowels in place of the pull string so the plunger can both be set and used while the tool is overhead, and a hinge system on the baster attachment so the tool can increase its range of motion for particularly hard-to-reach standing water. Despite any future changes, the purpose of the GutterSnipe remains to expand on current mosquito inspection methods by sampling a unique overlooked habitat type in a safe and time-effective manner.
We thank D. Dirkswager and the Twin Cities Metropolitan Mosquito Control District for assistance and resources for working on the GutterSnipe prototype. We thank A. Fallon for providing mosquito larvae for testing and J. Lundquist and P. J. Leonard for assistance with prototype testing and construction.