In the northeastern part of the greater Chicago metropolitan area, the North Shore Mosquito Abatement District (NSMAD) treats approximately 50,000 catch basins each season with larvicide tablets as part of its effort to reduce local populations of the West Nile virus (WNV) vector Culex pipiens. During the 2014 season, an NSMAD technician monitored a subset of 60–195 basins weekly for 18 weeks among the communities of the District for the presence of mosquitoes. Monitoring found no clear evidence in the reduction of mosquitoes with the use of larvicides, and visual inspections of 211 larvicide-treated basins found that the majority (162, 76.8%) were missing tablets 1–17 weeks after applications. This loss of treatment may be due to the rapid dissolution or flushing of larvicides and would help explain why the larvicide appeared to be ineffective.
Introduction
As part of efforts to decrease local vector populations of West Nile virus (WNV) and other mosquito-borne diseases, stormwater catch basins are often targeted for routine larvicide applications in urban areas around the world.1–23456789101112131415–16 The North Shore Mosquito Abatement District (NSMAD) treats approximately 50,000 catch basins each season in the District's 207 sq km (80 sq mi) located in the northeastern greater Chicago IL metropolitan area. Treatment is with extended-release, 180-day larvicides, and the specific objective is to reduce local populations of the local WNV vectors Culex pipiens and Cx. restuans. NSMAD technicians begin treatment each year in May by applying a single dose of an extended-release larvicide to the catch basins. Once all basins have been treated, typically taking 10–12 weeks, an additional round of larvicide application begins in basins treated earliest in the first round as supplies and availability of seasonal staff permit. For the 2014 season, Natular™ XRT extend-release tablets (XRT: Clarke Mosquito Control Products, Inc.) were applied to basins beginning in the last week of May and continued until the second week of August. This product was chosen based on smaller scale trials performed in 2011,17 2012,18 and 201319 by NSMAD, which suggested a range of effective control from 8 to at least 14 consecutive weeks. Basins in the southernmost 13 sq km (5 sq mi) of the district that had received the first applications in May were treated with an additional Natular™ T30 30-day tablet20 (T30: Clarke Mosquito Control Products, Inc.) 1–2 weeks after the last XRT application (Fig. 1). The use of these two pesticides' active ingredient, spinosad, for mosquito control, including the effect on nontarget species, has been described at length elsewhere.21–2223242526–27 The additional T30 applications were ended by September when seasonal technicians were no longer available.
Methods
As part of regular catch basin monitoring performed by NSMAD, 5–45 catch basins were inspected for the presence of mosquito larvae and pupae weekly from selected basins in 12 of the 13 communities within the NSMAD operational area during 18 weeks in June to September. The 13th, a small community with a total area of only 1.2 km sq (0.45 square mi) and only 60 catch basins, was not included in inspections. Monitoring was performed by removing the circular grate of each structure with a manhole hook and taking two dip samples using a standard 350-mL dipper. The average number of mosquitoes per two dips in treated basins and those basins that had yet to be treated (“untreated”) was then used to inform NSMAD staff on larvicide effectiveness. In those cases where catch basin sump water was clear enough to see the entire bottom of the structure, the presence of XRT and/or T30 tablets was noted. When possible, a search of the bottom was made using the dipper to ensure that larvicide tablets were not hidden by sediment. Two-sample t-tests were used to compare average dip samples between treatments (untreated vs XRT and XRT vs XRT + T30) each week. Precipitation data were downloaded from a nearby weather station of the National Oceanic and Atmospheric Administration (NOAA) National Weather Service Forecast Office located at the Chicago O'Hare Airport ( http://www.nws.noaa.gov/climate/index.php?wfo=lot). These data were used to compare the amount of rainfall in the 2014 season with rainfall in earlier years.
Results
Of the 1,521 total basins monitored, most (1,095) did not hold mosquitoes at the time they were inspected. The structures that were found to harbor larvae and/or pupae at the time of inspection had from 1 to approximately 200 mosquitoes in a dip. During the 9 weeks in which untreated and XRT-treated basins were both monitored (weeks 24–32), average dip samples were observed to be significantly different between these two treatment types in only 2 weeks. In week 30, treated basins had fewer mosquitoes than untreated basins. In week 32, treated basins had more mosquitoes than untreated basins (Table 1, Fig. 2). During the final 6 weeks of the study, when comparisons between XRT-treated and XRT + T30-treated basins could be made, no significant difference in mosquito numbers was observed between these types (Table 1, Fig.2). Two-hundred and eleven treated basins had visible bottoms, of which 162 (76.8%) appeared to be missing tablets including 10 XRT + T30-treated basins that were observed with either an XRT or a T30 tablet only (9 of these missing an XRT tablet and 1 missing a T30 tablet) and 16 XRT + T30-treated basins missing both tablets. Tablets were observed missing in basins in all 12 communities from 1 to 17 weeks after applications. The total rainfall from June to September 2014 was the highest of the past five seasons (Table 2).
Table 1.
Comparison of average mosquitoes per dip taken weekly from 1,521 untreated, Natular™ XRT- treated, and XRT + Natular™ T30 -treated catch basins in the greater Chicago metropolitan area in 2014.
Table 2.
Total rainfall in cm (in) from June to September 2010 to 2014. The 30 year average for this time period is 38.76 cm.
Discussion
Clearly, the results suggest that the effectiveness of both larvicides was less than ideal during the 2014 season. The XRT-treated basins harbored similar numbers of mosquitoes as those without larvicides, and no difference was observed between XRT + T30-treated basins when compared to those without the additional T30 application. Factors that can potentially contribute to a decreased period of efficacy are stated in these larvicides' product labels and include high rainfall or strong water flows that can flush larvicides out of structures and/or increase their rate of dissolution. Given the large number of treated basins that were observed without tablets, flushing and rapid dissolution likely occurred widely, and some direct evidence of both was observed (Figs. 3 and 4). That a large percentage of observable basins were missing treatments is clearly the most significant finding from the 2014 monitoring. The results of comparisons among untreated, XRT-treated, and XRT + T30-treated basins should be interpreted with caution, as basins associated with these treatments were spread out among the 207 sq km operational area and were likely influenced differentially by a number of site-specific factors (eg, amount of runoff and debris).
Unfortunately, by nature of their design and function to capture runoff, the influx of potentially disruptive runoff water flows is common, if not expected, in catch basins. As the XRT tablets are similar in size and weight to other commonly used extended-release larvicides (ie, Zoecon Altosid® XR Extended Residual Briquets (Wellmark International) and FourStar™ Briquets (FourStar Microbial Products LLC), a comparable degree of loss could be expected with other products. In general, the number of mosquitoes observed in NSMAD basins in 2014 appeared smaller than in some previous years,17–18–19 and it is possible that runoff from this season's higher rates of rainfall could have flushed both larvicides and mosquitoes out of structures.2,3,28–29–30 Currently, there are no plans by the NSMAD to monitor the dispersal of lost pesticide active ingredient from the catch basins.
Although the potential for flushing and/or rapid dissolution of catch basin pesticides is noted on the pesticide labels, the degree to which this occurs is not well documented and likely varies widely by locality, structure, and type of pesticide formulation (granular, tablets, vapor, etc). Certainly, that flushing and rapid dissolution occur is not surprising, particularly given that rainfall during monitoring was so high, but the degree to which these phenomena were indirectly observed was unexpected. As it is logistically difficult for mosquito control programs to monitor both mosquitoes and associated treatments within even a small percentage of their catch basins (ie, removal of heavy manhole grates, proximity to vehicular traffic), there is paucity of studies on this subject. Depending on the size and resources of local programs, monitoring of mosquitoes within a certain portion of catch basin may not even occur at all. Finally, it is not routine for mosquito control personnel to publish studies in more formal and academic publications, and thus such work cannot be found by common research search engines such as PubMed and Google Scholar.
Most monitored basins in this study did not hold mosquitoes at the time of inspections; however, long term monitoring by the NSMAD has found that local mosquito populations fluctuate greatly and during some seasons a much greater percentage of these widespread and abundant structures will harbor mosquitoes at various points in times. Due to their great prevalence and propensity to harbor Cx. pipiens in urban areas, the use of catch basin larvicides is an important consideration for mosquito control efforts, but ensuring that these pesticides remain in structures for their designed duration is a significant challenge. As an alternative to pesticide applications, the use of manhole inserts to deter mosquitoes from entering basins has seen some success in small-scale experimental trials,31–32–33 but this have yet to be implemented on a wider scale. Such an intervention would likely require collaborations with local stormwater agencies and a prohibitively large financial investment in stormwater infrastructure.
Author Contributions
Conceived, designed, and performed the experiments: JEH, MH. Analyzed the data: JEH. Wrote the first draft of the manuscript: JEH, DZ, CX, JEL, MH, MOR. All authors reviewed and approved of the final manuscript.