Man-made stormwater and sewage infrastructure, particularly roadside catch basins, provides widespread habitats for immature mosquitoes in urban and suburban environments. Historically, throughout much of the USA, stormwater, sewage, and industrial wastewater were conducted together through “combined” sewer systems, discharging a combination of stormwater and wastewater into streams. Within recent decades, many cities have replaced these combined sewers with “stormwater only” systems that separate stormwater from wastewater. The objective of this research was to evaluate the implications of this infrastructure conversion for production of Culex pipiens, a primary vector for West Nile virus. On a weekly basis over 14 wk, 20 catch basins (10 combined sewer and 10 stormwater only) were sampled for mosquito larvae and emerging adults using the dipping collection method and floating emergence traps. Abundance of larval Cx. pipiens was higher in combined sewer compared with stormwater-only catch basins, while to the contrary, abundance of adult Cx. pipiens was lower in combined sewer compared with stormwater-only catch basins. This study is the first to reveal that habitat attractiveness and quality for Cx. pipiens may vary between combined sewer and stormwater-only catch basins, and our results contribute to a growing body of research to inform vector management and urban planning efforts as municipalities consider the environmental and public health implications of conversion from combined sewage management to separation of stormwater and wastewater.
West Nile virus (WNV) is the most prevalent mosquito-borne pathogen in the USA, and the abundance of important vector species can be increased greatly by man-made stormwater and sewage infrastructure, a primary habitat for the growth of immature mosquitoes in urban and suburban environments (Geery and Holub 1989). In particular, roadside catch basins (i.e., underground reservoirs that reduce flooding by collecting and conducting surface runoff, debris, and sewage through the subterranean storm drain system) long have provided widespread and protected permanent and semipermanent water sources that are ideal habitats for the development of Culex species (Munstermann and Craig 1977, Crans 2004). These artificial concrete structures typically hold water for extended durations, including during periods of dryness in the summer months (Calhoun et al. 2007), and contain high concentrations of decomposing organic matter, which enhance mosquito production both by attracting gravid females to the habitat via physical and chemosensory oviposition cues (Murrell et al. 2011, Nguyen et al. 2014) and by providing nutritional resources for developing larvae (Murrell et al. 2011). Prior research has demonstrated that there is significant fine-scale spatial and temporal heterogeneity in mosquito production in catch basins, and the presence of larvae within and abundance of adults emerging from these aquatic mesocosms may be influenced by factors including ambient and aquatic temperatures, rainfall (Gardner et al. 2012), water chemistry and pH (Gardner et al. 2013, Arana-Guardia et al. 2014), leaf litter inputs (Gardner et al. 2018), insecticide use (Nasci et al. 2017, Harbison et al. 2020), and catch basin structural properties (grate opening size; O'Meara et al. 1989).
Historically, throughout much of the eastern USA, stormwater, sewage, and industrial wastewater have been conducted together through “combined” sewer systems, discharging a combination of stormwater and wastewater into streams (Tibbetts 2005). Within recent decades, in recognition of the numerous ecological and public health hazards posed by the discharge of untreated domestic sewage into local waterways (Su et al. 2003, Tibbetts 2005), many cities have begun replacing these combined sewers with “stormwater only” systems that separate stormwater from wastewater. Among the environmental impacts of combined sewage overflows, prior research has revealed that Culex quinquefasciatus Say larvae reared in water from a sewage overflow stream may develop faster and grow to larger adult body sizes compared with larvae reared in tap water (Chaves et al. 2009) and that WNV infection in mosquitoes, corvids, and humans are spatially clustered around combined sewage overflow–affected streams in the southern USA (Vazquez-Prokopec et al. 2010). However, no efforts to date have directly compared vector mosquito production in combined versus stormwater-only sewer systems within a narrow geographic radius. Here, we tested the hypotheses that abundance of larval Cx. pipiens L. and production of adult Cx. pipiens in catch basins vary between the 2 sewer systems in an urban setting with mixed combined sewer catch basins and stormwater-only catch basins in close spatial proximity.
This study was conducted from May 30 to August 27, 2018, in Bangor, ME. We selected 20 catch basins (10 per each type of catch basin); all catch basins had circular or rectangular open grates, were constructed of precast reinforced concrete, and were located on the edges of residential streets within a 3-km radius. Water was present in all catch basins throughout the study period, and no larvicides or other mosquito abatement measures were applied throughout the entire city for the duration of the study. Because types of stormwater and sewage infrastructure are spatially aggregated by neighborhood throughout the city, the combined sewer and stormwater-only catch basin types unavoidably also were spatially aggregated within the study area. To understand any potential variables spatially confounding the effect of catch basin type, we compared housing density, median home price (i.e., a proxy for socioeconomic status), and vegetation cover, all of which may be associated with mosquito production (Ruiz et al. 2007, Gardner et al. 2013, LaDeau et al. 2013), and found no significant differences across the treatment areas.
On a weekly basis, all 20 catch basins were sampled for mosquito larvae and emerging adults under conditions of no rainfall. To test the hypothesis that abundance of larval mosquitoes in catch basins varies between combined sewer catch basins and stormwater-only catch basins, larvae were collected using a standard dipping method. In brief, a 10.3 × 10.3-cm aquarium net attached to the end of conduit pole 3 m in length was passed over the water surface in a single figure eight, and the net was inverted into a container and flushed with water. Larvae were stored on ice for further processing; all instars were counted in aggregate and identified to species using taxonomic keys (Andreadis et al. 2005). To test the hypothesis that production of adult mosquitoes in catch basins varies between combined sewer catch basins and stormwater-only catch basins, floating emergence traps were deployed at the beginning of the study period and continuously monitored for adult mosquitoes (Hamer et al. 2011). These traps float on the surface of the water and capture adult mosquitoes in a conical collection cup as they fly upward out of the catch basins. Once per week, adult mosquitoes were collected from the traps, counted, and identified to species (Andreadis et al. 2005). The effects of catch basin type on abundance of larvae and emerging adult Cx. pipiens were analyzed using separate Poisson regression models. These models tested the fixed main effects of catch basin type (i.e., combined sewer or stormwater only), week, and their interaction and the random effect of catch basin ID (i.e., a unique identifier associated with each catch basin). Due to the well-known correlations between temperature and rainfall and mosquito production in catch basins (Rueda et al. 1990, Gardner et al. 2012), mean ambient temperature (°C) and cumulative rainfall (cm) on collection dates were included as fixed covariates. Weather data were obtained from the Fairmount weather station (44.80°N, 68.80°W) in Bangor.
A total of 3,809 larval and 1,510 adult Cx. pipiens were collected throughout the study period; no natural predators of mosquitoes (e.g., copepods, dragonfly naiads) were observed. Other species that were collected but not included in statistical analyses due to small sample sizes were 813 larval and 30 adult Cx. restuans Theobald, 570 larval and 15 adult Culex sp. unidentifiable to species due to poor specimen quality, and 32 larval and 16 adult Aedes japonicus Theobald. Larval and adult mosquito abundance were extremely low during the 1st 7 wk and increased during the latter half of the study period. Abundance of larval Cx. pipiens was higher in combined sewer compared with stormwater-only catch basins (Z = −1.89, P = 0.05; Fig. 1a) and was positively correlated with temperature (Z = 25.99, P < 0.01) and negatively correlated with rainfall (Z = −5.17, P < 0.01; Table 1). To the contrary, the abundance of adult Cx. pipiens was lower in combined sewer compared with stormwater-only catch basins (Z = 2.57, P = 0.01; Fig. 1b), but adult abundance also was positively correlated with temperature (Z = 7.06, P < 0.01) and negatively correlated with rainfall (Z = −16.21, P < 0.01; Table 1).
Our study was the first to reveal that habitat attractiveness and quality for Cx. pipiens may vary between combined sewer and stormwater-only catch basins within a narrow geographic radius, with consequences for mosquito production. Interestingly, while the abundance of larvae was significantly higher in combined sewer compared with stormwater-only catch basins, the opposite held true for adult Cx. pipiens emerging from the catch basins. The finding that larval densities are high in combined sewer catch basins is consistent with prior research that has demonstrated the attractiveness for oviposition of combined sewage overflows to gravid female Cx. quinquefasciatus (Chaves et al. 2009). Further study is necessary to address whether the cues that enhance oviposition are related to high microbial activity (Ponnusamy et al. 2008, 2010) and nutrient load, particularly nitrogen (Gardner et al. 2013, Nguyen et al. 2014), which are characteristic of these environments, or to the presence of conspecifics or absence of other invertebrates, including competitors and predators (Kiflawi et al. 2003, Reiskind and Wilson 2004). Although our observation that adult emergence was lower in combined sewer relative to stormwater-only catch basins was inconsistent with previous research that suggests that nutrient-rich combined sewage overflows may enhance mosquito development and promote WNV transmission (Calhoun et al. 2007, Vazquez-Prokopec et al. 2010, Chaves et al. 2011, Nguyen et al. 2014), these prior studies focused on mosquito production in streams and seminatural experimental settings. In the comparatively closed mesocosms created by stormwater and sewage infrastructure, eutrophication (Mogi and Okazawa 1990, Noori et al. 2015) and intraspecific competition (Reiskind and Wilson 2008) may inhibit immature mosquito growth, even in generally pollution-tolerant species like Cx. pipiens (Savage and Miller 1995). Future research should clarify the specific biological mechanisms by which catch basin type alters larval mosquito development and adult emergence success. Overall, our results contribute to a growing body of research that may inform vector management and urban planning efforts as municipalities consider the environmental and public health implications of conversion from combined sewage management to separation of stormwater and wastewater.
The authors thank Cassie Steele, Troy Cloutier, and Audrie French for their technical assistance, the Gardner Lab at the University of Maine for their feedback on the development and execution of the study, and the Public Works Department in the City of Bangor, ME, for permission to conduct the research. Support for this research was provided by the United States Department of Agriculture (USDA) Northeastern IPM Center, Project Number 73984-10929 and the USDA National Institute of Food and Agriculture, Hatch Project Number ME021826 through the Maine Agricultural and Forest Experiment Station.