Impact of underground storm drain systems on larval ecology of Culex and Aedes species in urban environments of Southern California
Ethics and vertebrate animalsThe field surveys and collections were conducted on accessible public areas or private residential areas with property owners’ permission. The study did not involve human participants, or endangered or protected species. Laboratory mice were used as a blood source for mosquitoes. All experimental protocols were approved by the Institutional Animal Care and Use Committee (IACUC) of the University of California, Irvine (UCI) (IACUC protocol number: AUP-19-165). All methods were carried out in accordance with relevant IACUC guidelines and regulations.Study sites and mosquito larval habitat surveillanceThe study was carried out in Orange County, California, USA. Orange County is a highly urbanized county with an estimated population density of approximately 1470 people/km2 according to U.S. Census Bureau, an average annual low/high temperature range of 13–25 °C, 65% relative humidity, and annual precipitation of about 350 mm according to U.S. Climate Data. Annual rainfall was 261 mm, 311 mm, 198 mm and 475 mm for 2016, 2017, 2018 and 2019, respectively. A major drought event occurred in December 2017 and February 2018 when the total rainfall in the 3-month period was 20.6% of the 30-year average. Both Ae. aegypti and Ae. albopictus were discovered in the county in 20158. Culex quinquefasciatus is the most abundant mosquito in the county and breeds readily in a variety of residential, commercial and USDS water sources, and is the primary vector of West Nile virus in southern California18.Larval mosquito surveillance in Orange County was conducted from 2016 to 2019 by the Orange County Mosquito and Vector Control District (OCMVCD) through its routine mosquito surveillance and treatment program, following the recommendations of the California Department of Public Health and the Mosquito and Vector Control Association of California19. Briefly, OCMVCD staff conducted routine inspection for aquatic habitats in randomly selected public areas, and performed door-to-door mosquito larval and adult sampling on residential or commercial premises upon the request of the residents or business owners while distributing public education materials for vector control and personal protection. Arial photography was used to examine the presence of abandoned swimming pools in residential areas. In addition to surface aquatic habitats, subsurface habitats (e.g., catch basins, underground drains, manhole chambers, and public utility vaults) were examined for larval abundance of all mosquito species. In 2019, OCMVCD completed 5,622 mosquito service requests, and conducted 11,813 inspection and treatments on routine sites using a variety of public health-approved adulticides and larvicides. A total of 38,099 underground drains and catch basins and 6925 km of flood channels were treated. In addition, a total of 17,783 km of gutters and 3562 neglected swimming pools were inspected and treated. The larval distribution data reported here were based on this extensive field sampling effort20.Larval sampling used standard mosquito dippers or pipettes, and specialized modifications of these to sample hard to reach areas. Mosquito larvae from each source were collected, transferred into a uniquely-numbered vial with isopropyl alcohol (70%), and submitted to the laboratory for identification; if present, live pupae were collected and held in site-specific labelled rearing chambers (BioQuip Products, Inc., Rancho Dominguez, CA) until emergence. Third and fourth instar mosquito larvae (1–100, depending on sample size) and emerged adults were identified to species using a stereo microscope (40–50x) and morphological features described in taxonomic keys21,22. Results were uploaded to OCMVCD’s data management system, along with collection date, GPS location, and habitat type for each sample site. For this study, larval habitats were classified into six types: small container, underground system, ornamental water features, marsh, pools/spas, and creek (Table S1). The container classification included flowerpots/vases, saucers, tires, bowls, boxes, buckets, dishes, tree holes, etc. Underground storm drain system referred to larval habitats such as catch basins, manhole chambers, underground drains, and public utility vaults that were below the ground. Water feature included flood control channels, ponds, fountains, birdbaths, street gutters and small reservoirs, etc. Marsh included both fresh and salt water marshes.Mosquito strains and water source for laboratory studiesWe examined the effect of USDS water on oviposition substrate preference and larval development in microcosms in an insectary with climate control (27 ± 1 °C, 70 ± 10% relative humidity, and 12 h light/12 h dark photoperiod) at UCI. To minimize potential bias on behavior and ecology from mosquito colonization, this study did not use previously established laboratory mosquito colonies. Instead, we used Ae. aegypti and Ae. albopictus adults reared from field-collected eggs using ovicups in residential areas of Orange and Los Angeles Counties, California, respectively. Culex quinquefasciatus were also reared from eggs of field-collected, blood-engorged adult mosquitoes using gravid traps in Orange County23.All experiments reported here used two types of habitat water: (1) USDS water collected from seven manhole chambers or catch basins (33°47′01.9″N, 117°53′19.0″W, Orange City, manhole; 33°52′25.0″N, 117°57′02.6″W, Fullerton City, manhole; 33°44′44.4″N, 118°06′24.2″W, Seal Beach City, manhole; 33°55′38.9″N, 117°56′51.4″W, La Habra City, manhole; 33°52′48.9″N, 117°55′21.4″W, Fullerton City, catch basin; 33°54′35.2″N, 117°56′02.5″W, Fullerton City, catch basin; 33°52′25.0″N, 117°57′02.6″W, Fullerton City, catch basin); and 2) flowerpot water from vases of three cemeteries in Orange County (33°50′29.0″N, 117°53′57.9″W; 33°46′21.5″N, 117°50′35.8″W; 33°46′12.3″N, 117°50′21.4″W). Water (including sediments) from each breeding source was collected with mosquito dippers and mixed together by habitat type into 18.9 L (five-gallon) Nalgene™ containers. The containers were transported to the laboratory in shaded ice containers, and stored overnight in a refrigerator at 4 °C. The experiments described below were conducted on the field-collected water for the two habitat types. We selected flowerpot water as the comparison substrate because flowerpot containers showed the highest larval positivity rate in the study area.Oviposition preference testTo examine whether USDS water attracts or repels egg laying by Ae. aegypti and Ae. albopictus mosquitoes, a two-choice oviposition preference test was conducted. Briefly, this experiment used two ovicups placed within a mosquito cage (1 × 0.5 × 0.5 m3), one ovicup with 200 ml USDS water and another with 200 ml flowerpot water. Adult mosquitoes were bloodfed on mice; fully engorged females 3-days post-bloodfeeding were used for oviposition preference tests. Ten gravid Ae. aegypti females were released into a cage and allowed to lay eggs for three days, and the number of eggs in each ovicup were counted. Five replicates were used. The same experiment was conducted for Ae. albopictus.To evaluate whether the presence of Cx. quinquefasciatus larvae has any impact on the egg laying behavior of invasive Aedes mosquitoes, the two-choice oviposition preference test described above was used. One ovicup contained 200 ml USDS water and ten first-instar Cx. quinquefasciatus larvae, while the second ovicup contained 200 ml USDS water only. Ten gravid Ae. aegypti or Ae. albopictus females were released into a cage and allowed to lay eggs for three days. Five replicates were used. We also conducted this experiment using flowerpot water with the same design and same number of replicates to determine whether the impact of Cx. quinquefasciatus larvae on Aedes mosquito egg laying behavior was similar across different water substrate types.Egg hatchingTo investigate the effects of different habitat water sources on egg hatching, 50 Ae. aegypti or Ae. albopictus eggs on separate filter papers were introduced into ovicups with 200 ml USDS water or flowerpot water. Deoxygenized distilled water that we routinely use in laboratory mosquito colony maintenance was used as a positive control. The experiment was conducted in an insectary with climate control (27 ± 1 °C). The number of larvae hatched were counted daily for six days continuously. Five replicates were used.Larval survivorshipA life table study was conducted on Ae. aegypti and Ae. albopictus larvae to determine the effect of USDS water and flowerpot water on larval development and survivorship. Twenty-five newly hatched Ae. aegypti or Ae. albopictus larvae were introduced into a microcosm that contained 200 ml USDS or field flowerpot water. The number of dead and surviving larvae was recorded daily until they pupated. Pupae were counted, and removed to different paper cups for emergence to adults. Four replicates were used for each type of habitat water per species. We included Cx. quinquefasciatus in the larval life table study for method validation purposes because the larvae of this species were known to successfully develop into pupae and adults in USDS water in southern California10.Larval survivorship experiments were conducted in two different seasons. The first was in the summer (August–September) 2019 when the density of invasive Aedes species peaked19, and also insecticide runoff from mosquito and residential/agricultural pest control applications were at the highest levels in southern California24. The second was in the winter (December) 2019 when there was little insecticide treatment for mosquito and pest control. This design enabled us to examine seasonality in larval survivorship and the impact of environmental insecticide runoff in USDS water. To determine whether USDS water’s nutritional deficiency plays a major role in limiting Aedes larval development, we repeated the larval survival experiment by adding 0.1 g Tetramin Tropical Flakes, the standard larval mosquito diet in insectaries, to the microcosms every 2 days. The number of dead and surviving larvae, pupae, and emergent adults was recorded daily.Data analysisAll aquatic habitats that were positive or negative for the larvae of Ae. aegypti, Ae. albopictus and Cx. quinquefasciatus (the predominant species), were mapped using ArcGIS 10.7.1. The proportion of aquatic habitats positive for Ae. aegypti and Cx. quinquefasciatus was calculated for each habitat type from 2016 to 2019. To examine variation in Aedes and Culex larval positivity rate among different groups of larval habitats within the USDS, larval positivity rates for Ae. aegypti and Cx. quinquefasciatus were calculated for underground water retention vaults, underground catch basins/manholes, and underground pipelines/tunnels. The Chi-square test was used to examine the statistical significance. Culex quinquefasciatus was analyzed because it was the most common species, whereas Ae. albopictus was not included in the analysis due to insufficient number of Ae. albopictus positive habitats. To determine whether USDS water attracted or repelled oviposition of invasive Aedes mosquitoes, a pairwise t test was used to compare egg number in USDS water ovicups to flowerpot water ovicups for each Aedes species. Similarly, a pairwise t-test was used to test the effect of Cx. quinquefasciatus larvae on Aedes mosquito oviposition choice.To examine the effect of water sources on egg hatching, the t-test was used to analyze the egg hatching rate. The analysis of larval life table study data focused on pupation rates and larval-to-pupal development times. The pupation rate was calculated as the proportion of first-instar larvae that molted into pupae. The effect of water sources and larval food supplementation on pupation rate was analyzed using non-parametric Wilcoxon test. The t-test was used to analyze the duration of larval-to-pupal development. Kaplan–Meier survival analysis was used to determine the effects of food supplementation and water source on larval development for each species, and the log-rank test was conducted to determine their statistical significance. All statistical analyses were performed using JMP software (JMP 14.2, SAS Institute Inc.). More
