The Impact of Hurricane Irma on Population Density of the Black Salt-Marsh Mosquito, Aedes taeniorhynchus, in Collier County, Florida

The black salt-marsh mosquito, Aedes taeniorhynchus (Wiedemann), is a common mosquito species found in the eastern coastal areas of the Americas. While having the ability to transmit several viruses and filarial worms to humans and other animals, Ae. taeniorhynchus is primarily considered a nuisance in Florida (Agramonte and Connelly 2014). Gravid Ae. taeniorhynchus females lay eggs in moist soil of the salt-marsh and mangrove swamps just above the waterline where eggs are subjected to flooding by rainfall or high tide (Ritchie and Addison 1992). Aedes taeniorhynchus are migratory insects (Nielsen 1958), and often leave their emergence sites during large population explosions (Kennedy 1961). The majority of spring adulticiding applications performed by Collier Mosquito Control District (CMCD), located in Southwest Florida, are due to Ae. taeniorhynchus. Adult females not only emerge from local mangrove swamps in Collier County, FL, but also enter the district by migrating on the wind as far as 20–60 mi (32–96 km) (Harden and Chubb 1960) from breeding sites within the Everglades National Park, Ten Thousand Islands National Wildlife Refuge, and Rookery Bay National Estuarine Research Reserve.

The southeastern USA and Caribbean endured an extremely active year for tropical disturbances during the 2017 hurricane season (June 1–November 30, 2017). On September 10, 2017, Hurricane Irma made a 2nd landfall in Florida, striking Collier County. Areas around the county affected by Hurricane Irma are abundant in swamps dominated by red (Rhizophora mangle L.) and black (Avicennia nitida Jacq.) mangroves, prime habitat for Ae. taeniorhynchus mosquitoes (Provost 1973). To our knowledge, the salt-marsh season (May–June) of 2018 represents one of the lowest Ae. taeniorhynchus activities recorded in Collier County. In 1983, the county also experienced extremely low Ae. taeniorhynchus numbers linked to a strong El Niño/Southern Oscillation (ENSO) event (Ritchie 1984), representing the impact of severe weather disturbances on population density.

This study consisted of entomological data collected in Collier County from 2016 to 2018 during the peak months of Ae. taeniorhynchus activity. Mosquito population data consisted of May and June human landing rate counts and real-time trap data derived from the BG-Counter (Biogents AG, Regensburg, Germany). Landing rates were determined by counting the number of adult female mosquitoes landing on an individual within 2 min. Landing rate study sites selected were based off of previously known areas with populations of Ae. taeniorhynchus. Data collected included 17 sites from Marco Island (3.5-mi [5.6 km] radius from the point: 25°56′57.50″N, 81°41′24.12″W) and 11 sites from Henderson Creek (4-mi [6.4 km] radius from the point: 26°3′14.61″N, 81°39′42.41″W). Landing rate average for each year is expressed as the mean count for the area during May–June (mean ± standard deviation). During 2017 and 2018, 2 sites were monitored using the BG-Counter remote surveillance technology. BG-Counters were placed in Marco Island (25°56′0.55″N, 81°42′58.19″W) and Henderson Creek (26°1′3.07″N, 81°37′58.00″W). Average nightly catch for each year is expressed as the mean catch for the area during May–June.

Human landing rate and BG-Counter data showed dramatic year-to-year fluctuations in population density of Ae. taeniorhynchus mosquitoes during the months of May and June (Fig. 1A, 1B, 1E, 1F). Data collected from human landing rates in 2016 displayed typical Ae. taeniorhynchus counts, with an average rate of 6.69 ± 3.37 and 6.70 ± 4.40 adult female mosquitoes for Marco Island (Fig. 1A) and Henderson Creek (Fig. 1E), respectively. In 2017, Collier County experienced a population explosion of Ae. taeniorhynchus, with an average landing rate of 18.08 ± 9.71 for Marco Island (Fig. 1A, 1D) and 21.37 ± 14.35 (Fig. 1E, 1H) for Henderson Creek. For some individual sites within Marco Island and Henderson Creek, Ae. taeniorhynchus landing rate counts often exceeded 100+ adult female mosquitoes within 2 min (data not shown). In 2017, BG-Counters were deployed by CMCD for the 1st time. During the months of May and June in 2017, BG-Counter trap data displayed an average nightly catch of 2,055.10 ± 1,789.43 in Marco Island (Fig. 1B) and 6,186.73 ± 4,676.99 in Henderson Creek (Fig. 1F). The following year, a population crash of Ae. taeniorhynchus was observed, with average landing rates of 0.86 ± 1.38 and 1.41 ± 1.89 for Marco Island (Fig. 1A, 1D) and Henderson Creek (Fig. 1E, 1H), respectively. Furthermore, BG-Counter trap data indicated an average nightly catch of 76.87 ± 80.60 adult mosquitoes per night in Marco Island (Fig. 1B) and 1,524.93 ± 1,442.26 in Henderson Creek (Fig. 1F). It is important to note that increased freshwater species in Henderson Creek in June (determined by landing rate counts) likely account for the high BG-Counter trap data in Henderson Creek. These population changes depict 95% reduction (P < 0.0001) for Marco Island (Fig. 1D) and 93% reduction (P < 0.0001) for Henderson Creek (Fig. 1H) in the Ae. taeniorhynchus population following Hurricane Irma. Taken together, these data represent large fluctuations in year-to-year population density of Ae. taeniorhynchus mosquitoes in Collier County.

Next, we asked whether heavy rainfall or unusual tide levels may have played a role in the dramatic reduction in Ae. taeniorhynchus populations in 2018. Historical trends have shown that increased rainfall from ENSO events may result in decreased levels of salt-marsh mosquitoes (Ritchie 1984). Meteorological data from 2016 to 2017 were obtained from the Big Cypress Basin monthly hydrologic reports (SFWMD 2018) and are expressed as monthly rainfall in inches. Spring rainfall in 2016–18 displayed typical trends; however, the summer of 2017 had above average rainfall from June to October (Fig. 2A). Tidal data from the Naples Station (8725110) were downloaded from the National Oceanic and Atmospheric Administration tide and currents database. Tide measurements analyzed consisted of monthly highest tide from January 2016 to October 2018 and is expressed as water level relative to mean sea level (MSL) (NOAA 2018). Historical average of highest tide was calculated using the previous 10 years of monthly highest tide data. Tide data from 2016 to 2018 indicated average high tides, with a slight rise in highest tide during May and June correlating to the salt-marsh season for Ae. taeniorhynchus in Southwest Florida (Fig. 2B). These high tide levels seen in May and June represent typical tides capable of intermittent flooding of Ae. taeniorhynchus breeding sites within salt-marsh and mangrove swamps. In 2017, an above average high tide (>5 ft [8 km] above MSL) was observed in September as a result of Hurricane Irma on September 10, 2017 (Fig. 2B). These observations suggest that increased summer rainfall and high tide as a result of Hurricane Irma may have had an influence on Ae. taeniorhynchus populations during the 2018 salt-marsh season.

Next, we asked whether other tropical disturbances have had an impact on Ae. taeniorhynchus populations. On October 24, 2005, Hurricane Wilma made landfall in Collier County. We obtained historical human landing rate data from 2005 and 2006 for 17 sites from Marco Island and 11 sites from Henderson Creek, as described above. Average Ae. taeniorhynchus populations were detected during the months of May and June in 2005, with 6.69 ± 4.31 adults on average in Marco Island (Fig. 1C, 1D) and 9.92 ± 2.85 (Fig. 1G, 1H) mosquitoes on average in Henderson Creek. Similar to 2018, Ae. taeniorhynchus populations were drastically reduced in 2006 following Hurricane Wilma. In May and June of 2006 an average of 3.20 ± 3.45 and 2.25 ± 1.20 adults were detected in Marco Island (Fig. 1C, 1D) and Henderson Creek (Fig. 1G, 1H), respectively. These data represent a 31% reduction (P = 0.0043) for Marco Island (Fig. 1D) and 77% reduction (P < 0.0001) for Henderson Creek (Fig. 1H) in the Ae. taeniorhynchus population following Hurricane Wilma. Along with human landing rate counts and trap data before and after Hurricane Irma, these results suggest that dramatic population crashes can be expected after a severe tropical disturbance.

Population crashes of Ae. taeniorhynchus have been associated with high water levels, with potential mechanisms including increased populations of larvivores, submergence of mosquito oviposition sites, and death of unhatched submerged eggs (Ritchie 1984). Abnormal, prolonged flooding that supports increased larvivorous predator populations was not observed in CMCD aerial inspections between 2016 and 2018; however, after Hurricane Irma we observed widespread damage and sediment deposition in mangrove swamps located in Collier County. While we cannot completely rule out physiological exhaustion (Ritchie 1984) from the 2017 population boom or another mechanism for the reduced salt-marsh mosquito numbers, we suggest that hurricane activity in 2005 and 2017 played a role in population crashes of Ae. taeniorhynchus. The apparent association of reduced Ae. taeniorhynchus numbers and hurricane activity suggest that high tides from Hurricane Wilma and Hurricane Irma likely submerged mangrove swamps harboring Ae. taeniorhynchus eggs—flushing breeding sites of unhatched eggs. In addition, damage and increased sedimentation within Ae. taeniorhynchus breeding habitats may have also affected mosquito populations. Reports from the National Aeronautics and Space Administration indicate that Hurricane Irma drastically affected the landscape of mangrove swamps located in the Everglades National Park and Ten Thousand Islands National Wildlife Refuge (NASA 2018). Hurricanes have been shown to shape the landscape in mangrove swamps of the Everglades through storm surge–associated sediment deposition (Smith et al. 2009). Further research is required in order to assess the effect of landscape changes to mangrove swamps and salt-marsh mosquito populations. Regardless, hurricane activity may serve as a useful predictor of salt-marsh mosquito populations in Southwest Florida.

To our knowledge, Ae. taeniorhynchus populations during 2018 are one of the lowest ever recorded for Collier County. Adulticide usage is directly correlated to mosquito surveillance data, with landing rate data driving the majority of CMCD's operational decision-making (Fig. 2C). We have observed a decreased usage of adulticiding materials following tropical disturbances (Fig. 2C, 2D), with usage increasing with higher landing rates during a “typical” season or population explosion (r² = 0.8616, P = 0.0228; r² = 0.9265, P = 0.0087) (Fig. 2C). An 80% reduction in adulticide usage (Fig. 2D) and 81% reduction in acres treated (Fig. 2E) was observed in 2018 compared with the previous year. The 2017 “salt-marsh season” led to approximately 63 adulticiding missions, including the output of over 3,500 gal of naled (Dibrom®; Adapco, Sanford, FL) and 170 gal of phenothrin (Anvil 10-10®; Clarke Inc., St. Charles, IL) (Fig. 2D). In comparison, 9 adulticiding missions targeting Ae. taeniorhynchus mosquitoes were performed in May and June of 2018. This included just over 500 gal of Dibrom, 170 gal of Anvil 10-10, and 30 gal of Merus™ 2.0 (Clarke Inc.) (Fig. 2D). Furthermore, nearly 1 million acres were treated with adulticide for Ae. taeniorhynchus in May and June of 2017 when compared with the following year's treatment of 180,053 acres (Fig. 2E). Similar results were observed before and after Hurricane Wilma, with an 89% reduction in both adulticide usage (Fig. 2D) and acres treated (Fig. 2E). Furthermore, over 3,000 service requests were submitted during the 2017 “salt-marsh season,” as opposed to 316 in 2018 (Fig. 2F). These results signify the impact of tropical disturbances and severe weather patterns on mosquito control operations. Therefore, an enhanced understanding of the causes of Ae. taeniorhynchus population booms and crashes may allow for improved operational planning and decision-making.

The authors thank the Collier Mosquito Control District (CMCD) Board of Commissioners and all the employees at CMCD who participated in trap maintenance, sample collection, and technical assistance, including Richie Ryan, Derrick Klein, Daniel Anez, Drew Luckow, Jorge Puente, McCoy Ward, and Jeffery Gallucci. We also thank CMCD's Communications Department, including Robin King and Andrea McKinney for technical editing and public outreach.