The arroyo chub, Gila orcutti, is native to the watersheds of southern California and has been proposed as a replacement for the non-native western mosquitofish, Gambusia affinis, as a biological control agent for mosquitoes in sensitive watersheds throughout the region. Because mosquito-eating fish are not strictly larvivorous, they also have the potential to impact non-target organisms in systems where they are introduced. The impact of arroyo chub stocking density on the invertebrate community, and the suitability of a riverine constructed wetland as conservation habitat for the arroyo chub population, in the Santa Ana River were evaluated at the Prado Wetlands in Riverside County, California. Invertebrate abundance and community structure did not differ significantly among three stocking densities (0, 1.5 and 6 kg/ha) during a five-week study in cage mesocosms. Approximately 16 months after stocking G. orcutti into a 0.9-ha wetland, nearly 3,700 fish were collected; however, no arroyo chubs were recovered. All of the fish collected were invasive species in the Santa Ana River system and piscivory by non-native species most likely caused the extirpation of the arroyo chubs in the wetland.

The ovipositional responses of female Culex tarsalis to semiochemicals produced by aquatic organisms in three aquatic guilds (predaceous insects, algivorous fish and predatory/larvivorous fish) were examined in laboratory choice assays. Previous binary choice trials have shown that female mosquitoes avoid laying eggs on water that contains fish exudates and will instead choose to oviposit on water without kairomones released by fish. Culex tarsalis was not deterred from ovipositing in cups containing semiochemicals from invertebrate predators and algivorous fish, but was deterred from ovipositing in cups containing arroyo chub semiochemicals.

Mosquitoes are important vectors of many devastating human diseases. Much of the research to date has focused on their adult biology as vectors, but the field of microbiome research is rapidly developing. Commensal and symbiotic bacteria may play important roles in nutrition, digestion, or chemical defense for mosquitoes. The effects of larval mosquito presence and nutrient input on aquatic and larval mosquito microbiomes and on ecosystem structure and function were studied in 1 m2 mesocosms. Three levels of enrichment with organic and inorganic nutrients were cross-classified with larval mosquito presence or absence. Water quality variables (total nitrogen, nitrates, nitrites, ammonium, total phosphorous, and chemical oxygen demand), mosquito oviposition rates, invertebrate abundance, phytoplankton biomass, planktonic particle size spectra, primary production, community metabolism, and microbial diversity were measured during the 1-month study. The oviposition rate of Culex mosquitoes was directly related to nutrient enrichment. Mosquito presence significantly influenced phosphorus concentration, but did not significantly affect the concentration of other water quality variables. The presence of larval mosquitoes with high levels of nutrient input increased the photosynthetic productivity of aquatic mesocosms, with a corresponding decrease in bacterial diversity. Conversely, when mosquitoes were prevented from ovipositing in mesocosms, water column bacterial diversity increased at high levels of nutrients. This work shows that larval mosquitoes have profound impacts on aquatic ecosystems, especially when those ecosystems are affected by high nutrient input.

Mosquito management in wetlands is complex partly because some of the wetland management operations are known to enhance mosquito production. I aimed to describe the effects of nutrients and integrated mosquito management strategies on mosquitoes produced in wetlands in the following three studies. First, we evaluated the growth characteristics of alkali bulrush (Schoenoplectus maritimus), an alternative aquatic plant of potential bioremediation importance, and its associated mosquito production across a gradient of nitrogen enrichment including the high ammonium nitrogen regimens typically found in constructed wetlands. Mosquito larvae abundance, dominated by the western encephalitis mosquito (Culex tarsalis Coquillett), was significantly greater in enriched versus unenriched mesocosms. Alkali bulrush survived high nitrogen loadings, but its biomass was significantly suppressed at high (>50 mg/liter) ammonium nitrogen concentration in the water. Second, we characterized microbiota associated with C. tarsalis larvae, water column and epibionts of two (alkali and California) bulrushes using 16S rRNA gene sequences generated with Illumina sequencing platform. The diversity of microbiota in mosquitoes sampled from wetlands containing the two bulrushes did not differ appreciably, and was chiefly dominated by Thorsellia (Gammaproteobacteria). The epibionts of the bulrushes and bacterioplankton were dominated by members of Actinobacteria, Bacteroidetes, Cyanobacteria, Proteobacteria and Verrucomicrobia. Overall, nearly 49% of the Bacteria taxa found in the mosquito gut were also found in the habitat, suggesting a strong Culex larvae-Bacteria interaction. Finally, we evaluated the effects of a one-time application of two rates of a key mosquito biopesticide, Bacillus thuringiensis subsp. israelensis (a granular form of VectoBac G) on native microbiota and physicochemical variables in the feeding zone of Culex larvae in experimental mesocosms. Beta diversity of Bacteria communities revealed that samples from low Bti and untreated control mesocosms were significantly separated from high Bti mesocosms and were dominated by unidentified OTU of Cyanobacteria, Cytophagales and Cyclobacteriaceae (Bacteroidetes), and Sphingomonas (Alphaproteobacteria). Phytoplankton (chlorophyll a), sestonic particulates, nutrients, pH and other physicochemical variables in the water column were also significantly reduced in the high Bti mesocosms. Important implications of the results of these studies on integrated mosquito management in aquatic ecosystems are discussed.

Culex tarsalis, the western encephalitis mosquito, predominantly uses large water bodies as oviposition sites. These water bodies are home not only to macroinvertebrate predators, but larger predators such as fish. The western mosquitofish, Gambusia affinis, has been used worldwide for the control of larval mosquito larvae for several decades.. My research demonstrates that Cx. tarsalis can detect the presence of semiochemicals associated with the western mosquitofish, Gambusia affinis, without the fish being physically present, leading to a decrease in oviposition rate.

Oviposition onto fish-conditioned water was reduced relative to that in control treatments in bioassays using different densities of fish to condition water, in no-choice assays comparing mosquito oviposition on fish-conditioned water with aged tap water controls, and in bioassays using aged tap as well as pond water against Gambusia-associated semiochemicals at different concentrations to simulate “real-world” conditions. Assays with individual females provided responses to fish-conditioned water that were congruous with the results from binary choice experiments using multiple gravid females in replicate cages. Mosquito oviposition onto fish-conditioned water that had been filter-sterilized was reduced relative to filter-sterilized control water suggesting that the semiochemical was still present after bacteria were removed. Female Cx. tarsalis were deterred from ovipositing in the majority of trials conducted, and the response appears to be correlated with the concentration of Gambusia-associated semiochemicals present in oviposition sites.

I identified two non-volatile and three volatile chemical compounds in water in which G. affinis had been held using a variety of chemistry techniques. Oviposition bioassays were conducted using standards of the three volatile compounds identified: dimethyl disulfide (DMDS), dimethyl trisulfide (DMTS) and S-methyl methanethiosulphonate to determine their role in mosquito oviposition behavior. The non-volatile compounds isolated from the Gambusia-treated water could not be fully characterized, but one appears to contain a ring structure similar to that of cholesterol. Female Cx. tarsalis were deterred from ovipositing into sites containing solid phase extracts of Gambusia-treated water. This may indicate that it is the non-volatile compounds that the females are detecting, leading to a decrease in oviposition rate.