Our results demonstrate that Culex larval abundance varies spatially and temporally across urban landscapes and is strongly influenced by environmental characteristics, including the surrounding vegetation structure and aquatic chemistry. Using regression and machine learning techniques, we described the landscape features and fine-scale microhabitat characteristics of four Chicago neighborhoods and examined the implications of these measures for larval abundance in adjacent catch basins. We determined that aquatic pH, ammonia, and nitrates, terrestrial deciduous and flowering shrub area, and tree density are predictors of larval production in catch basins. However, the relative importance of these effects varies temporally, with aquatic chemistry influential in the early season and vegetation increasing in significance in the late season. These data may be used to inform mosquito control efforts by demonstrating the consequences of landscaping decisions on local mosquito production, and subsequently aid in the reduction of human risk of exposure to mosquito-borne disease in residential neighborhoods.
Potential effects of vegetation on larval abundance generally fall into two categories: First, trees and shrubs may offer resting habitats and sugar sources to adult females and their avian blood meal hosts, encouraging the mosquitoes to oviposit in nearby catch basins. Second, trees and shrubs may provide direct detrital inputs to the catch basins. Leaf litter and fallen fruit may affect mosquito production by influencing aquatic chemistry, introducing biomolecules to the larval habitat, and altering abundance and composition of the microbial community within the catch basin. Collectively, these effects have the potential to alter the attractiveness of the aquatic habitat to gravid females as well as the quality of the aquatic habitat for developing larvae.
Because there is often a positive correlation between abundance of adult mosquitoes and abundance of larvae in urban storm water systems [45, 46], shrubs may have an important role in determining larval production in catch basins by affecting the abundance of gravid females nearby. We found a marginally significant positive association between larval abundance and area of shrubs <1 m height within 25 m of each catch basin. In contrast, area of flowering plants was negatively associated with larval abundance. A possible explanation for this pattern is that male and female mosquitoes as well as potential avian blood meal hosts are attracted to non-flowering shrubs as resting sites and sources of fruit, encouraging gravid females to oviposit in catch basins and other aquatic habitats proximate to the site of their blood meals. Conversely, flowering plants containing pyrethroid compounds (e.g., chrysanthemum) and natural volatile oils (e.g., geranium, peppermint, and rosemary) may repel adult mosquitoes, and thus reduce the number of adults present and potential for oviposition in nearby catch basins [47, 48].
This relationship between vegetation proximate to catch basins and larval production was observed in trees as well as shrubs. Although linear regression indicated that total tree density was not correlated to larval abundance, this finding did not reflect broadly posed unimportance of trees in determining Culex production, but rather the sum of non-uniform effects of different genera on larval abundance. Regression tree and random forest models showed that key tree genera may have significant positive or negative effects on larval production. For example, in the early season, arborvitae and spruce trees were positively associated with larval abundance, suggesting that these shrubby, low-hanging trees serve a similar ecological function to shrubs in providing a resting and feeding site for Culex females and their avian hosts. The increased importance of pear trees to larval abundance during the fruit-bearing period in the late season indicates that trees not only provide habitat to adult mosquitoes, but may also offer a possible sugar source for females and males and an attractant to blood meal hosts . Few other fruiting trees were observed in the study area in sufficiently high densities for consideration in our models, and the effect of plums, mulberries, and other fruit-bearing trees on mosquito diet and host presence remains open to future study.
Vegetation also provides a direct source of detritus to nearby catch basins, with several potential implications for larval abundance. First, vegetation has been demonstrated to alter the aquatic chemistry of larval environments in naturally occurring and artificial container habitats, such as tree holes and used tires , and our findings in the catch basin ecosystem were consistent with these previous results. We found that larval abundance in catch basins was positively correlated with broad, seasonal trends in aquatic nitrate and ammonia variation. The organic detritus of plants introduced throughout the year, including pollen, flowers, fruit, seeds, and leaves, likely is a primary source of nitrogen in catch basins, explaining an increase in nitrates in catch basins during the late season after a critical mass of trees had dropped fruit and entered senescence [26, 27]. Elm trees, positively correlated to larval abundance, may have contributed substantially to this effect in our study area because their leaves lack a thick wax coating and therefore decompose quickly in water.
The impact of natural inputs on aquatic chemistry in the urban environment also may be compounded by artificial inputs and xenobiotics, such as lawn fertilizers, road salt, pesticides, and herbicides. Stemflow has been shown to dilute the chemical concentration of treehole microhabitats , and a similar effect of groundwater could explain additional variance in our models. Further, it is noteworthy that catch basins are subject to constant fluctuations in pH, ammonia, and nitrates, as we observed in the eight catch basins tested weekly for water quality concurrent with larval sampling. From our qualitative interpretation of that sample size, weekly larval abundance appears to track with these parameters. Our early season and late season models based on fewer aquatic chemistry measurements are significant because they capture the influence of broad scale, seasonal patterns in water quality [38, 39] on larval production, knowledge that is important to the accurate prediction of annual and inter-annual variation in mosquito abundance. Further research could clarify the function of fine-scale temporal aquatic chemistry fluctuations in determining vector production.
Second, vegetation introduces biomolecules to catch basin aquatic communities. These compounds, including tannins, cellulose, and glucose, may have significant effects on Culex production. While the catch basin sample size and water quality parameters taken in our current study were insufficient for correlation analysis examining the relationship between vegetation characteristics and biomolecules, prior studies have demonstrated that the compounds contained in the leaves of certain tree species may create environments inhospitable to Culex. For example, the tannins contained in oak leaves are toxic to the aquatic stages of mosquitoes , so that catch basins near these trees may have low larval abundance. Laboratory study could reveal whether detritus of ash and maple trees, such as wind-dispersed samaras, leaves, and branches, have comparable physiological effects on developing mosquitoes.
Finally, vegetation detritus may alter aquatic habitat quality and mosquito production via its effect on the microbial flora in catch basin mesocosms. It is understood that microbial abundance is related to leaf biomass in aquatic habitats [53, 54], and that microbes contribute to the larval diet through nutrient cycling and by providing a direct food source to developing larvae [55, 56]. However, microbial communities likely vary substantially among substrates of different leaf genera and thereby may alter mosquito production and fitness. One mechanism that may be responsible for associations between tree genera surrounding catch basins and mosquito production is differences in abundance and composition of microbial communities that grow on each leaf genus. These relationships between leaf detritus type, bacterial flora, and mosquito abundance and distribution should be examined in future investigations.