Evaluating disease-vector mosquitoes on green roofs and in other urban green spaces : towards mosquito-sensitive urban design

Abstract

Urban green infrastructure (UGI) constitutes a core component of sustainable urban planning. Green spaces can provide a multitude of ecosystem services and mitigate challenges such as climate change and urban heat islands. However, uncertainty remains as to whether such spaces can inadvertently engender habitats for disease vectors such as mosquitoes. Moreover, urbanization itself presents a number of conditions that may predispose populations to emerging and re-emerging infectious diseases. The lack of understanding in urban vector ecology represents a gap in the current knowledge and thus serves as a persistent hindrance to effective greening policies.

The chief objectives of this research are: 1) to evaluate the urban habitat preference of female vector mosquitoes for extensive green roofs in a humid subtropical climate vis-à-vis positive and negative control sites; 2) to identify the underlying microclimatic factors that explain site differences in vector abundance; 3) to investigate whether male presence also differs across site types; and 4) to identify knowledge gaps and appraise potential mosquito risks of other urban green infrastructure.

To these ends, three empirical studies and a comprehensive survey of entomological literature on UGI were undertaken. Adult mosquito surveillance was conducted for a year from 2015 to 2016. Data were collected fortnightly from seven sites representing three experimental treatment groups: green roofs (GR), low-elevation blue-green spaces as positive controls (PC), and bare roofs as negative controls (NC). Human-biting mosquito species known to transmit diseases were identified. Adult vector abundance (VA) captured by baited traps can gauge human-mosquito contact risk. Concurrent microclimatic parameters were measured to evaluate site-specific microclimatic effects on VA.

Generalized linear models (GLM) revealed site type to be a significant predictor of vector mosquito abundance, in that PC had considerably higher VA than GR and NC, despite all sites being located within the normal flying range of the studied mosquitoes. Therefore, contrary to common preconceptions, urban green roofs are not particularly preferred by mosquitoes compared with control sites.

GLM with time-lagged and no-lag microclimatic measurements consistently identified temperature and wind speed as significant factors. Wind speed accounted for VA differences across site types, while temperature exerted its effect on VA mainly through seasonal variation. High wind exposure on elevated buildings, which impeded mosquito flight, was found to be the underlying microclimatic factor.

A stringent binary measure of male presence/absence through logistic regression revealed significantly higher male presence in PC than GR and NC. No significant difference was found between GR and NC. The results suggest male presence as a potentially convenient and rapid site assessment measure for evaluating mosquito biting risk.

The extensive UGI review identified several high-risk UGI elements, all of which contain a conglomerate of pertinent mosquito life-cycle resources. Together with the empirical findings, a mosquito-sensitive urban design approach that encourages evidence-based and proactive mosquito-resource interventions within the scope of the World Health Organization’s Integrated Vector Management, is proposed. The findings provide strategic insights and practical recommendations for urban landscape design and planning, as well as disease control and management.