In Australia, the re-introduction of large, permanent water storage containers (known as rainwater tanks) as a response to climate instability has increased the risk of establishment of mosquito invasions and therefore the transmission of disease. An estimated 300,000 rainwater tanks have been installed into south east Queensland since 2006. This project aimed to identify the risk they pose.
To understand the role that rainwater tanks played in past epidemics of vector-borne disease, such as dengue, this thesis reviewed the history and potential mechanisms for the disappearance of Aedes aegypti (L.) from its historical, southern range in Australia. Government surveillance records from Brisbane (the capital of Queensland, Australia) were collated from the period when Ae. aegypti was extant until its last detection in 1957. Several factors were likely responsible for the disappearance of the species including the introduction of reticulated water and the removal of rainwater tanks, and increased mosquito surveillance backed up by regulatory enforcement. This thesis demonstrates the value of historical data to help identify potential invasion risks and indicates the dramatic impact that mosquito surveillance, supported by government regulations, can have on Ae. aegypti persistence at the margins of its distribution.
Field observations and laboratory trials were conducted to measure the temperature buffered environment of rainwater tanks and determine growth and survival of Ae. aegypti in these containers during winter in Brisbane. Abiotic conditions in rainwater tanks were measured and data loggers were placed outdoors in buckets and rainwater tanks during winter and summer. For laboratory experiments daily temperature fluctuations were selected based on the coldest week of the year. Tropical and sub-tropical colonies of Ae. aegypti were established from material collected from different location in Queensland to test for their capacity to survive Brisbane winter temperatures. Greater than 48% survival (egg to adult) was observed in all treatments and mean adult development times (egg to adult) took 32 days. As such, a population of Ae. aegypti could survive in both buckets and rainwater tanks exposed to the coldest water temperatures in Brisbane.
To test the hypothesis that climate played a role in the disappearance of Ae. aegypti in south eastern Australia, I exploited an existing bio-climatic model (using CLIMEX). Survey records from the past 100 years were collated for the species and combined with a number of biological parameters to predict: a) the historical distribution based on the presence of permanent water storage (rainwater tanks), and b) a natural distribution based on the absence of artificial and permanent water sources. Predictions for Australia were validated against a) international distribution records, and b) the annual population phenology of Ae. aegypti from Cairns, Botucatu and Sao Paulo state. If suitable water storage was present, such as rainwater tanks, climate was unlikely to have played a major role in the disappearance of Ae. aegypti in eastern Australia. However, when permanent water storage was absent, the majority of locations tested from the historical distribution were unfavourable for the persistence of the species.
To examine the role that rainwater tanks play in the presence and movement of Aedes species in urban environments, a mark, release and recapture experiment was designed and undertaken. Results suggested that roads act as barriers to male Ae. aegypti but not to female Ae. notoscriptus (Skuse), another disease vector that inhabits rainwater tanks. Female Ae. aegypti were four times more likely to be captured in traps placed proximal to rainwater tanks than in locations where rainwater tank were absent. Finally, an ‘isotropic Gaussian dispersal kernel’ was used to describe mosquito dispersal as a diffusion process of movement over time. This provides compelling evidence that permanent domestic water storage facilitated the historical persistence of Ae. aegypti across much of eastern Australia.
This thesis examined the role that a network of unsealed rainwater tanks may play in the spread of Ae. aegypti in Brisbane. Using a spatially explicit network model in Repast Simphony I simulated the movement of mosquito populations through known locations of rainwater tanks across 140 Brisbane suburbs. Movement was parameterized using an isotropic Gaussian dispersal kernel, a maximum distance of movement, average life expectancy and the probability of Ae. aegypti crossing a road over a two week period that was empirically derived. The model was run against a number of scenarios that examined population spread through rainwater tanks based on compliance rates (unsealed or not) and real road grids. Finally, this thesis described key network properties and produced risk maps to inform areas for future mosquito and rainwater tank surveillance.
The outcomes of this thesis demonstrate that large, permanent water storage containers present ideal larval habitat for the re-establishment and persistence of Ae. aegypti in areas south of its current distribution in Queensland, Australia. Large numbers of rainwater tanks have been present in south east Queensland for over a decade with little monitoring of their structural integrity or vector presence. Ongoing management of rainwater tanks in south east Queensland will be essential to prevent the establishment of disease vectors and the consequent risk of seasonal transmission of disease.