Bird calls reveal population secrets
Sensor networks developed by QUT researchers are
tracking rare birds without disturbing the population.

Remote-controlled sensor networks developed by the Queensland University of Technology (QUT) are helping scientists track rare bird populations living at Brisbane Airport.

Acoustic signals captured by a network of smart phones track the location and population of different bird species, including the elusive Lewin's Rail, said Richard Mason, from the Microsoft QUT eResearch Centre in the Faculty of IT.

"Sound provides the heartbeat of the environment," Mr Mason said. "Like listening to someone's heartbeat, it reveals a lot of information about health."

Mr Mason said Lewin's Rail was a difficult bird to study because of its shyness, camouflage and ground-dwelling nature.

However, as a direct result of the sensor network, other researchers, such as QUT ecologist Jennifer Gibson, already had a better understanding of the activity periods of the birds.

Brisbane Airport is situated on an area containing wetlands and other habitats suitable to the Lewin's Rail. Their habitat requirements are currently being researched by Ms Gibson.

Mr Mason said it was important to monitor bird populations in such areas, but the traditional methods of carrying out surveys with observations on site, often by a number of observers in different areas at the same time, could be time-consuming and expensive.

"Previously when biologists wanted to take a census or survey bird populations like this, they had to rise at 3:00 am to be on site before dawn and continue to make frequent labour-intensive site visits throughout the day to establish activity periods of interest," he said.

"These repeated visits raise the risk of disturbing the very creatures under investigation, altering their behaviour in response to the presence of biologists and disturbing the habitat."

By contrast, the QUT wireless acoustic network is based on four smart phones which have the site wired for round-the-clock sound, and a web-based interface for sensor control and data management, Mr Mason said.

Information collected via the smart phones is fed into an acoustic database with automated pattern recognition software developed for bird species' recognition.

Crucially, all data can be viewed remotely, minimising disruption to the birds' natural behaviours and significantly reducing the time spent on site by the researchers.

Mr Mason said that while the network had been designed for this specific application, it was sufficiently generic for use in a wide range of other situations.

"In future, that could mean smart phones are not just for the birds, but also frogs, bats and other wild creatures," he said.

Editor's Note: Original news release can be found here.