Australasian Science: Australia's authority on science since 1938

Australia’s Acoustic Environmental Accounts

Birds, bats, frogs and a few insects will dominate the data collected

Birds, bats, frogs and a few insects will dominate the data collected by the network but it is still a cheap and effective start to monitoring biodiversity at a continental scale

By Professor Hugh Possingham

A network of acoustic monitoring boxes spread across Australia’s bioregions could provide a cheap continent-wide biodiversity surveillance system providing feedback on how biodiversity is changing over time.

Interest in developing environmental accounts for Australia is enormous. The reason is simple – people have realised that if we can’t find credible and transparent metrics to quantitatively predict the consequences of policy and management on environmental issues like biodiversity, then those issues will always take a back seat to economic measures in policy development. This means that our growth in GDP, and other metrics like interest rates, will dominate our leaders’ attention rather than our declines in biodiversity.

This is an issue at all scales, from natural resource management bodies through to the Federal government and internationally. While many levels of government are now taking environmental accounts seriously, one of our greatest challenges is to cost-effectively gather repeatable and meaningful bio­diversity data in a continent of more than seven million square kilometres. Here I propose a possible solution that I stole from a variety of colleagues, including Gordon Grigg, Peter Cosier and Peter Grace.

The technology for measuring, recording and identifying sounds has been advancing rapidly. The hope is that soon we will have a cheap reliable box – let us call it the acoustic monitoring box – that will not just record and store sounds, but also identify them and create data files that indicate the number of calls of different fauna species over time.

As digital networks expand it should be possible for all of this data to be downloaded remotely – from say a computer in my office in Brisbane. Using existing sound libraries, many species can be identified now.

Unidentified sounds would be also be collected, stored and analysed in the future. Artificial intelligence software to identify novel sounds and report them to the user is still being perfected, but it’s not far off.

Indeed, if such a network could be constructed it’s important to get it going as soon as possible because statistically significant change is usually only detectable after many years.

So here’s the plan. Scatter, in a clever sciencey way, 2000 of these boxes across the continent. We have 85 bioregions, and most bioregions have about 10–20 structural vegetation types, so let’s say two boxes per bioregion vegetation structure combination (as a stratified random sample). In some cases it would be good to have the boxes close to an existing weather monitoring station because the Bureau of Meteorology will be looking after them (maybe I should have warned them about this).

Each box records an hour of sound four times per day, 1 day per week: the first starting at dawn, then an hour before dusk, an hour after dusk, and an hour in the middle of the night. All this data is then delivered to the analysts in terms of the number of calls of each species per day for each location.

Over the years we would be able to cost-effectively detect changes in the abundance of hundreds of species across the entire continent, all based on calling rate. It would be the first continent-wide survey of any group of fauna in Australia with consistent replication in space and time.

Why would we bother? In its simplest form it is background surveillance – like listening to the heart beat of the continent for an early warning sign of ill-health. Any disturbing changes would need to be investigated – like the call of a starling crossing the Nullabor.

At its best it provides a baseline against which we can start to assess the benefits of environmental interventions and test innovative management strategies. For scientists, it enables us to answer new questions, especially about the dynamics of many of our nomadic and/or erratic arid zone species.

Of course, the system is not comprehensive. Birds, bats, frogs and a few insects will dominate the data, and we won’t get much else. However, it is a cheap and effective start.

I figure that $8 million might buy and install the boxes, with $400,000 per year needed for maintenance and analysis. An endowment of $8 million would maintain it forever. That’s a total once-off cost of $16 million, or about one wing of a military airplane.

Professor Hugh Possingham is Director of the Applied Environmental Decision Analysis centre at the University of Queensland.