It's hard to overstate the importance of rainforests in keeping the world a place we want to go on living in. Yet they are being cleared at a terrifying rate, in part because methods to verify their protection are failing. The Nature Conservancy think they have a solution by listening to the rainforests' voices, and researchers they have partnered with have published a paper in Science confirming its viability.
There is considerable goodwill, particularly in richer parts of the world, for saving rainforests. Products are marketed as "rainforest friendly" and tens of millions of dollars are donated to buy territory for protection. Many nations in which the rainforests survive legislate to protect portions. Yet all too often, the greed of ranchers and plantation owners triumphs. Vast areas that were supposed to be protected turn out to be anything but and, without good verification procedures, the culprits get away with it.
Forest monitors struggle to keep up with what is happening in areas that are, by definition, enormous and remote from population centers. Satellite images can flag complete destruction, but they do a poor job of measuring when a forest's diversity is degraded.
Researchers have started strapping small, solar-powered sound recorders to trees, setting them to eavesdrop at regular intervals, particularly dawn and dusk when the rainforest is most alive. The recorders provide an indication of the animal sounds for hundreds of meters in all directions. This marks a major advance over camera traps, which of course only point in one direction and are blocked from seeing far.
Reviewing several studies on the workings of these, Dr Zuzana Burivalova of Princeton University and co-authors report that these sound recorders supply a rich volume of information about the forests' true condition, far more than can be revealed by other remote sensors. Moreover, it is far cheaper to visit an area once to install a recorder than to stick around for extended measurements.
Some recorders do require return visits to collect the data, whereas others are located in places where data can be uploaded to the cloud, increasing savings still further.
Burivalova and colleagues also note some less obvious advantages. Once the data is uploaded, it can be analyzed by anyone, avoiding issues of researcher bias. Algorithms or deep learning programs can be used to tie sounds to their makers. Calls can be assessed in many ways, revealing both the number of noisy animals in the recorders' vicinity and the diversity of species that make them.
The authors call for “a global organization to host a global acoustic platform” to provide an enormous database of rainforest sounds, allowing comparisons between healthy and degraded rainforests half a world apart.