Drones can track critically endangered birds that scatter over vast areas
Cutting-edge technologies, such as autonomous drones, are being developed to track critically endangered parrots as they scatter widely over challenging terrain
NOTE: This piece was a Forbes Editor’s Pick.
Originally published under this title: “Tracking Critically Endangered Swift Parrots Using Drones”
As I’ve mentioned before (here), tracking parrots in the wild is quite difficult because they tend to wander over large, remote areas that frequently are difficult for humans to traverse. For this reason, it’s easiest for parrot researchers to find a place where wild parrots congregate on a regular basis, such as a clay lick (more here), and wait for them to come to you. But what if the parrots that you’re studying don’t make daily pit stops at one particular location?
This is precisely the situation faced by Australian conservation biologists working with the Critically Endangered swift parrot, Lathamus discolor. The swift parrot is cloaked in a variety of greens with a bright red moustache and throat patch, a brilliant blue forehead, turquoise cheeks, a horn-coloured beak, pale pink feet and a pointed tail. Although it’s a distant relative to the rosellas, its dietary requirements are similar to those of the lories, which are nectar-feeding parrots.
Thanks to rampant habitat destruction and runaway logging that is decimating the large, old trees that it relies on for nest hollows on Tasmania, the swift parrot is very rare, with fewer than 1000 pairs alive in the wild today. A more recent threat to these tiny parrots is the invasive sugar glider, Petaurus breviceps, an arboreal gliding marsupial that was introduced to Tasmania where it is not native. The nocturnal, squirrel-like sugar glider is an effective predator of small birds, animals and insects, and it often invades nest hollows in the dead of night and kills whatever it finds. This means that the swift parrot mortality rate is astonishingly high: sugar gliders alone kill as much as half of all nesting females, and their chicks and eggs, each breeding season. Thus, despite conservation efforts, the population of these parrots continues to be in steep decline.
The swift parrot is quite small: it’s not much larger than that common household pet, the budgerigar. Its color and size make this little parrot difficult to spot on a good day, but its habits make it even more difficult to locate because the swift parrot is a seasonal migrant and a long-distance winter wanderer. It breeds on the southern island of Tasmania and migrates to Australia where it disperses widely throughout the southeastern portion of the Australian mainland during winter.
This raises an important question for scientists: because swift parrots are nomadic, particularly in winter when they scatter unpredictably over a huge area, how can researchers identify which forests are most important to these rare birds so they can be protected from the worst ravages that people mete out?
Taking a cue from conservation biologists in Minnesota, who have been successfully using autonomous robotic drones to track invasive carp for years (i.e.: ref & ref) to reduce the manual labor required by traditional wildlife telemetry methods, Australian scientists and engineers have been developing drones to track swift parrots. But drones have not been tested for autonomously locating live radio-tagged birds, especially in rugged terrains. Until now, that is.
Australian researchers from the University of Sydney, Australian National University, and University of Technology Sydney, teamed up to publish a study that validates this robotic monitoring method, and to demonstrate that it performs comparably to experienced human trackers using radio telemetry.
According to the published study, direct comparisons of a drone system with human trackers for locating radio-tagged swift parrots in the wild revealed that in six of eight trials, the estimated locations of the target bird was within 50 metres (164 feet) of its true location, which was sufficient for recapture, detailed field observations, or data readout. In the remaining two trials, the target bird moved during the flight and visual confirmation was lost.
This study demonstrates that drones can locate radio-tagged birds in the wild faster and easier than can human trackers, and moreover, drones can find birds in a larger search area than those covered by human trackers. This study highlights the idea that robotic wildlife telemetry tracking is can be used either as an autonomous system or to assist researchers in real-life field conditions. It also is better than trudging across dangerous terrain for hours with one arm held over your head in search of a few radio-tagged study animals. Of course, I’m waiting for drones to be used to locate flightless kakapos, Strigops habroptila, another critically endangered parrot species that can be very difficult to locate on their rugged New Zealand island homes.
Certainly, validating the effectiveness of drones for tracking small radio-tagged animals will help motivate further engineering development that could enable more widespread use of autonomous robots to precisely monitor animal movements, particularly those of small, rapidly flying species such as swift parrots.
Oliver M. Cliff, Debra L. Saunders, and Robert Fitch (2018). Robotic ecology: Tracking small dynamic animals with an autonomous aerial vehicle, Science Robotics, 3:eaat8409 | doi:10.1126/scirobotics.aat8409
Originally published at Forbes on 7 November 2018.