Scientists need better ways to survey wildlife, and one possibility is drones. So naturally, a team of scientists decided to scatter thousands of fake birds across a landscape to test the accuracy of the technology.
Their conclusion? Drones can be more accurate and less biased than traditional counting methods, according to the research published today in the Methods in Ecology and Evolution. The findings don’t come as a total surprise—other studies, like this recent one on seals, have concluded that drone surveys require less effort than humans on the ground, while being at least as accurate. Drones have also risen to prominence as a tool for keeping a watchful eye on wildlife threatened by poachers.
But as a reminder, this study used plastic birds—and for a good reason.
“In a wild population, the true number of individuals is not known. This makes it very difficult to test the accuracy of a counting approach,” Jarrod Hodgson, an ecology PhD candidate at the The University of Adelaide and lead author on the new study, told Earther in an email.
To overcome that uncertainty, Hodgson and his colleagues created life-sized replicas of the Greater Crested Tern, a common seabird. They set thousands of these fake birds out in ten “simulated Greater Crested Tern breeding colonies” on a metropolitan beach in Adelaide.
Then, the challenge began. Experienced wildlife counters armed with binoculars or telescopes were pitted against citizen scientists armed with imagery captured by off-the-shelf quadcopters, to see who more accurately estimate the actual number of “birds.”
To cut right to the chase, the noobs armed with aerial photographs won—their counts were 43 to 96 percent more accurate, on average. The researchers then put the citizen scientists out of work, by devising a computer algorithm that could count the plastic birds automatically with just as good results. All hail our machine overlords.
“This will be of particular interest in today’s research environment where funding for conservation is limited and researchers are under ever more pressing time constraints,” the researchers wrote.
There are some limitations to this study. Wildlife that can be seen clearly from above, like aggregating birds and seals, are particularly well-suited to drone surveys, Hodgson noted. “Additional experiments will be useful to assess the ability of drones to survey animals that prefer to stay hidden and those within complex habitats,” he said.
David Steen, a research ecologist at the Georgia sea turtle center and executive director of The Alongside Wildlife Foundation, told Earther in a Twitter DM that studies like this are valuable because they “generate defensible estimates to conclusively show that they represent an improved method for surveying birds in a nesting colony.”
He added that it’s probably important to replicate the research with animals that are actually moving around, or camouflaged. Finally, how real birds (or any animals) would respond to a drone flying overhead will be important to suss out, Steen and Hodgson both noted.
So, still some work to be done, but thousands of plastic birds later the future of wildlife monitoring is looking a bit more promising.
That’s good news, because as managers are forced to make increasingly high-stakes decisions to protect rapidly-dwindling populations, we need all the good data we can get.