Biodiversity Conservation Ecology Environment Nature News Round-up Projects Research Wildlife

Researchers use laser-mounted drones to find tiny insects in big landscapes

Researchers use laser-mounted drones to find tiny insects in big landscapes

How do you monitor some of the world’s smallest animals in one of the world’s largest environments? An international team of researchers have devised a solution to this problem using drones, laser scanning and machine learning. They then successfully used this novel method to identify the presence of ants in the East African Savanah. Their findings are published in the British Ecological Society journal Methods in Ecology and Evolution.

Instead of looking for the ants themselves, the researchers used drones and LiDAR (light detection and ranging) scanning to find a much bigger species, the whistling thorn acacia. Not only is this a keystone species of the East African Savanah, supporting grazers like elephants and giraffes, but it also has a symbiotic relationship with four species of ant which protect the trees from parasites in return for housing and nectar.

When ants are present on the trees they change the canopy shape, and it was this this ‘extended phenotype’ that let the researchers spot the presence of the insects.

The method was so successful that the researchers were able to survey an area of 16 ha and 9,680 acacia trees in just one hour-long drone flight, a task that would manually take 1000 hours to perform.

“Many plants have their phenotype altered when inhabited by insects.” Explained Dr Zhengyang Wang, Harvard University, lead author of the research. “But these are usually very difficult to survey at large scales. Our method provides a way to survey ‘extended phenotypes’ of insects and other pathogens quickly and effectively across large landscapes— through remote sensing and machine learning.”

Understanding the presence and ecosystem functioning of insects – the ‘little things that run the world’ – is an integral part of understanding the health of landscapes. The researchers also say that their method could be used to detect insect pests in agriculture or forestry plantations saving both time and money compared to ground-based surveys.