
Eric Nyquist
Tree identifier for tropical rarities
These days, AI-powered tree identification apps can help you name the species along your morning walk. But for scientists studying biodiversity in the tropics, tree identification is a real problem. “When you walk into a tropical forest site where you have a study going on, there can be hundreds of tree species all found in an area the size of a few footballs fields,” says Liza Comita, a professor of tropical forest ecology in the School of the Environment. Many of these species are hard to tell apart, especially when they are seedlings. Only a handful of people on the planet have the training and visual memory to do it well. “It can be a real bottleneck in terms of quantifying and monitoring biodiversity,” she says.
What if, Comita and her seed grant collaborators thought, there was an AI app trained on research forests in Panama where every tree has been painstakingly identified? Such a tool could let researchers around the world keep track of what’s sprouting in reforestation projects, for instance. So far, with their 2024 seed grant, the team has been busily photographing seedlings and experimenting as well with taking images that show wavelengths of light not visible to the human eye, something that might help in identifying and studying the plants.