If a rare, ancient tree burns in a bushfire but no-one has ever marked it on a map, would we ever notice?
Scientists don't think we would. And that's why a new project to "fingerprint" vulnerable trees and pick them out of aerial footage is taking off.
The need for the project was apparent after a dry lightning strike hit the rugged terrain of the Gell River region in Tasmania's south-west in 2018, marking the beginning of a devastating bushfire season.
The subsequent fires, which began in the Gell River region, burnt through 210,000 hectares — destroying several homes and requiring the evacuation of hundreds of people.
But 6 per cent of Tasmania's World Wilderness Heritage Area also went up in flames.
A review later found the large majority of threatened vegetation in those areas had evolved some ability to cope with fire, four areas that burned held ancient, vulnerable species and are expected to never recover.
Part of the challenge of protecting these rare, fire-vulnerable species stems from the fact we do not know exactly where they all are.
Jayne Balmer is part of the team working to fix that.
She is a senior ecologist at the Department of Natural Resources and Environment Tasmania.
"We're trying to improve the mapping of fire-sensitive communities in the Tasmanian Wilderness World Heritage Area … among those are things like Huon pine," she said.
Huon pine is the longest-lived tree species in Australia — with records showing some giants have lived for thousands of years.
Like King Billy pine, pencil pine and fagus, the Huon pine species evolved before fire became a major part of the landscape.
"These communities date back in their ancestry to Gondwanan times — Huon pine is about 68 million years old in terms of its evolutionary history," Dr Balmer said.
Vegetation mapping for Tasmania's Wilderness World Heritage areas began in 1990 — and while some information is collected on foot, most of the mapping is based on interpretations of images taken from aircraft.
But it's tricky to pick out specific trees and plants from so high above.
"They all look green and so it's a bit like 'Where's Wally' from the air, as well as when you're looking at aerial photos," Dr Balmer said.
"Conventional red, green, blue visible spectrum aerial imagery doesn't clearly highlight any differences between Huon pine and the species it co-occurs with.
Dr Balmer approached University of Tasmania professor Arko Lucieer, an expert in high-precision aerial surveys, for help.
Hyper spectral sensors are a special type of camera that can capture 170 bands of light within each image — rather than just the red, green and blue that humans see — allowing researchers to gather images in the near-infrared.
Because there are subtle differences in the way every species of tree and plant reflects light, they each have a kind of spectral 'fingerprint'.
"So it's this unique spectral fingerprint of Huon pine that we're trying to detect from the air, and with the drone technology, work out the right combination of sensor settings to achieve this," said Professor Lucieer, who is the head of the School of Geography, Planning, and Spatial Sciences.
They started by measuring the 'fingerprint' of a Huon pine at the Tasmanian Botanical Gardens in Hobart, before taking the technology to the south west wilderness and flying over areas where they knew Huon pines are found.
The team also developed a computer algorithm to analyse every pixel in the mass of imagery they collected.
They found that using this algorithm, they could successfully make the Huon pines pop out amidst the other vegetation.
"There's a handful of teams globally working with hyper spectral remote sensing on drones so it's quite a specialist area," Professor Lucieer said.
But there are limits on where, and how far, drones can fly.
Now, they're hoping to collaborate with aircraft operators in the state to test the technique on a larger scale.
"The drones have allowed us to fly low and slow and trial this technique, what's next is to mount our sensors on an aircraft and fly much larger areas, so we can map larger remote areas in the south-west … and produce a map of the Huon pine trees," Professor Lucieer said.
In the future, the technique could be applied to map other species — allowing them to monitor the biodiversity of a region and the impacts of climate change, he said.
