Skip to main navigation Skip to main content
The University of Southampton
News

The high-tech hunt for a lonely plant’s partner

Published: 28 May 2024
Woman looking up at large plant
Dr Laura Cinti next to the E. woodii at Kew Gardens

Artificial intelligence is being employed in the hunt for a female partner for a critically endangered male plant species that predates dinosaurs.

A research project led by the University of Southampton is scouring thousands of acres of forest in South Africa, where the only known ‘Encephalartos woodii’ – or E. woodii – was found, in a bid to find a female.

And if the hunt doesn’t pay off, the researchers are also exploring if they can change the plant’s sex to create a female version.

The only known E. woodii was discovered in the Ngoye Forest, South Africa, in 1895. It was moved several years later for safe keeping, with samples sent to botanical gardens – including Kew in London – where it is still propagated and grown today.

But with only one male found, all subsequent propagated samples are male clones, so the plant cannot naturally reproduce. The Ngoye Forest has, until now, never been fully explored to determine if a female could exist.

Aerial image of tree tops
Drone footage of the tree canopy in the Ngoye Forest. Credit: C-LAB

Hunting with drones

Dr Laura Cinti , research fellow at the University of Southampton’s Winchester School of Art, is leading the first project to use drones and AI to search for a female E. woodii . She said: “This plant is, as far as we know, extinct in the wild. I was very inspired by the story of the E. woodii , it mirrors a classic tale of unrequited love. I’m hopeful there is a female out there somewhere, after all there must have been at one time. It would be amazing to bring this plant so close to extinction back through natural reproduction.”

Dr Cinti is collaborating with Dr Howard Boland, a creative technology expert working in AI, and Dr Debbie Jewitt, a conservation scientist and drone pilot based in South Africa.

Initial drone flights in 2022 took tens of thousands of images and used a multispectral sensor to capture features beyond what can be seen from high above with the naked eye – such as distinguishing whether plants are living or dead, and to identify species.

Aerial image of tree tops in bright colours
Aerial image with multispectral sensing. Credit: C-LAB

But with no E. woodii yet found, the drone imaging is ongoing – now with the power of AI. They have covered 195 acres of the Ngoye Forest so far – and there are 10,000 acres in total.

“With the AI, we are using an image recognition algorithm in order to recognise plants by shape,” explained Dr Cinti. “We generated images of plants and put them in different ecological settings, to train the model to recognise them.”

Dr Cinti is also working on a new partner project investigating whether it is possible to change the sex of the E. woodii using either chemical or physiological manipulation, and then generate vegetative plants from that material.

She said: “There have been reports of sex change in other cycad species due to sudden environmental changes such as temperature, so we are hopeful we can induce sex change in the E. woodii too.”

Black and white image of two men with trees
The last two E. woodii at Ngoye Forest in the early 1900s

The story of the E. woodii

The Encephalartos woodii is a type of plant known as a cycad. Cycads are the oldest surviving seed-bearing plants, dating back more than 300 million years and surviving multiple mass extinctions and environmental changes.

They are dioecious, which means they are either male or female, and they produce cones from which pollen is transported by insects for reproduction.

Despite their longevity, they are now the most endangered organisms on our planet, and the Encephalartos woodii is the rarest of all.

Dr Cinti said: “Cycads sell on the black market for hundreds of thousands of pounds, and are kept in cages in botanical gardens due to the risk of theft.”

More information is available on the project website .

Privacy Settings