Increasing human pressure on ecosystems complicates finding a way to effectively protect all species in nature and mitigate the escalating climate crisis.
One limitation in protecting endangered species is the lack of money for conservation projects. To tackle this issue, a team of researchers from Sweden, Switzerland and the United Kingdom has developed new software to determine where to most effectively establish protected areas to prevent species extinction in a region or country.
The researchers developed the new CAPTAIN program by relying on biology, environmental economics and computer science.
“To optimise the program, we simulate an artificial world that includes many species that are exposed to human pressure and climate change,” says computational biologist Daniele Silvestro at the University of Gothenburg and the University of Fribourg in Switzerland. He is the lead author of the new article published in the prestigious journal Nature Sustainability.
The software called CAPTAIN (Conservation Area Prioritization Through Artificial Intelligence) coordinates biodiversity data, a budget set aside for conservation, human pressure and climate change.
“We let the algorithm play the role of decision maker, just like in a video game, where the reward is the number of species saved from extinction at the end of the game. By repeating the game many times, the program learns how to best choose which areas to protect in this simulated world. After this training phase, the algorithm can be applied to real data.”
Exceeds conservation policies in rality
According to the researchers, the AI-optimised models outperform alternative conservation policies.
“Simply protecting the most species-rich areas or the largest extent of land based on a limited budget will result in losses of species that can be prevented. AI-based solutions can instead use the budget in the best way to minimise the extinction of species,” says Silvestro.
“In reality, we face many threats – from cities, mines and deforestation to climate change,” says Thomas Sterner, a professor of environmental economics at the University of Gothenburg and a member of the research team. “We need to understand and follow up changes in reality and take both the economy and ecology into account to optimise work with protected areas.”
Regular surveillance is crucial. Using the new tool, biodiversity can best be protected when detailed knowledge about the diffusion of species is available or can be collected, either through experts or new technology such as drone surveillance.
“Considering that not a single one of the 20 international goals for biodiversity set for the 2010–2020 period was met, it is clear that we need to reconsider how we devise effective conservation strategies,” says Alexandre Antonelli, a professor at the University of Gothenburg and Director of Science at the Royal Botanic Gardens, Kew, in Great Britain.
“We believe that AI can revolutionise species conservation in the near future. Our tool can already help decision makers and landowners make the best use of available data and stop the irreversible loss of biodiversity,” says Antonelli, who led the project together with Daniele Silvestro.
Contact:
Daniele Silvestro, affiliated with the University of Gothenburg and the University of Fribourg in Switzerland (speaks English), +41-26 300 88 57, daniele.silvestro@bioenv.gu.se
Alexandre Antonelli, professor of biodiversity at the University of Gothenburg and director of science at Royal Botanic Gardens, Kew, in London (speaks Swedish), +44-73 84 87 76 64, a.antonelli@kew.org
Thomas Sterner, professor of environmental economics at the University of Gothenburg, +46-708 163306, Thomas.sterner@gu.se
