AI could help win ‘race against extinction’ of vital plants, say botanists

The rise of AI and digitisation could be a turning point in the “race against extinction” faced by botanists trying to identify and save vital plants before they vanish, according to a major report from Royal Botanic Gardens, Kew.

New technology is enabling scientists to track how flowering times have shifted by weeks around the world, rapidly identify new specimens and even get crucial genetic data from 180-year-old fungus specimens, potentially opening a “genomic goldmine”. Digitisation and online access to millions of specimens that were until now only accessible in archives is also producing new insights, especially in the global south.

Plants and fungi underpin all life on Earth, supplying food and medicines, storing carbon and regulating the climate. However, about 40% of the 70,000 plant species that have been assessed are at risk of extinction, while another 330,000 have yet to be analysed. There are also believed to be another 100,000 plant species still to be named by scientists.

About 2,000 new plant species are recorded each year but this “barely scratches the surface” said Prof Alexandre Antonelli, the executive director of science at RBG Kew.

It means potential new medicines and sustainable crops are going extinct before even being discovered.

The situation is even more stark for fungi, with 90% of an estimated 2m species still unknown to science and less than 1% of known species assessed for extinction risk.

“While documenting and protecting all life on Earth remain formidable challenges, digitisation and accompanying technologies make me increasingly hopeful that we’ll succeed,” said Antonelli.

AI can learn how to identify challenging plants, for example, such as sedges and peat mosses whose distinguishing features are microscopic, meaning new or vulnerable species can be spotted faster. “These AI models can sometimes now identify better than specialists – that’s incredibly exciting,” he said.

Digitising images and collection data of plant and fungi specimens also speeds up international collaboration and can open up important but rarely accessed collections in biodiversity hotspots, such as Madagascar.

Landy Rajaovelona, a senior botanist at Kew Madagascar, says: ‘Madagascar is one of the world’s most extraordinary biodiversity hotspots. By digitising [37,000 physical specimens], we’ve unlocked a treasure of knowledge spanning centuries, offering invaluable insights into today’s biodiversity.”

RBG Kew has now digitised all 7.4m of its specimens, including those collected by Charles Darwin, and these are freely available online. The four-year programme involved taking 20,000 high-resolution images a day at its peak. In total, there are 145m digital specimens now online globally, but this is less than 16% of the total held in herbariums, leaving “huge blind spots in understanding”, the scientists said.

The report also features a global study using an AI model trained to spot flowers that analysed 8m digitised specimens. It revealed flowering has shifted by an average 2.5 days a decade over the last century due to the climate crisis. Changing rainfall patterns, as well as rising temperatures, have meant some flowers arriving later and others coming earlier.

This can severely disrupt ancient relationships between plants and the pollinators and other animals that depend on them at specific times of year. For example, a study using herbarium specimens showed that about 80% of the kindal trees in India’s Western Ghats, which are important for timber, used to flower at the same time. By the 1990s, that fell to less than half.

New technology is also unlocking genetic secrets from fungi, with scientists now able to produce high-quality genomes from very old specimens, some up to 180 years old. The researchers said the breakthrough makes historical fungarium specimens a “genomic goldmine” for new medicines and the prediction of disease outbreaks. Penicillin and statins were derived from fungi.

“Fungi are very opportunistic and they love heat and humidity,” said Dr Esther Gaya, senior research leader at RGB Kew. “Some human pathogens seems to be spreading from warmer places as the hot season gets longer in temperate regions.”

There are concerns over the heavy use of energy and water by AI datacentres, acknowledged by the report’s authors. The Guardian reported in May that datacentres are now consuming 6% of electricity in the UK and US. OpenAI’s boss, Sam Altman, said in February: “It also takes a lot of energy to train a human.”

The report, produced by 400 scientists across 40 countries, cautions that using digitisation and AI could amplify existing biases and inequalities unless the underlying data is expanded and improved. It called for partnerships between the technology companies and environment organisations and for governments and funders to invest in plant and fungi collections.