Nearly thirty thousand plants, and barely a fifth assessed

Every two years, the Royal Botanic Gardens, Kew, publishes a stocktake of what is known about plants and fungi, and this year’s edition landed on 16 June. Compiled by more than 400 scientists from over forty countries, the report is the tenth in the series, and its message is a study in both alarm and quiet optimism. The alarm is that the scale of the extinction crisis is still barely measured. The optimism is that digital tools are finally catching up to the task.

The numbers are the kind that stop the eye. Some 29,748 species of plants are now classified as threatened with extinction, alongside 411 species of fungi. Fewer than a thousand plant species have ever been formally declared extinct, which sounds like a small number until the authors explain why. Only about 18 per cent of known plant species have been assessed for their extinction risk at all, and for fungi the figure is 0.6 per cent. The known threatened list, in other words, is the visible tip of a much larger iceberg, and the true number of species that have already disappeared is almost certainly far higher than the roll of declared extinctions suggests.

The discovery side of the ledger is stranger still. An estimated 100,000 plant species and around two million species of fungi are thought to exist but have never been formally described. In the reporting year alone, scientists described 4,600 new plant species and 7,800 new fungi. Fungi are the dark matter of biology: a quarter of a million species classified so far, perhaps millions more waiting, and a group so under-studied that assessing their extinction risk is barely off the ground.

The report’s central theme, this year, is the digital biodiversity revolution. The idea is that the crisis cannot be fought with notebooks and field trips alone, because the baseline of what exists is still being drawn. Kew has just completed the full digitization of its herbarium and fungarium, some 7.4 million specimens, in the largest single scientific investment in the gardens’ 267-year history. The images and data are being opened to the world through the Global Biodiversity Information Facility, the portal that connects natural-history collections across continents. Machine learning and digitization, the authors argue, are turning identification, range-mapping and threat-assessment from the work of lifetimes into the work of algorithms.

The tension the report does not shy away from is that better data reveals a deeper problem. The more specimens are digitized and compared, the clearer it becomes how much has already been lost and how little anyone noticed. A species that has never been named cannot be on an official red list, and a fungus that no one has sequenced cannot be flagged as endangered. The revolution in measurement is, at the same time, a revolution in disclosure.

What the report asks for is patience as much as money. Assessing the remaining 82 per cent of plants and nearly all fungi is a generational project, and the species waiting to be assessed do not wait politely. Yet the digitized collections also mean that, for the first time, the workload is not strictly linear. The gap between what is known and what is unknown is, at last, being attacked from both ends.