- Joined
- Jan 20, 2026
- Messages
- 345
- Reaction score
- 2,392
Old herbariums suddenly became a map to the species that science risks losing forever.
Botanical archives have kept millions of samples available only to a narrow circle of specialists for decades, but AI and digitization turn old collections into a tool for the search and conservation of plants and mushrooms that may disappear sooner than the scientific description is obtained.
A major report by Royal Botanical Gardens (Kew) shows that new technologies are accelerating the work with species on which food, medicine, carbon storage and climatic processes depend. About 40% of the 70 000 already estimated plant species are at risk of extinction, another 330 000 species have not yet passed such a rating. Approximately 100 000 plants, according to experts, remain unnamed.
For mushrooms, the gap is even greater. Science knows only a portion of the estimated 2 million species, and the risk of extinction is estimated for less than 1% of the mushrooms already described. Because of this gap, potential sources of new drugs and resistant crops can disappear before specialists have time to study their properties.
AI helps to recognize complex groups of plants, where the differences are visible only under a microscope. Such models find new or vulnerable species faster among samples of sewage, peat mosses and other difficult to detect plants. Digital pictures and collections also make it easier to work together between countries and provide access to materials from regions with high biodiversity, including Madagascar.
Kew has already digitized all 7.4 million of his samples, including Charles Darwin’s collections. At the peak of the project, the team took 20 000 high-resolution images per day. The world now has 145 million digital samples available, but this array is less than 16% of the materials that store herbariums.
Digitized collections have already helped show how the climate is changing plant life cycles. The AI model studied 8 million samples and revealed that over the past century, flowering dates have shifted by an average of 2.5 days per decade. Due to the change in temperature and precipitation, some plants begin to bloom earlier, others later, which can disrupt ties with pollinators and animals dependent on seasonal rhythms.
Technology also opens access to the DNA of old mushroom samples. Experts were able to obtain high-quality genomes from specimens up to 180 years. Such data helps to look for substances for future drugs and better understand the spread of mushroom-related diseases.
The authors of the report admit that AI has its own risks, including the high expenditure of energy and water in data centers, but without such tools, a significant part of the biological collections will remain invisible to science.
The report also warns that digital tools can reinforce existing distortions if the databases remain incomplete. For such systems to bring more benefits, broader collections, partnerships between technology companies and environmental organizations, as well as financing plant and mushroom archives are needed.
Botanical archives have kept millions of samples available only to a narrow circle of specialists for decades, but AI and digitization turn old collections into a tool for the search and conservation of plants and mushrooms that may disappear sooner than the scientific description is obtained.
A major report by Royal Botanical Gardens (Kew) shows that new technologies are accelerating the work with species on which food, medicine, carbon storage and climatic processes depend. About 40% of the 70 000 already estimated plant species are at risk of extinction, another 330 000 species have not yet passed such a rating. Approximately 100 000 plants, according to experts, remain unnamed.
For mushrooms, the gap is even greater. Science knows only a portion of the estimated 2 million species, and the risk of extinction is estimated for less than 1% of the mushrooms already described. Because of this gap, potential sources of new drugs and resistant crops can disappear before specialists have time to study their properties.
AI helps to recognize complex groups of plants, where the differences are visible only under a microscope. Such models find new or vulnerable species faster among samples of sewage, peat mosses and other difficult to detect plants. Digital pictures and collections also make it easier to work together between countries and provide access to materials from regions with high biodiversity, including Madagascar.
Kew has already digitized all 7.4 million of his samples, including Charles Darwin’s collections. At the peak of the project, the team took 20 000 high-resolution images per day. The world now has 145 million digital samples available, but this array is less than 16% of the materials that store herbariums.
Digitized collections have already helped show how the climate is changing plant life cycles. The AI model studied 8 million samples and revealed that over the past century, flowering dates have shifted by an average of 2.5 days per decade. Due to the change in temperature and precipitation, some plants begin to bloom earlier, others later, which can disrupt ties with pollinators and animals dependent on seasonal rhythms.
Technology also opens access to the DNA of old mushroom samples. Experts were able to obtain high-quality genomes from specimens up to 180 years. Such data helps to look for substances for future drugs and better understand the spread of mushroom-related diseases.
The authors of the report admit that AI has its own risks, including the high expenditure of energy and water in data centers, but without such tools, a significant part of the biological collections will remain invisible to science.
The report also warns that digital tools can reinforce existing distortions if the databases remain incomplete. For such systems to bring more benefits, broader collections, partnerships between technology companies and environmental organizations, as well as financing plant and mushroom archives are needed.
