A palm can be a familiar sight on a promenade and almost absent from its native habitat. This difference changes how we view a plant often reduced to a holiday silhouette. The number of cultivated specimens tells us neither how many wild populations remain nor whether they can still regenerate.
The exhibition “Tropical Autumn – Palms, Treasures and Secrets”, scheduled to run until 23 November 2026 at the Grandes Serres du Jardin des Plantes in Paris, offers a starting point for exploring this diversity. Behind the large leaves lie very different ecological stories. To understand the threats, we need to move from the scenery to the species, and then from the species to the territory it actually occupies.
A familiar silhouette, very different habitats
The National Museum of Natural History presents palms through their morphology, habitats, history and uses. These monocotyledons develop, among other things, a stem, often mistaken for a trunk. The exhibition route highlights their role in food, crafts and culture, as well as the resources they provide for other organisms.
This diversity means we cannot speak of “the palm” as a single species that would respond in the same way everywhere. A plant tended in a greenhouse or garden and a population established in a forest do not live under the same conditions. Seeing the former grow is not enough to conclude that the latter is safe. The right indicator depends on the question: are we trying to find out whether a plant survives, or whether a population has a future in its habitat?
Madagascar shows what a threat figure represents
A study by Mijoro Rakotoarinivo and colleagues, published in PLOS ONE in 2014, provides a particularly well-documented example. It analysed assessments of 192 palm species endemic to Madagascar, meaning they occur naturally only on this island. The assessments were published on the IUCN Red List in 2012.
The figure often cited is 83% of species threatened. But this represents neither the proportion of individual palms that have died nor a forecast of extinction within a set period. It is the central estimate of the proportion of species exposed to a risk of extinction, with explicit treatment of those for which knowledge was insufficient.
The authors had sufficient data for 179 species; 149 were classified as vulnerable, endangered or critically endangered. A further thirteen lacked information. Depending on how these unknowns were accounted for, the estimated proportion ranged from 78 to 84%. The 83% figure assumed that poorly known species were threatened in the same proportion as the assessable species. These historical results shed light on a method and a situation; they are not a new census conducted in 2026.
The problem often starts with habitat
The study drew on 2,160 geolocated observations and specimens, collected between 1834 and 2010, as well as fieldwork. A large proportion of Madagascar’s palms depended on humid forests. Degrading or fragmenting these forests could reduce plant survival and the establishment of new generations.
Two mechanisms compounded the pressure. The first involved changes to the environment, particularly for agriculture. The second was the direct exploitation of certain plants: for building materials, palm hearts or harvesting for the horticultural trade. The authors distinguished these pressures rather than reducing them to a single cause.
This precision is useful in the face of overly simplistic explanations. Climate change is one of the issues in conservation, but it does not, on its own, explain every documented threat to every population. In this study from Madagascar, habitat loss and the use of biological resources were at the forefront. Protecting palms therefore also requires understanding local residents’ needs and how resources are harvested.
Better knowledge of a species can change its classification
Another less intuitive finding: an improvement in official status does not necessarily mean that the situation on the ground has improved. Comparing assessments from 1995 and 2012, the researchers noted several changes linked to a better understanding of populations and their distribution.
A species thought to be restricted to one place may be found elsewhere. Its risk is then reassessed, even though no new forest has grown between the two studies. Conversely, a recently described species may immediately appear highly threatened because it is known only from a small area.
To understand a trend, we therefore need to look at the method, the dates and the species included. Adding together categories from two different inventories can create the illusion of progress or collapse. The study also noted genuine declines linked to deforestation and overexploitation: recognising the effect of increased knowledge does not mean denying real losses.
What artificial intelligence can contribute
On a global scale, research presented by Kew Gardens in September 2022 combined existing assessments with machine-learning predictions. It produced risk information for 1,889 species, or around three quarters of the family. Of these, 56% could be threatened.
This result is not a new Red List fully verified in the field. Modelling helps identify priorities across a group that is still poorly documented. Kew noted that 620 species had to be excluded for lack of data and that predictions for a further 663 were based on little information. The tool guides research; it does not make the gaps disappear.
A reserve on a map is not enough
The study in Madagascar highlighted the value of expanding protected areas, while also showing that some species remained outside their boundaries and that threats persisted inside them. The legal protection of an area and the effectiveness of its conservation are two things that need to be assessed separately.
The issue then takes on a different shape: preserving places where populations can persist, monitoring their development and involving the communities that use these plants. A botanical collection can showcase this richness. To know whether it will survive in the wild, we must also look beyond the greenhouses, to the patches of forest where palms live and reproduce.
Illustrative photograph: palms and tropical vegetation on the island of Bohol, in the Philippines. Credit: Vyacheslav Argenberg / Wikimedia Commons — CC BY 4.0.
