Severe drought raises projected habitat loss for most cacti, study finds

When 993 cactus species were modelled with future severe droughts as well as average climate, the share projected to lose more than four-fifths of their suitable climate rose from 1–15% to 25–35% across emissions scenarios, a projection of habitat, not a count of extinctions.

Peer reviewed · Modelling study (projections to the end of the century) · No species has been reassessed yet

Cacti are built for dry places, but a study published in Nature on 11 November 2026 suggests many species may not be built for the droughts a warmer world brings. Michiel Pillet, a postdoctoral researcher at the University of South Florida, Brian Enquist of the University of Arizona and 19 colleagues added projected severe drought to computer models of where 993 cactus species could grow. Far more species lost suitable climate than in models using averages alone. The results are model projections: no one has yet shown these species will die out.

Evidence at a glance

Is this a big deal?

A solid step for how forecasts are made, and a warning rather than a verdict for cacti. The strongest result is that drought-aware models fit the real distribution of cacti better than average-only ones, and that this choice changes the answer a lot. Whether a given species will actually vanish is a separate question this study cannot settle on its own.

What scientists found

Under average-climate models, a majority of cacti already lose ground: 58–61% of species are projected to lose some suitable climatic area, assuming they can spread up to 100 km. Add severe drought and that becomes 69–72%, or roughly 690 to 710 of the 993 species.

The bigger change is in how deep the losses go: the share projected to lose more than a quarter of its suitable area jumps from 26–43% to 55–58%. The share losing more than four-fifths rises from 1–15% to 25–35%, about 250 to 350 species. Nature‘s Research Briefing places that horizon at 2100.

If cacti cannot spread at all (a 0 km dispersal assumption, closer to reality for slow-growing plants over decades), 84–85% of species lose suitable area in the drought models. Losses also pile up in places: five bioregions could lose more than 25% of their species. The university releases name Chile, Peru and northeastern Brazil, all cactus diversity hotspots, as places where declines could be especially severe.

How they found it

The team started with occurrence records: places where a named cactus species has been reliably seen. They gathered 648,316 of them from the BIEN botanical database, private contributors and the community science network iNaturalist, then checked every record’s name and native status. Because poachers target rare cacti, the precise locations are available only on request.

A species distribution model learns which conditions a species lives in, then asks where those conditions will exist in future. The usual inputs are long-term averages such as mean temperature and yearly rainfall. Here the team also offered the models severe drought variables built from the Standardized Precipitation Evapotranspiration Index (SPEI), which compares water supply from rain with water demand from heat and dry air, judged against each place’s own history. The drought projections came from a global dataset published in 2025 by Diogo Araujo of Rutgers University with several of this study’s authors, and distributed through NASA’s Socioeconomic Data and Applications Center.

Think of judging a town’s water supply by its average year. The reservoir looks fine on paper, but it is the run of bad years that empties it. The analogy has a limit: cacti carry their own reservoir in their stems, so a single dry year matters less to them than to most plants. The study asks what happens when droughts grow longer and more intense than that store was built for.

Each species was modelled with and without the drought variables under three global climate models (GFDL-ESM4, MPI-ESM1-2-HR and UKESM1-0-LL) and three emissions pathways (SSP1-2.6, SSP3-7.0 and SSP5-8.5). The models were fitted with MaxEnt, widely used software for this kind of mapping. A statistical test that penalises extra complexity preferred the drought models in 91% of cases, and the authors report that the pattern held across climate models, cut-off choices and other sensitivity tests.

“When we started modeling and crunching the data, we were really skeptical of the initial findings,” Brian Maitner, an assistant professor of integrative biology at USF and co-author, said in the university’s release. “Then we consulted with cactus experts across North and South America and they confirmed that what we were modeling, they were already seeing in the field.”

Why it matters to you

If you live in the American Southwest or Mexico, this is about the plants on your doorstep. In the University of Arizona release, Enquist says severe drought is already visible in Arizona, “with large-scale mortality of chollas, prickly pears, and even the mighty saguaro.” Pillet adds that recent droughts have stunted saguaro growth and reproduction, which also affects the animals that rely on cacti for food and shelter.

For everyone else, the lesson is about forecasts. Climate-impact studies that use averages alone may be underestimating risk to wildlife, because averages smooth over the droughts and heat waves that do the damage. This paper is a test case. It changes nothing you need to do this year.

What the evidence does not show

It does not show that any species will go extinct by a given date. Suitable climatic area is where the climate matches what a species tolerates today; a plant can hang on in a spot the map calls unsuitable, or fail in one it calls fine. The releases say losses of 80% or more could qualify species as Critically Endangered under IUCN criteria and would more than double the number in that category. That is the authors’ reading of the projections. No Red List status has changed yet.

The models also leave things out. They do not represent water stored in plant tissue, soil moisture, how seedlings and adults differ in drought tolerance, interactions with pollinators and seed dispersers, or land-use change. SPEI measures drought relative to local history, not absolute dryness, a point one reviewer challenged. Results for rare species rest on few records, though the authors report the conclusions held when poorly sampled models were excluded.

How this fits what we already knew

The baseline was already grim. A 2015 global assessment in Nature Plants found 31% of 1,478 evaluated cactus species threatened, with land conversion and the collection of wild plants and seeds for the horticultural trade among the main threats. Those authors ranked cacti among the most threatened groups assessed; for comparison, a recent Nature study put 44% of assessed reef-building corals in threatened categories. In 2022, Pillet and colleagues modelled 408 species in Nature Plants and found that 60% would see favourable climate shrink in forecasts for 2050 and 2070, challenging the idea that cacti would thrive in a hotter world. The new study more than doubles the species count and adds the drought extremes that the earlier work left out.

What the peer reviewers said

SciQuest has not yet obtained independent comment; the clearest outside views are in the paper’s open peer review file. One anonymous referee called the methods disappointing, arguing that models of plant physiology, phenology and dispersal would do better and concluding that “there is undefined confidence in the study’s forecasts.” Karl Hülber and Johannes Hausharter of the University of Vienna, who reviewed the paper together, replied that process-based models need species data that does not exist for hundreds of cacti, and that “we regard the use of the MaxEnt-model as feasible”. The authors answered that the data needed for physiology-based models is not available at this scale and that their sensitivity tests support the results.

How other headlines framed it

• The Nature Podcast went furthest, saying extreme droughts could push iconic species to extinction. The paper models habitat, so “could” is doing a lot of work.
• Nature‘s Research Briefing led with the method, noting that biodiversity forecasts usually ignore extreme events, and put the deepest losses at more than 25% of species by 2100.
• Arizona’s Family made it local, centring saguaros and quoting Pillet’s personal view that some species will be lost in the wild.

The university releases say “more than a quarter” of species could lose at least 80% of their habitat. The abstract gives 25–35% across scenarios, so “about a quarter to a third” is the more precise reading.

What happens next

Pillet says the results will feed into updated IUCN Red List assessments for cacti. The researchers also call for work on how drought affects different life stages and whether other plant groups face the same blind spot in forecasts. The analysis code is public, so other teams can test the approach on different plants.

In one sentence

Cacti are drought specialists, but forecasts that include future severe droughts project that most of the 993 species studied lose suitable climate and a quarter to a third lose more than four-fifths of it, which is a warning about habitat, not yet proof of extinction.

Read next

• 44% of assessed reef-building corals are threatened, Nature study finds
• Most tracked bee species in Europe lost range over a century
• What happens if we reach net zero emissions?

Sources and disclosure

1. Pillet, M. D., Maitner, B. S., Arakaki, M., Araujo, D. S. A., Baker, M. A., Barrios, D. et al. Severe drought and warming increase the extinction risk of cacti. Nature (2026), published 11 November 2026. SciQuest read the abstract, figure and Extended Data legends, data availability statement, author information, Supplementary Table 3 and the open-access peer review file; the main text is behind a paywall.
2. Nature Research Briefing. Cacti are severely threatened by drought in a warming world, 11 November 2026.
3. University of Arizona. New research finds that even cacti are severely threatened in a warmer, drier world, 11 November 2026 (also on EurekAlert!).
4. University of South Florida. New USF-led research shows that even cacti are severely threatened in a warmer, drier world, 11 November 2026.
5. Pillet, M. et al. Elevated extinction risk of cacti under climate change. Nature Plants 8, 366–372 (2022).
6. Goettsch, B. et al. High proportion of cactus species threatened with extinction. Nature Plants 1, 15142 (2015).

The authors declare no competing interests. Funding came mainly from the US National Science Foundation, with Spanish and Chilean research grants, as listed in the paper.

Disclosure: This article was drafted with AI assistance and checked against the sources above. Species counts (for example, 25–35% of 993 is about 250 to 350) are SciQuest calculations from the abstract’s percentages; the graphic plots only the abstract’s figures. Spotted an error? See our corrections policy.