The research team, led by Hassan M. Alzain of the Yale School of the Environment, together with Hassan A. Alkhunaizi and Heba Bedair, set out to answer three questions that conservationists had never systematically addressed for this species: where could Ballota adenophora live today and in the future, how will climate change alter its habitat suitability, and which environmental variables control its distribution. The authors note that no global conservation assessment of the species exists under modern climate change scenarios, despite the plant’s endangered status and its restriction to one of the harshest environments on the planet.

To build their predictions, the researchers turned to species distribution modeling, or SDM, a technique that links known occurrence records with environmental data to map where a species can potentially survive. The team assembled 65 occurrence records from field surveys, published literature, unpublished reports, and the Global Biodiversity Information Facility. Because such a small sample can introduce serious bias into model predictions, the researchers generated ecologically realistic pseudo-absence points from areas where the species has not been reported, maintaining a balanced ratio of presences to pseudo-absences. They validated these points by comparing environmental conditions with true presence sites and ran sensitivity analyses to test how different selection strategies affected model performance.

Environmental data came from WorldClim 2.1 at a 2.5 arc-minute resolution, beginning with 19 standard bioclimatic variables. To avoid overfitting, the team screened the predictors for multicollinearity using the variance inflation factor, discarding any variable with a VIF above 5 or a correlation above 0.75. That filtering left three uncorrelated variables to drive the models: mean diurnal range, isothermality, and the mean temperature of the driest quarter. Future projections were based on the IPSL-CM6A-LR global circulation model, with ensemble averages applied to two contrasting Shared Socioeconomic Pathways for the far future period of 2061 to 2080. SSP126, the sustainability scenario, assumes low emissions and strong mitigation, while SSP585, the fossil-fueled development scenario, assumes high emissions with minimal climate action. By bracketing the range from optimistic to pessimistic futures, the study aimed to capture the full spectrum of possible outcomes.

The modeling itself used an ensemble approach, combining five algorithms within the sdm package in R: generalized linear models, random forest, support vector machines, radial basis function networks, and classification and regression trees. Ensembles are favored because they reduce the uncertainty and bias inherent in any single algorithm. Seventy percent of the data trained the models and thirty percent was reserved for testing, with accuracy evaluated using the area under the receiver-operating characteristic curve and the true skill statistic. The ensemble achieved a mean AUC of 0.75 and a TSS of 0.71, which the authors describe as exceptional accuracy for this kind of prediction. Random forest, support vector machines, and classification trees performed best among the individual algorithms.

The results paint a precise picture of where the plant can persist. High habitat suitability today concentrates in Madinah, Tabuk, the Hisma range, Jabal Al Lauz, and Jabal Radua north of Yanbu, with the northwestern and southwestern parts of the country emerging as the most suitable zones. Among the three climatic drivers, the mean temperature of the driest quarter dominated, explaining nearly 55 percent of the variance, while mean diurnal range and isothermality contributed the remainder. The response curves revealed a telling pattern: the probability of presence falls as the diurnal temperature range and the temperature of the driest quarter increase, but rises with greater isothermality, meaning the plant favors places where day-to-night temperature swings are proportionally large relative to seasonal swings.

Under both future scenarios, the story was similar and sobering. Potential distribution declined by more than four percent by 2061 to 2080, while newly gained suitable areas amounted to only about 0.9 percent. The climatically stable zones, where conditions should remain hospitable, cluster in western Saudi Arabia in the Abha, Mekkah, and Madinah regions and in the northwest near the Red Sea, with some gains appearing in the Uruq Bani Ma’arid protectorate. These refugia, the authors argue, are the places where conservation investment will pay the greatest dividends, because they could shelter the species as the surrounding landscape grows hotter and drier.

The emphasis on diurnal temperature variation is more than a statistical curiosity. Diurnal temperature range is known to regulate key plant functions, including stomatal opening, central carbon metabolism, and the photoperiodic timing of flowering and the transition from vegetative to reproductive growth. The authors link the species’ sensitivity to the extreme thermal environment it inhabits, where daytime heating is intense and nighttime cooling can be sharp. They also note that observed warming trends in southwestern Saudi Arabia, in regions such as Abha, Al-Baha, and Jazan, project temperature increases of roughly 0.8 to 2.1 degrees Celsius across the mid-century periods, compounding the stress on a plant already adapted to the edge of what a desert can offer. A decline in diurnal temperature range, driven by rising nighttime minimum temperatures, cloud cover changes, greenhouse gas emissions, and urbanization, may further disrupt these delicate physiological balances.

The study is candid about its limitations. Saudi Arabia lacks systematic surveys of its endemic plants, so most occurrence records come from herbaria, databases, and literature that skew toward easily accessible locations and conspicuous species. Some populations, and potentially even species new to science, may have been entirely overlooked. The authors acknowledge that the moderate accuracy of their models and the limited underlying data may introduce bias into the projections, and it remains unclear whether the species currently occupies all the areas flagged as suitable. They call for extensive ground-truthing surveys to validate the predictions and to uncover the full extent of Saudi Arabia’s endemic diversity, warning that vegetation loss in the kingdom may exceed that of any other Middle Eastern nation.

Even so, the conservation implications are clear and actionable. Saudi Arabia, a party to the Convention on Biological Diversity, currently protects roughly four percent of its land through fifteen protected areas that span the country’s major physiographic regions, wetlands, mountain and coastal ecosystems, and viable populations of threatened species. The study’s authors urge policymakers to declare the identified hotspots as protectorates, arguing that focusing on areas of high potential habitat suitability would protect the largest number of species per unit area. They further recommend building transboundary ecological corridors to connect fragmented habitats, which would facilitate gene flow and boost genetic diversity, alongside enforcement programs, habitat restoration, stakeholder engagement, and adaptive monitoring. In a country already celebrated for rescuing the Arabian oryx and other flagship species from the brink, the tiny yellow-flowered mint of the northwest shrublands may be next in line, provided the maps produced by these models are translated into protected ground before the climate closes in.

Subject of Research: Climate change impacts on the potential distribution of the endangered endemic plant Ballota adenophora in Saudi Arabia

Article Title: Potential distribution of the endangered Ballota adenophora, endemic to Saudi Arabia under climate change scenarios: toward conservation prioritization

Article References: Alzain, H. M., Alkhunaizi, H. A., & Bedair, H. (2025). Potential distribution of the endangered Ballota adenophora, endemic to Saudi Arabia under climate change scenarios: toward conservation prioritization. Discover Ecology, 1(1), Article 6. https://doi.org/10.1007/s44396-025-00007-8

Image Credits: AI Generated

DOI: 10.1007/s44396-025-00007-8

Keywords: Ballota adenophora, species distribution modeling, climate change, Saudi Arabia, endemic plants, conservation prioritization, ensemble modeling, habitat suitability, SSP scenarios, bioclimatic variables, protected areas, desert biodiversity