Tecomella undulata under threat: The impact of climate change on the distribution of a valuable tree species using a machine learning model

PLOS One | https://doi.org/10.1371/journal.pone.0326609 July 9, 2025 13 / 18 such as input data quality and interpolation methods. Despite these uncertainties, gridded data are increasingly used in ecological [52,53] and agricultural research, as demonstrated by studies evaluating their utility in fields like crop modeling [54,55]. Threats facing T. undulata T. undulata, a versatile plant species significant for arid-rural communities, is currently under the threat of extinction due to rampant over-exploitation and inadequate conservation initiatives. Traditional propagation techniques relying on seeds are rendered ineffective as several obstacles such as human activities, environmental shifts, short seed viability, and improper harvest practices hinder the plant's natural regeneration process [20]. Compounding these issues is a lack of sustained conservation efforts from local populations. Concurrently, T. undulata’s population has been seeing a continual decrease within its natural habitats, primarily driven by numerous anthropogenic factors. Activities such as illegal deforestation, overharvesting, and habitat destruction significantly contribute to this alarming decline, emphasizing its fragile nature. Essentially, urgent and effective conservation measures are necessary to prevent further degradation of T. undulata’s pop- ulation, highlighting the requirement for an immediate strategic approach in its preservation [15,18]. Impact of climate change on T. undulata distribution

With temperature variables being the most influential predictors, any shifts in these conditions due to climate change could have pronounced effects on the species’ distribution. Temperature seasonality (bio4) emerged as the most signifi- cant variable, suggesting that T. undulata has a distinct phenological niche, tied closely to seasonal temperature variation. This specificity could imply a vulnerability to climate change, especially if temperature fluctuations become more erratic or deviate from the current seasonal patterns. The model's emphasis on temperature-related variables, including the mean temperature of the wettest (bio8) and the driest quarters (bio9), coupled with precipitation variables such as precipitation seasonality (bio15), underscores a potentially intricate environmental interplay that T. undulata depends on for its survival and reproduction. The species’ apparent preference for warmer temperatures during the wet quarters may hint at thermo- philic reproductive or germination strategies, which are at odds with the projected increase in unsuitable areas under all climate scenarios (SSP245 and SSP585). The findings indicate a pronounced decrease in areas of high suitability over the coming decades, with a 98% reduction by the year 2090 under the SSP585 scenario. Such drastic contractions could render T. undulata vulnerable to extinction, particularly because these regions represent the core of its climatic niche. On a regional level, the current distribution

centered in the warmer southern provinces of Iran like Fars and Bushehr will face significant challenges as the unsuitable areas are predicted to increase. This geographic concentration suggests a limited adaptive capacity unless the species can migrate to or establish in new areas, which in itself is uncertain given the overall reduction in suitable habitats. The consistent increase in unsuitable habitats and the noteworthy reduction in optimal ones suggest that by 2090, T. undulata may be confined to extremely small pockets of suitable environments, if at all. Phenology and environmental variables The flowers of this plant start to bloom from the beginning of April and this process continues until the end of May. The fruits of this plant were formed only in one region of the country and they existed until the end of May when the fruits started to fall as the weather got warmer. In some habitats, this species produces flowers twice a year. Once in early April and the second time in early November. The fall of leaves occurred in January and February (in Delfard of Jiroft habitat) and of Bushehr in July and August (in Shahnia and Moghdan habitat) [56]. Flowering time (April-May) coincides with the warmest and wettest period as preferred by the species (high Bio8). Moderate temperatures and precipitation during this time allow successful flowering. Fruiting in May requires maintaining temperature and humidity conditions. High