1. Introduction
Tamarix boveana Bunge has strong root growth [1], with a strong adaptive ability to grow under arid and saline conditions [2] and is widely distributed in the Mediterranean region along riverbanks [3]. It has some economic value as a potential industrial raw material and source of animal feed because of its high protein and unsaturated fatty acid contents [4]. Tamarix boveana is used as an astringent, a diuretic, a stimulant of perspiration, and an aperitif in some folk medicines [5]. The essential oil of the species has antimicrobial properties; in addition, ethyl acetate and methanol in its extracts exhibit strong antioxidant activity [6] and are effective at killing larvae, thus acting as natural bioinsecticides [7]. Additionally, this plant plays an important role in ecological restoration and pollution management, and can eliminate nitrogen and phosphorus pollution [8].
Global warming has emerged as a critical threat to global ecosystems [9], especially in terms of species distribution, habitat quality, and biodiversity, with far-reaching impacts [10,11]. Environmental factors such as temperature, precipitation, and carbon dioxide concentration are key determinants of species growth and range, and changes in these environmental factors directly affect the habitat size and suitable distribution area of species [11,12,13]. In particular, shifts in temperature and rainfall patterns could differentially impact species [14], with significant differences in how different species respond to these changes [15]. Amid predicted atmospheric changes, the effects of global warming will continue to intensify [11], potentially leading to fragmentation of species’ habitats [16] and significant shifts in species’ ranges, which subsequently will affect ecosystem function and sustainability [10].
The International Union for Conservation of Nature Red List of Threatened Species serves as the world’s most extensive catalog documenting the extinction risk levels of flora and fauna worldwide and is also considered the most authoritative indicator of the status of biodiversity. The IUCN Red List criteria evaluate species extinction risk through quantitative thresholds A to E, with criterion B specifically addressing geographic range size, fragmentation patterns, and habitat decline. Tamarix boveana qualifies as Vulnerable (VU) under subcriterion B2ab (iii, iv, v) by meeting three key requirements: B2 exhibiting a restricted distribution under 20,000 km2, a demonstrating severe population fragmentation, and b showing sustained declines in iii habitat area and quality, iv mature individual counts, and v number of subpopulations. This classification confirms significant range disruption from combined climate change and habitat degradation impacts, indicating heightened wild population extinction risk [3]. The Mediterranean occupies an intermediate position separating central Europe’s humid temperate conditions from North Africa’s dry desert climate, experiencing influences from both systems. Owing to this particular geographic location, the Mediterranean exhibits heightened vulnerability to climate shifts [17]. This area has been recognized as a critical zone for projected climatic shifts, exhibiting heightened sensitivity to future environmental changes [18] and has been considered one of the hotspots of species diversity [19]. The native range of Tamarix boveana is located along the Mediterranean coast [3], and climate change in this region significantly modifies the species’ long-term sustainability and habitat occupancy. Thus, the Mediterranean seaboard serves as an ideal model system to assess how environmental changes might modify the organism’s ecological niche parameters and spatial occupation, yielding important data about its survival prerequisites and anticipated relocation tendencies that must direct safeguarding measures and prevent biological impoverishment.
Species distribution modeling (SDM) serves as a pivotal methodology to assess how shifting climatic patterns influence biogeographic ranges, utilizing species occurrence records coupled with environmental variables to project habitat suitability across diverse climate scenarios [20]. From 2000 to 2019, the application of species distribution modeling in ecology, forestry, and other fields has shown an increasing trend annually [21]. Species distribution modeling algorithms include the generalized linear model (GLM), generalized additive model (GAM), ecological niche factor analysis (ENFA), classification tree analysis (CTA), maximum moisture model (MaxEnt), and ensemble modeling, etc. Compared with other individual models, the MaxEnt modeling approach offers distinct benefits including not requiring complete data, handling small samples, and having high prediction accuracy [22]. Additionally, the maximum entropy approach demonstrates reduced sensitivity to sampling bias [23,24]. The prediction accuracy of ensemble models for future plant distribution ranges is not superior to that of individual modeling algorithms [25], a finding that also holds true for forecasting distribution areas of animals [26] and insects [27]. Furthermore, due to its limited data processing capacity, the ensemble modeling platform is less suitable for large-scale, high-resolution, and multi-scenario predictions, making the MaxEnt model a more appropriate choice under such circumstances [28]. Standard parameterization in maximum entropy approaches tends to produce excessively complex models that diminish predictive validity. This limitation has driven growing research interest in algorithmic refinements for improved ecological forecasting accuracy. The kuenm package provides an effective optimization scheme that can significantly improve the generalization ability of the model and reduce unnecessary complexity by automating the selection of the regularization multiplier and the feature type [29]. In this study, to enhance the precision of Tamarix boveana’s habitat suitability projections, we employed the kuenm package for MaxEnt model parameter optimization.
Previous studies on Tamarix boveana have mainly concentrated on investigating its chemical composition and stress resistance traits, while research on its suitable distribution areas remains limited. To address this knowledge gap, this study comprehensively considers multiple environmental variables, including climatic, edaphic and topographic factors, to evaluate both existing and potential distribution areas suitable for Tamarix boveana. Within the framework of the Coupled Model Intercomparison Project Phase 6 (CMIP6), we selected four representative Shared Socioeconomic Pathways for simulation. These scenarios cover a range of carbon and other heat-trapping gas-release trajectories: SSP126 represents a low emission pathway, SSP245 corresponds to a medium emission pathway, SSP370 reflects a medium-to-high-emission pathway, and SSP585 stands for a high-emission scenario. The selection of these diverse climate scenarios was designed to improve prediction accuracy of potential distribution shifts for Tamarix boveana under future environmental change conditions. The study objectives focused on (1) determining key bioclimatic drivers influencing Tamarix boveana’s geographic range; (2) modeling its spatial distribution under contemporary and projected climatic conditions; and (3) quantifying range dynamics through habitat area fluctuations and centroid displacements. This study provides theoretical guidance and scientific foundation for biodiversity preservation and ecosystem governance of Tamarix boveana.
2. Materials and Methods
2.1. Data Acquisition and Preparation
2.1.1. Species Distribution Data of Tamarix boveana
Occurrence records of Tamarix boveana were sourced from Global Biodiversity Information Facility (
2.1.2. Sources and Processing of Environmental Data
This study utilized 36 environmental variables, including climatic, topographic, and soil factors. Climate and elevation data were obtained from WorldClim. Using ArcGIS’s (version 10.4) terrain modeling tools, we generated both slope and aspect layers from the digital elevation model. Soil factors are also key drivers of plant distribution [31], particularly for plants adapted to arid and saline environments [32]. Soil factors were sourced from the Harmonized World Soil Database (HWSD,
Multicollinearity may lead to redundant factors and degradation of model performance. Therefore, we conducted Pearson correlation analysis on 36 ecological variables using the ENMTools tool [35]. For pairs of variables with strong intervariable correlations (r) > 0.8 [36], we retained the more ecologically relevant variable based on their contribution rate and permutation importance values [35]. Ultimately seven key predictors were selected for modeling: t_pH_h2o, bio1, bio3, bio15, bio18, bio19, and elevation. The complete set of ecological predictors is documented in Table S1, with corresponding correlation matrices presented in Figure 2.
2.2. Model Parameter Optimization
The MaxEnt modeling employs two fundamental parameters to regulate model complexity and feature selection: the regularization multiplier, denoted as RM, serves as a penalty coefficient that mitigates overfitting by constraining feature coefficient magnitudes, with a default value of 1 where increased values impose stronger restrictions; concurrently, Feature Classes abbreviated as FC represent mathematical transformations applied to environmental variables, specifically encompassing Linear through L, Quadratic via Q, Hinge using H, Product with P, and Threshold by T designations. These can be combined into 29 different types of feature combinations. This study utilized the kuenm package in R version 4.2.3 to establish a regularization multiplier range of 0 to 4 with an increment of 0.1. The RM and FC were cross-combined separately for validation, and there were a total of 1160 possible parameter settings. Finally, the combination with an omission rate < 5% and delta AICc = 0 was selected as the optimal parameter combination for inclusion in the model [29].
2.3. Model Construction and Evaluation
In the modeling process, we input the screened 7 environmental variables with 186 distribution points of Tamarix boveana into the MaxEnt model. The analysis configuration specified 10,000 environmental background samples and permitted up to 5000 algorithmic iterations. Model development utilized 75% of randomly selected observations, while the holdout 25% served for predictive performance evaluation. All these procedures were implemented within the MaxEnt modeling framework. To account for stochastic variability in MaxEnt and enhance result reliability, we conducted 10 replicate model runs across all time periods, selected logistic output format, and used the ensemble mean of all iterations as the final prediction. The predictive power of our model was quantified through AUC analysis derived from ROC curves. As a standardized evaluation metric spanning 0–1, increased AUC values correspond to enhanced model discrimination [37].
2.4. Tamarix boveana Habitat Classification, Patterns, and Changes in Centroid Migration
MaxEnt model outputs were processed in ArcGIS 10.4, where habitat suitability was reclassified into four distinct tiers using natural breaks classification: nonfitness zone (0–0.038), low-fitness zone (0.038–0.162), medium-fitness zone (0.162–0.371), and high-fitness zone (0.371–0.813).
The probability of the existence of Tamarix boveana ≥ 0.084 was considered the suitable zone, and the probability of the existence of Tamarix boveana ≤ 0.084 was defined as the unsuitable zone. Using ArcGIS’s SDMTools, we computed spatiotemporal variations in Tamarix boveana’s habitat extent and geographic centroids across climate scenarios and time periods. The area was derived by counting grid cells within the four suitability classes previously established. The centroid is a statistical indicator of spatial distribution that describes the geometric center of a species’ fitness zone, indicating the center of a species’ potential distribution. The centroid migration of the distribution range was determined by computing the geometric center coordinates, including latitude and longitude, and displacement distances across different time periods. These analyses enable the interpretation of potential distribution shifts for Tamarix boveana across varying climatic conditions.
3. Results
3.1. Model Accuracy Assessment
When modeling was based on these default parameters, a delta AICc value of 53.93 was obtained, indicating poor model performance. Following parameter optimization via the kuenm R package (version 4.4.1) simulations were conducted for the potential fitness zones of Tamarix boveana. The parameter settings at a delta AICc value of 0 were selected, i.e., a parameter combination of RM = 0.8 and FC = LQPT, which reduced the model complexity while maintaining a high predictive ability. With this parameter, the MaxEnt model was run and cross-validated 10 times, which indicated the mean AUC value of Tamarix boveana was 0.989 (Figure 3) with a standard deviation of 0.002. This outcome reflects robust estimation performance.
3.2. Main Environmental Variables Affecting the Distribution of Tamarix boveana
Jackknife analysis was employed to assess variable contributions, systematically evaluating each ecological parameter’s role in the predictive model. The results (Figure 4) demonstrated that isothermality (bio3), annual mean temperature (bio1), soil pH (t_pH_h2o), and precipitation of warmest quarter (bio18) exhibited relatively high regularized training gain when used in isolation (“with only variable”), indicating their independent explanatory power in predicting species distribution. In contrast, the other environmental factors showed lower independent contributions, suggesting their effects may depend on synergistic interactions with other variables. By integrating the Jackknife test results with the percent contribution analysis (Table S2), these four key environmental variables collectively accounted for 72.2% of the total contribution, confirming their dominant role in shaping the geographic range of Tamarix boveana.
3.3. Results of Tamarix boveana Habitat Prediction During the Current Period
Contemporary suitable habitats for Tamarix boveana are shown in Figure 5. Except for Croatia and Montenegro, Tamarix boveana is found in all other Mediterranean coastal countries, with a clear geographic preference for its distribution range. Specifically, it is concentrated in arid and semiarid areas along the southern Mediterranean shoreline and in coastal areas west of the Mediterranean Sea. The range has a clear geographic preference and is concentrated in the arid and semiarid areas of the southern Mediterranean coast and the coastal areas of the western Mediterranean. Optimal habitat conditions for the species are concentrated in the coastal areas of northern Algeria, northern Morocco, eastern Tunisia, southeastern Spain, central Greece, France, and southern Italy, with an area of 3.05 × 105 km2, accounting for 14% of the total suitability area. The moderately suitable areas are located in western and eastern Morocco, northwestern Algeria, Tunisia, northwestern Libya, central and northern Spain, and the Mediterranean coasts of other countries, with an area of 3.05 × 105 km2, and the Mediterranean coasts of other countries, with an area of 5.33 × 105 km2, accounting for 24% of the total suitable area. The low suitable area is located in west-central Spain, Algeria, and northern Libya, with an area of 14.02 × 105 km2, accounting for 62% of the total suitable area.
3.4. Modeling Results of Potential Tamarix boveana Habitat Areas in the Future
Figure 6 shows that under different future carbon emission scenarios, the distribution of Tamarix boveana remains mainly in southeastern Spain, northern Algeria, northern Morocco, and eastern Tunisia. Nevertheless, habitat suitability exhibits an overall declining pattern and further deteriorates with increasing emission intensity and time. In the 2030s (2021–2040s), the suitable areas under the low-emission pathway (SSP126) show the largest change from the current situation, with the highly suitable areas decreasing to 2.68 × 105 km2 (Chart 1). By the 2050s (2041–2060s), the suitable areas under the SSP245, SSP370, and SSP585 scenarios all show significant reductions in medium- and high-suitability zones, with the highly suitable zones decreasing to 2.15 × 105 km2 in the high-emission scenario (SSP585). By the 2070s (2061–2080s), the highly suitable zones for Tamarix boveana decrease further, especially in North Africa and Spain. In the southern part of the country, the highly suitable areas decline to 2.08 × 105 km2 under the SSP585 scenario, representing the lowest area of all the scenarios. This suggests that the distribution of the species will be further restricted under high-emission scenarios, with a significant reduction in the area of its core habitat. This may lead to a high degree of habitat fragmentation. Notably, under the low-emission scenario (SSP126), the low- and medium-suitability zones are relatively stable. The high-suitability zones decrease more moderately, and its distribution range is more stable. These findings suggest that lower levels of global warming have relatively manageable impacts on the Tamarix boveana habitat. Therefore, reducing carbon emissions may be an important strategy to mitigate the adverse effects of future climate change on the potential distribution of this species.
3.5. Changes in the Migration Patterns of Potentially Suitable Areas for Tamarix boveana over Time
Across all four emission scenarios, Tamarix boveana exhibited concurrent range gains and losses that intensified temporally, though habitat expansion rates remained consistently lower than contraction rates, resulting in net range reduction (Chart 2 and Figure 7). In the SSP126 scenario, the contracted areas were mainly located in southwestern Spain, Morocco, northern Algeria, and the Mediterranean coast of other countries. Specifically, the highest value for the contracting area was 2.67 × 105 km2 in the 2050s–2070s, and the highest value for the expanding area was also 1.31 × 105 km2 under present climatic conditions; range expansion primarily occurs in Spain’s north-central regions and Turkey’s central plateau. In the SSP245 scenario, in the 2050s–2070s, the contraction area is the largest at 3.41 × 105 km2, and the stabilization area decreases to 5.39 × 105 km2. The contraction area is primarily concentrated in the interior of North Africa, and the contraction area is more extensive than it is in the SSP126 scenario. The expansion area is more similar to that of the SSP126 scenario; however, more of the area is located in the Turkish region relative to the SSP126 scenario. Under the SSP370 scenario, the region of contraction increases significantly in the 2030s–2050s from 1.39 × 105 km2 to 3.52 × 105 km2 compared with the current period to the 2030s. The stabilization area decreases significantly from 7.41 × 105 km2 to 5.28 × 105 km2 compared with the current period to the 2030s, and it continues to decrease from the 2050s to the 2070s. Under the SSP585 scenario, the magnitude of the area change among the contracting, stabilizing, and expanding zones is more similar to that under the SSP370 scenario, with a greater increase in the area of the contracting zone from 2.07 × 105 km2 in the current period to 3.93 × 105 km2 in the 2030s. A significant decrease in the area of the stabilizing zone from 6.73 × 105 km2 in the current period to 4.88 × 105 km2 in the 2030s is observed. The contraction zone reaches the highest value of area for all scenarios at 4.78 × 105 km2 in the 2050s–2070s. The stabilization zone area reaches the lowest value at 4.02 × 105 km2.
3.6. Changes in the Centroid of Tamarix boveana over Time
The significant geographical differences in the distributions of suitable areas in Mediterranean coastal countries leads to interference with precise centroid calculation. When Mediterranean coastal countries are combined for centroid calculation, the position of the centroid may shift to the ocean because of the influence of the oceanic region. This finding is not in accordance with the actual distribution. To minimize marine interference and enhance centroid calculation precision, we conducted separate transfer analyses for the two primary suitable habitat zones within Mediterranean coastal nations. Specifically, Spain served as a separate region, and Morocco, Algeria, and Tunisia represent another region for centroid migration analysis. This helps to reveal more clearly the centroid transfer patterns of species suitability zones within different land areas (Figure 8). The longitude and latitude of the centroid and migration distances across different periods are presented in the Supplementary Materials section.
Centroid Shift in Spain (Supplementary Table S3A): Across all climate scenarios, the geographic center of Spain’s suitable habitat exhibits a general northeastward shift, with varying magnitudes across different time periods. In the SSP126 and SSP245 scenarios, the shift is relatively moderate, with the most substantial change occurring between the present and the 2040s. By comparison, under the SSP370 and SSP585 scenarios, the centroid shift becomes more pronounced in the century’s closing decades, with maximum displacement distances reaching 142.24 km (SSP370) and 97.00 km (SSP585), respectively.
Centroid Shift in Morocco, Algeria, and Tunisia (Supplementary Table S3B): in North Africa, the centroid of the suitable habitat for Tamarix boveana consistently shifts northeastward across all scenarios, although minor deviations are observed in certain time periods. Under the SSP126 and SSP245 scenarios, the shift remains relatively gradual, with single-stage displacement distances ranging from 7.16 km to 45.86 km. In contrast, under the SSP370 and SSP585 scenarios, the northeastward shift becomes more pronounced, with a cumulative centroid displacement of 49.82 km by the 2080s under the SSP370 scenario.
4. Discussion
4.1. Model Prediction Accuracy
The MaxEnt model’s discrimination ability is affected by species distribution records, environmental factors, and model parameter settings [38]. In this study, the effects of sampling errors and uneven sampling on the prediction results of the MaxEnt model were reduced by screening the distribution records of Tamarix boveana. The number of distribution points after screening was 186, and the model’s forecasting performance was significantly enhanced and stabilized when it exceeded 120 [39]. Preparatory tests were conducted before formal modeling, and seven climate variables were identified through aggregation of the predictor contribution score factors and replacing the contribution scores with the results of the Pearson correlation analysis. This prevented model overfitting resulting from multicollinearity among environmental predictors. The four carbon emission pathways in the BCC-CSM2-MR model in CMIP6 were selected for future environmental data to more clearly reflect the effects of different carbon emission intensities on the distribution range of Tamarix boveana. The model was optimized using the R kuenm data package, and appropriate feature combinations and modulation octaves were selected. The average AUC value of the predicted habitat extent of Tamarix boveana was predicted by the model to be 0.989, and the ROC curve tended to increase to the left. This indicated that the model prediction results were accurate and plausible and could be used to analyze the distribution range of Tamarix boveana.
4.2. Environmental Factors Affecting the Distribution of Tamarix boveana
Within the environmental gradients influencing the distribution of Tamarix boveana, isothermality (bio3) was the most critical determinant, followed by annual mean temperature (bio1), soil pH, and precipitation of the warmest quarter (bio18) (Figure 9). This demonstrates that precipitation, temperature, and soil factors collectively constrain the distribution of Tamarix boveana. Temperature and precipitation can directly influence the physiological activities of plants [40], shape different soil characteristics [41], and are critical factors for species growth [40,42]. Based on the environmental variable response curves, Tamarix boveana exhibited a peak isothermality value of 36, suggesting that this moderate seasonal temperature fluctuation may influence species distribution by maintaining essential phenological triggers [43]. The optimal annual mean temperature was 15 °C, consistent with long-term climatic data from Mediterranean regions [44]. The peak soil pH preference was 8.0, which likely affects the species’ adaptive range by modulating sodium ion uptake efficiency [45] and altering rhizosphere microbial community structure [46]. This indicates that Tamarix boveana prefers weakly alkaline soils, which aligns with its ecological distribution pattern in arid and saline-alkali regions [47]. Notably, the optimal pH range identified in this study differs from earlier reports on other Tamarix species, reflecting species-specific adaptations [48]. The peak precipitation of the warmest quarter was 50 mm. Under high-temperature conditions, precipitation directly determines ecosystem productivity and stability by regulating soil moisture, plant transpiration efficiency, and physiological stress. In Mediterranean coastal regions, warmest-quarter precipitation serves as a critical resource for vegetation to survive summer drought periods.
Owing to the limitations of the available environmental variables, this study projected the current soil pH and elevation factor to the future. Although soil pH can be affected by climate change and human activities, in general, it changes at a slow rate and not significantly in the short term. Thus, the stability of soil pH in the short- to medium-term time period of the 2021–2040s makes it highly predictable. However, in the long-term prediction of the 2041–2080s, the results have some bias. Changes in elevation factors in future periods may alter the elevation patterns currently occupied by species, and the results may also be biased. In addition to the above factors, the intensity, extent, and mode of human activities and species interrelationships may also change the land use pattern, thus affecting the spatial distribution of species to varying degrees [49]. Future studies could further investigate their impacts on Tamarix boveana distribution patterns by incorporating additional biotic and abiotic factors.
4.3. Distribution, Changes, and Analysis of Centroid Transfer in the Potential Fitness Zone of Tamarix boveana
Rising global temperatures are expected to intensify extreme meteorological phenomena [50] and profoundly alter the distribution patterns of plant communities [51], and the warming trend is projected to further intensify [52]. The range of Tamarix boveana remains mainly concentrated in Spain, Morocco, Algeria, and Tunisia under projected climate scenarios; however, the areas suitable for this species show a decreasing trend under different future emission scenarios, especially under the SSP585 high-carbon emission scenario. As an area sensitive to climate change, the Mediterranean region may experience higher temperatures and less precipitation in the future [17], especially in North Africa. This could be the primary driver behind the significant reduction in highly suitable habitat areas. Conversely, Spain’s optimal habitat zones demonstrate notable stability. This may be related to its complex topographic features (including mountains, hills, and plains) [53], which provide diverse habitats that buffer the detrimental impacts of climatic shifts. The range of Tamarix boveana is gradually expanding in the Turkish region as carbon emissions increase and over time. Turkey is geographically located at the border between the Mediterranean and continental climate zones, so it may have high regional variability in climate change impacts.
The migration trends of centroid under the low-emission scenarios (SSP126 and SSP245) were more similar between the two regions, with smaller center-of-mass shifts and phase adjustments. However, under the high-emission scenarios, the direction and distance of the center of mass migration fluctuated more in Spain, e.g., briefly to the southeast in the 2050s and then dramatically to the northeast in the 2070s under the SSP370 scenario. In contrast, North Africa maintained a more stable northeastward migratory trend but with a smaller overall migration amplitude. The observed range shifts in both direction and magnitude correlate strongly with elevational gradients, anthropogenic land modifications, and human disturbance [54]. Empirical evidence demonstrates that rising temperatures consistently push species distributions poleward [55,56]. This is consistent with the present study, where the centroid of Tamarix boveana shifted northward in both parts of Spain, Morocco, Algeria, and Tunisia and where the climatic conditions in these regions may become more suitable for the species in the future.
Although centroid shifts in suitable areas were relatively small in North Africa, the area of highly suitable areas decreased significantly. This phenomenon may be driven by a combination of factors. First, the overall centroid change in suitable areas primarily reflects the spatial center of gravity of the total suitable areas and does not represent changes in highly suitable areas alone. In North Africa, although the spatial location of the overall suitable area is relatively stable, future warming and increased drought may lead to the transformation of highly suitable areas to areas of medium and low suitability [57,58]. Taken together, contrasting with neighboring biogeographic regions, Spanish habitats show amplified reorganization under climate change pressures. with greater migration of the centroid. In contrast, the spatial location of suitable zones in the North African region is relatively stable, but the highly suitable zones are significantly reduced. This difference suggests that there are regional variations in the dynamics of suitable zones for Tamarix boveana across future climate scenarios [17]. Overall, the potentially suitable distribution areas of Tamarix boveana showed a gradual trend of fragmentation from the current period to the future period, which may be due to the increase in CO2 emissions, human activities, and other factors that exacerbate habitat fragmentation [59].
4.4. Tamarix boveana Conservation Status and Recommendations
Across contemporary and projected climatic conditions, the core habitats of Tamarix boveana show a predominant distribution along Mediterranean coastal regions. Existing nature reserves remain largely confined to inland areas, creating significant gaps in habitat protection coverage (Supplementary Figures S1 and S2). To address this critical conservation challenge, four priority interventions are proposed: First, strategic optimization of protected area networks should focus on key coastal zones exhibiting high habitat suitability, particularly along North African shores, southeastern Spain, and selected Italian coastal sectors. This spatial realignment must incorporate ecological corridor development to counter habitat fragmentation effects. Second, scientifically guided reintroduction programs should target environmentally analogous sites with stable habitat characteristics, with strict adherence to genetic diversity preservation protocols. Third, establishment of multilateral cooperation frameworks among Mediterranean coastal states is essential to enable systematic data exchange and coordinated research initiatives. Finally, implementation of proactive conservation measures including germplasm banking and controlled propagation programs will provide critical safeguards against future environmental variability. These integrated conservation strategies are designed to enhance species resilience to climatic changes while ensuring the sustained ecological functionality of Tamarix boveana populations across their native range. The proposed measures combine immediate habitat protection with long-term adaptive management approaches to address both current and emerging threats to this vulnerable species.
5. Conclusions
Through ensemble modeling with kuenm-enhanced MaxEnt, we quantified the present and potential future distributions of Tamarix boveana under four distinct climatic regimes (SSP126, SSP245, SSP370, SSP585), successfully addressing three key objectives: identifying current suitable habitats, determining environmental drivers, and projecting future range shifts. Currently, the species primarily occupies Mediterranean coastal regions, with Spain, Tunisia, Algeria, and Morocco harboring the most suitable areas, influenced mainly by isothermality, annual mean temperature, soil pH, and warmest-quarter precipitation. Future projections reveal a general habitat contraction, particularly in western Spain and inland North Africa, alongside a poleward shift in distribution at varying rates, increasing habitat fragmentation in high-suitability zones, underscoring climate change as a critical threat. Despite core habitats remaining in Spain, Morocco, Algeria, and Tunisia, existing protected areas cover only a minimal portion of its range, a persistent issue even as the species migrates northward, highlighting the urgent need for conservation strategies. However, this study has certain limitations, including uncertainties in climate projection data and insufficient consideration of anthropogenic disturbances and interspecies interactions. Future research should place greater emphasis on these factors.
Conceptualization and funding support, C.G. and C.Y.; data curation, methodology, data analysis, software, visualization, writing, S.D.; review, C.Y.; expert advice on data analysis and subject selection, H.W. All authors have read and agreed to the published version of the manuscript.
The original contributions presented in this study are included in the article and
The authors declare no conflicts of interest.
Footnotes
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Figure 1 Distribution point data of Tamarix boveana Bunge.
Figure 2 Spatial autocorrelation of environmental variables.
Figure 3 ROC of Tamarix boveana.
Figure 4 Jackknife test of environmental variables.
Figure 5 Potential suitable area of Tamarix boveana in countries across the Mediterranean under current climatic condition.
Figure 6 Suitable distribution changes of Tamarix boveana in countries across the Mediterranean under different climate change scenarios.
Figure 7 Changes of suitable area of Tamarix boveana in different periods.
Figure 8 Centroid changes of Tamarix boveana in different periods.
Figure 9 Response curves of the most influential variables.
Supplementary Materials
The following supporting information can be downloaded at
1. Arocena, J.M.; van Mourik, J.M.; Schilder, M.L.M.; Faz, C.A. Initial soil development under pioneer plant species in metal mine waste deposits. Restor. Ecol.; 2010; 18, pp. 244-252. [DOI: https://dx.doi.org/10.1111/j.1526-100X.2009.00582.x]
2. Terrones, A.; Morano, J.; Agullo, J.; Villar, J.L.; Vicente, A.; Alonso, M.A.; Juan, A. Influence of salinity and storage on germination of Tamarix taxa with contrasted ecological requirements. J. Arid. Environ.; 2016; 135, pp. 17-21. [DOI: https://dx.doi.org/10.1016/j.jaridenv.2016.08.001]
3. Beech, E. Tamarix boveana. IUCN Red List. Threat. Species; 2018; 2018, e.T79928148A123870302.
4. Ashori, A.; Nourbakhsh, A. Effect of press cycle time and resin content on physical and mechanical properties of particleboard panels made from the underutilized low-quality raw materials. Ind. Crops Prod.; 2008; 28, pp. 225-230. [DOI: https://dx.doi.org/10.1016/j.indcrop.2008.02.015]
5. Saïdana, D.; Mahjoub, M.A.; Boussaada, O.; Chria, J.; Chéraif, I.; Daami, M.; Mighri, Z.; Helal, A.N. Chemical composition and antimicrobial activity of volatile compounds of Tamarix boveana (Tamaricaceae). Microbiol. Res.; 2008; 163, pp. 445-455. [DOI: https://dx.doi.org/10.1016/j.micres.2006.07.009] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17223327]
6. Saidana, D.; Amel, B.; Boussaada, O. The antioxidant and free-radical scavenging activities of Tamarix boveana and Suaeda fruticosa fractions and related active compound. Eur. Sci. J.; 2014; 10, pp. 201-219.
7. Saidana, D.; Halima, M.K.B.; Haouas, D.H.; Tiba, B. Insecticidal activities of Tunisian halophytic plant extracts against larvae and adults of Tribolium confusum. Tropicultura; 2007; 25, pp. 193-199.
8. Kaddour, Z.; Khaled, M.T.O. Phytoremediation of Nutrient from Domestic Wastewater using Tamarix Boveana and Salsola Baryosma under salt stress. Asian J. Res. Chem.; 2021; 14, pp. 203-207. [DOI: https://dx.doi.org/10.52711/0974-4150.2021.00036]
9. Houtan, K.S.V.; Tanaka, K.R.; Gagné, T.O.; Becker, S. The geographic disparity of historical greenhouse emissions and projected climate change. Sci. Adv.; 2021; 7, eabe4342. [DOI: https://dx.doi.org/10.1126/sciadv.abe4342]
10. Duan, R.; Kong, X.; Huang, M.; Varela, S.; Ji, X. The potential effects of climate change on amphibian distribution, range fragmentation and turnover in China. PeerJ; 2016; 4, e2185. [DOI: https://dx.doi.org/10.7717/peerj.2185]
11. Terence, P.D.; Stephen, T.J.; Joanna, I.H.; Iain, C.P.; Georgina, M.M. Beyond predictions: Biodiversity conservation in a changing climate. Science; 2011; 332, pp. 53-58.
12. Nolan, C.; Overpeck, J.T.; Allen, J.R.M.; Anderson, P.M.; Betancourt, J.L.; Binney, H.A.; Brewer, S.; Bush, M.B.; Chase, B.M.; Cheddadi, R.
13. Liu, J.G.; Moody, H.; Hull, V.; Davis, S.J.; Gaskell, J.; Hertel, T.; Lubchenco, J.; Seto, K.C.; Gleick, P.; Kremen, C.
14. Parmesan, C. Ecological and evolutionary responses to recent climate change. Annu. Rev. Ecol. Evol. Syst.; 2006; 37, pp. 637-669. [DOI: https://dx.doi.org/10.1146/annurev.ecolsys.37.091305.110100]
15. Hanson, P.J.; Weltzin, J.F. Drought disturbance from climate change: Response of United States forests. Sci. Total Environ.; 2000; 262, pp. 205-220. [DOI: https://dx.doi.org/10.1016/S0048-9697(00)00523-4]
16. Ye, X.Z.; Zhang, M.Z.; La, W.F.; Yang, M.M.; Fan, H.H.; Zhang, G.F.; Chen, S.P.; Liu, B. Prediction of potential suitable distribution of Phoebe bournei based on MaxEnt optimization model. Acta Ecol. Sin.; 2021; 41, pp. 8135-8144.
17. Giorgi, F.; Lionello, P. Climate change projections for the Mediterranean region. Glob. Planet. Chang.; 2008; 63, pp. 90-104. [DOI: https://dx.doi.org/10.1016/j.gloplacha.2007.09.005]
18. Giorgi, F. Climate change Hot-spots. Geophys. Res. Lett.; 2006; 33, L08707. [DOI: https://dx.doi.org/10.1029/2006GL025734]
19. Myers, N.; Mittermeier, R.A.; Mittermeier, C.G.; Fonseca, G.A.B.; Kent, J. Biodiversity hotspots for conservation priorities. Nature; 2000; 403, pp. 853-858. [DOI: https://dx.doi.org/10.1038/35002501]
20. Elith, J.; Leathwick, J.R. Species distribution models: Ecological explanation and prediction across space and time. Annu. Rev. Ecol. Evol. Syst.; 2009; 40, pp. 677-697. [DOI: https://dx.doi.org/10.1146/annurev.ecolsys.110308.120159]
21. Guo, Y.L.; Zhao, Z.F.; Qiao, H.J.; Wang, R.; Wei, H.Y.; Wang, L.K.; Gu, W.; Li, X. Challenges and development trend of species distribution model. Adv. Earth Sci.; 2020; 35, pp. 1292-1305.
22. Philips, S.J.; Anderson, R.P.; Schapire, R.E. Maxinum entropy moedling of species geographic distributions. Ecol. Model.; 2006; 190, pp. 231-259. [DOI: https://dx.doi.org/10.1016/j.ecolmodel.2005.03.026]
23. Merow, C.; Smith, M.J.; Silander, J.A., Jr. A practical guide to MaxEnt for modeling species’ distributions: What it does, and why inputs and settings matter. Ecography; 2013; 36, pp. 1058-1069. [DOI: https://dx.doi.org/10.1111/j.1600-0587.2013.07872.x]
24. Phillips, S.J.; Anderson, R.P.; Dudík, M.; Schapire, R.E.; Blair, M.E. Opening the black box: An open-source release of MaxEnt. Ecography; 2017; 40, pp. 887-893. [DOI: https://dx.doi.org/10.1111/ecog.03049]
25. Crimmins, S.M.; Dobrowski, S.Z.; Mynsberge, A.R. Evaluating ensemble forecasts of plant species distributions under climate change. Ecol. Model.; 2013; 266, pp. 126-130. [DOI: https://dx.doi.org/10.1016/j.ecolmodel.2013.07.006]
26. Smith, A.B.; Santos, M.J.; Koo, M.S.; Rowe, K.M.C.; Rowe, K.C.; Patton, J.L.; Perrine, J.D.; Beissinger, S.R.; Moritz, C. Evaluation of species distribution models by resampling of sites surveyed a century ago by Joseph Grinnell. Ecography; 2013; 36, pp. 1017-1031. [DOI: https://dx.doi.org/10.1111/j.1600-0587.2013.00107.x]
27. Zhu, G.P.; Peterson, A.T. Do consensus models outperform individual models? Transferability evaluations of diverse modeling approaches for an invasive moth. Biol. Invasions; 2017; 19, pp. 2519-2532. [DOI: https://dx.doi.org/10.1007/s10530-017-1460-y]
28. Li, R.Z.; Hu, X.J.; Li, Q.Z.; Liu, L.Y.; He, Y.R.; Chen, C.Y. Gap analysis of Firmiana danxiaensis, a rare tree species endemic to southern China. Ecol. Indic.; 2024; 158, 111606. [DOI: https://dx.doi.org/10.1016/j.ecolind.2024.111606]
29. Cobos, M.E.; Peterson, A.T.; Barve, N.; Osorio-Olvera, L. kuenm: An R package for detailed development of ecological niche models using Maxent. PeerJ; 2019; 7, e6281. [DOI: https://dx.doi.org/10.7717/peerj.6281]
30. Gargano, M.L.; Mandracchia, G.; Venturella, G.; Calvo, R. A revision of Tamarix specimens (Tamaricaceae) kept in the BCN herbarium of Barcelona (Spain). Fl. Medit.; 2018; 28, pp. 393-397.
31. Wu, H.; Liu, Y.X.; Zhang, T.T.; Xu, M.X.; Rao, B.Q. Impacts of soil Properties on Species Diversity and Structure in Alternanthera philoxeroides-Invaded and Native Plant Communities. Plants; 2024; 13, 1196. [DOI: https://dx.doi.org/10.3390/plants13091196]
32. Ni, M.; Vellend, M. Soil effects on plant distributions and potential migration in Eastern North America. Res. Gate; 2022; 1, pp. 1-18.
33. Wu, T.W.; Lu, Y.X.; Fang, Y.J.; Li, L.; Li, W.P.; Jie, W.H.; Zhang, J.; Liu, Y.M.; Zhang, L.; Zhang, F.
34. Zhou, R.B.; Gao, Y.; Chang, N.; Ma, D.L.; Li, C.; Wu, H.X.; Wang, J.; Liu, Q.Y. Potential distribution of Triatoma rubrofasciata under different climatic scenarios in China. Chin. J. Biol. Control; 2022; 33, pp. 125-132.
35. Dormann, C.F.; Elith, J.; Bacher, S.; Buchmann, C.M.; Carl, G.; Gabriel, C.; Diekotter, T.; Jaime, R.G.; Gruber, B.; Lafourcade, B.
36. Warren, D.L.; Seifert, S.N. Ecological niche modeling in Maxent: The importance of model complexity and the performance of model selection criteria. Ecol. Appl.; 2011; 21, pp. 335-342. [DOI: https://dx.doi.org/10.1890/10-1171.1] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/21563566]
37. Wang, Y.S.; Xie, B.Y.; Wan, F.H.; Xiao, Q.M.; Dai, L.Y. Application of ROC curve analysis in evaluating the performance of alien species’ potential distribution models. Biodivers. Sci.; 2007; 15, pp. 365-372.
38. Guisan, A.; Thuiller, W. Predicting species distribution: Offering more than simple habitat models. Ecol. Lett.; 2005; 8, pp. 993-1009. [DOI: https://dx.doi.org/10.1111/j.1461-0248.2005.00792.x]
39. Chen, X.M.; Lei, Y.C.; Zhang, X.Q.; Jia, H.Y. Effects of sample sizes on accuracy and stability of maximum entropy model in predicting species distribution. Sci. Silvae Sin.; 2012; 48, pp. 53-59.
40. Liu, L.; Zheng, J.H.; Guan, J.Y.; Han, W.Q.; Liu, Y.J. Grassland cover dynamics and their relationship with climatic factors in China from 1982 to 2021. Sci. Total Environ.; 2023; 905, 167067. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2023.167067]
41. Davidson, E.A.; Janssens, I.A. Temperature sensitivity of soil carbon decomposition and feedbacks to climate change. Nature; 2006; 440, pp. 165-173. [DOI: https://dx.doi.org/10.1038/nature04514]
42. Guo, F.Y.; Liu, X.H.; Zhao, X.F.; Maimaiti, Z.; Wang, Y.G.; Lin, T.; Zhao, C.Y.; Xing, L.Y.; Wang, R.; Zhao, H.H.
43. Liu, Y.J.; Ge, Q.S.; Dai, J.H. Research progress in crop phenology under global climate change. Acta Geogr. Sin.; 2020; 75, pp. 14-24.
44. IPCC. Climate Change 2023: Synthesis Report. Contribution of Working Groups I, II and III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change; Core Writing Team Lee, H.; Romero, J. IPCC: Geneva, Switzerland, 2023.
45. Duan, Q.; Zhu, Z.; Wang, B.; Chen, M. Recent Progress on the Salt Tolerance Mechanisms and Application of Tamarisk. Int. J. Mol. Sci.; 2022; 23, 3325. [DOI: https://dx.doi.org/10.3390/ijms23063325]
46. Shen, X.T.; TU, B.H.; LIU, C.Y.; Zhao, Y.; Ding, J.X.; Liang, Y.T. Mechanisms of rhizosphere microorganisms in regulating plant root system architecture in acidic soils. Environ. Sci.; 2025; 46, pp. 570-578.
47. Francisco, G.M.; Fernando, M.T.; Esther, G.; Sergio, H.L. Salinity Tolerance of the Hygrophilous Plant Species in the Wetlands of the South of the Iberian Peninsula. Not. Bot. Horti Agrobot. Cluj-Napoca; 2012; 40, pp. 18-28.
48. Wu, T.G.; Wu, M.; Yu, M.K.; Xiao, J.H. Plant distribution in relation to soil conditions in hangzhou bay coastal wetlands, China. Pak. J. Bot.; 2011; 43, pp. 2331-2335.
49. Boivin, N.L.; Zeder, M.A.; Fuller, D.Q.; Crowther, A.; Larson, G.; Erlandson, J.M.; Denham, T.; Petraglia, M.D. Ecological consequences of human niche construction: Examining long-term anthropogenic shaping of global species distributions. Proc. Natl. Acad. Sci. USA; 2016; 113, pp. 6388-6396. [DOI: https://dx.doi.org/10.1073/pnas.1525200113]
50. Zhou, S.; Yu, B.F.; Zhang, Y. Global concurrent climate extremes exacerbated by anthropogenic climate change. Sci. Adv.; 2023; 9, eabo1638. [DOI: https://dx.doi.org/10.1126/sciadv.abo1638]
51. Cramer, W.; Guiot, J.; Fader, M.; Garrabou, J.; Gattuso, J.P.; Lglesias, A.; Lange, M.A.; Lionello, P.; Liasat, M.C.; Paz, S.
52. Filonchyk, M.; Peterson, M.P.; Zhang, L.F.; Hurynovich, V.; He, Y. Greenhouse gases emissions and global climate change: Examining the influence of CO2, CH4, and N2O. Sci. Total Environ.; 2024; 935, 173359. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2024.173359] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/38768722]
53. Hernaiz-Huerta, P.P. The geology of Sierra Calderona (SE Iberian Cordillera, Spain), a review. Results of a new 1:25.000 scale geological mapping. J. Maps; 2024; 20, 2302363. [DOI: https://dx.doi.org/10.1080/17445647.2024.2302363]
54. Ding, W.N.; Ree, R.H.; Apicer, R.A.; Xing, Y.W. Ancient orogenic and monsoon-driven assembly of the world’s richest temperate alpine flora. Science; 2020; 369, pp. 578-581. [DOI: https://dx.doi.org/10.1126/science.abb4484]
55. Wang, Y.J.; Zhao, R.X.; Zhou, X.Y.; Zhang, X.L.; Zhao, G.H.; Zhang, F.G. Prediction of potential distribution areas and priority protected areas of Agastache rugosa based on Maxent model and Marxan model. Front. Plant Sci.; 2023; 14, 1200796. [DOI: https://dx.doi.org/10.3389/fpls.2023.1200796]
56. Rodrigues, J.F.M.; Botero, C.A. The global determinants of climate niche breadth in birds. Nat. Commun.; 2025; 16, 3685. [DOI: https://dx.doi.org/10.1038/s41467-025-58815-1] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/40246840]
57. Tanarhte, M.; De Vries, A.J.; Zittis, G.; Chfadi, T. Severe droughts in North Africa: A review of drivers, impacts and management. Earth Sci. Rev.; 2024; 250, 104701. [DOI: https://dx.doi.org/10.1016/j.earscirev.2024.104701]
58. Bi, M.; Wan, L.; Zhang, Z.; Zhang, X.; Yu, C. Spatio-Temporal Variation Characteristics of North Africa’s Climate Potential Productivity. Land; 2023; 12, 1710. [DOI: https://dx.doi.org/10.3390/land12091710]
59. Nematollanis, S.; Fakheran, S.; Kienast, F.; Jafari, A. Application of invest habitat quality module in spatially vulnerability asseaament of natural habitats (case study: Chaharmahal and Bakhtiari province, Iran). Environ. Monit. Assess.; 2020; 192, 487. [DOI: https://dx.doi.org/10.1007/s10661-020-08460-6]
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Details
1 Northeast Asia Biodiversity Research Center, Northeast Forestry University, Harbin 150040, China; [email protected] (S.D.); [email protected] (H.W.), College of Forestry, Northeast Forestry University, Harbin 150040, China; [email protected]
2 College of Forestry, Northeast Forestry University, Harbin 150040, China; [email protected], State Key Laboratory of Tree Genetics and Breeding, Northeast Forestry University, Harbin 150040, China
3 Northeast Asia Biodiversity Research Center, Northeast Forestry University, Harbin 150040, China; [email protected] (S.D.); [email protected] (H.W.), College of Forestry, Northeast Forestry University, Harbin 150040, China; [email protected], State Key Laboratory of Tree Genetics and Breeding, Northeast Forestry University, Harbin 150040, China