Next Article in Journal
Landscape Aesthetics Quality in Subalpine Forests of Eastern Tibetan Plateau Will Greatly Decrease by the End of the Century?
Previous Article in Journal
Response of Eucalyptus Seedlings to Water Stress in a Warm Tropical Region in Brazil
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Forests
Volume 16
Issue 12
10.3390/f16121805
forests-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Predicting Optimal Sites for Ecosystem Restoration and Assisted Migration of Abies pinsapo Boiss. Using Species Distribution Modelling

by
Antonio Jesús Ariza-Salamanca
1,
Pablo González-Moreno
1,
José Benedicto López-Quintanilla
2 and
Rafael María Navarro-Cerrillo
1,3,*
1
Laboratory of Dendrochronology, Silviculture and Global Change, Dendrodat Lab-ERSAF, Department of Forest Engineering, University of Cordoba, Campus de Rabanales, Crta. IV, km. 396, 14071 Córdoba, Spain
2
Consejería Medio-Ambiente y Ordenación del Territorio, Plan de Recuperación del Pinsapo, 29071 Málaga, Spain
3
Instituto Interuniversitario del Sistema Tierra en Andalucía, Centro Andaluz de Medio Ambiente (IISTA-CEAMA), Avenida Mediterraneo, S/N, 18006 Granada, Spain
*
Author to whom correspondence should be addressed.
Forests 2025, 16(12), 1805; https://doi.org/10.3390/f16121805
Submission received: 3 November 2025 / Revised: 24 November 2025 / Accepted: 27 November 2025 / Published: 30 November 2025
(This article belongs to the Section Forest Ecology and Management)
Browse Figures
Versions Notes

Abstract

Climate change exacerbates the vulnerability of relict forests. However, plant taxa may buffer extinction risk through range shifts that track suitable habitats or through adjustments in their ecological niches, either via phenotypic plasticity or evolutionary adaptation to prevailing environmental regimes. In addition to these biological responses, the risks associated with climate change can also be mitigated through forest management practices and conservation strategies, including assisted migration. We used presence–absence data from Abies pinsapo Boiss. and environmental variables to describe the past and current natural distribution of the species by using species distribution models (SDMs). Then, we characterized future patterns of habitat suitability and identified potential areas for ecosystem restoration and assisted migration. The models predict a 77% loss of suitable habitat by 2060 and up to 99% by 2100 yet highlight climatically suitable areas outside the species’ current range—particularly in the Sierra Nevada National and Natural Park and Sierras de Cazorla, Segura y Las Villas Natural Park. These results provide spatially explicit guidance for restoration and assisted migration strategies. Our findings demonstrate the need for proactive conservation planning and show that SDMs can help identify climate refugia for long-term species persistence.

1. Introduction

Long-term shifts in the distribution of tree species are largely driven by climatic factors [1]. Species that have experienced progressive contractions of their original ranges in response to climate changes are identified as climate relicts [2]. In this context, Mediterranean firs (Abies species) constitute a relevant example of relict species with current limited ranges under the threat of climate change [3]. Their current distribution reflects adaptive processes associated with spatial shifts over time [4]. However, unlike past climatic changes, current global climate change is occurring over a very short timescale, potentially too brief to allow sufficient adaptation of species or successful migration. The mismatch in rates between climate change and tree migration velocity will strongly affect forest growth, composition, and genetic variability, particularly in relict species [5].
Consequently, proactive management actions have been proposed to ensure relict species conservation, promote their long-term persistence, and to maintain ecosystem services. A potential management approach is to support tree species and seed sources (populations) in following the displacement of their climatic niches through “assisted migration” [6]. Nevertheless, assisted migration is a complex strategy as it can conflict with traditional conservation objectives and paradigms while raising scientific, policy, and ethical concerns [7]. This approach requires human intervention to enable plant species to establish in areas expected to experience more suitable future climates [8]. Addressing this complexity necessitates comprehensive research and site-specific data to reliably predict and manage the outcomes of assisted migration interventions. To support this goal, species distribution models (SDMs) combined with global circulation models (GCMs) can be used to estimate habitat suitability for species under multiple possible climatic scenarios [9,10]. The most widely applied SDMs are correlative niche models, which statistically associate species occurrence records with environmental variables [11]. In this context, SDMs can be employed to characterize the ecological attributes of species, providing a basis for identifying candidates that are likely to be better adapted to future site conditions. This approach not only supports long-term replacement of species presumed to be less suited to changing environments but also provides a framework for assisted migration of threatened and endangered species to areas with higher probabilities of persistence under future climates [12].
Endemic to the southwestern Iberian Peninsula, Abies pinsapo Boiss. represents a fir species that has persisted as a climatic relict [13]. Because it occupies a highly specialized and restricted ecological niche, this species faces an elevated risk of extinction [14]. Consequently, it is listed as Endangered on the International Union for Conservation of Nature (IUCN) Red List of Threatened Species [15]. The pinsapo fir has traditionally been regarded as a conifer with low tolerance to water stress [16]. However, recent research has revealed an unexpectedly high resilience to climate-induced dieback and mortality [17]. Due to its physiological plasticity, the species is able to withstand areas exposed to seasonal summer drought [18,19]. However, several studies have reported an altitudinal shift in the distribution of pinsapo fir, attributed to the climatic constraints operating within its current realized niche [20]. This upward displacement suggests that the species is already responding to changes in temperature and water availability, which may further restrict its suitable habitat in the future. These concerns highlight the need to complement initiatives focused on extant populations with ex situ conservation measures, such as assisted migration. However, only a limited number of artificial stands exist outside the natural distribution range of pinsapo fir: one located in central Spain (Orcajo, Zaragoza province) and two established in the south (Granada province), within the protected natural areas of Sierra Nevada and Sierra de Huetor. The outstanding performance and regeneration of the pinsapo fir in these artificial plantations provides evidence of the feasibility of establishing additional sites outside its current range, where the species could survive and act as a refuge from climate change impacts.
This study addresses three critical research needs: (1) identifying the relative importance of climatic, edaphic, and topographic variables shaping the species’ current distribution; (2) predicting current and future suitable habitat distribution of Abies pinsapo under projected future climate changes, and (3) providing actionable spatial guidance for restoration and assisted migration efforts.
The findings of this study provide a decision-support framework for evaluating conservation strategies for Abies pinsapo, particularly those involving adaptive approaches such as assisted migration, and contribute to the development of evidence-based measures aimed at enhancing the species’ resilience under future climate scenarios.

2. Materials and Methods

2.1. Study Area

The study area covers the entire region of Andalusia (southern Spain), encompassing 87,600 km2 across eight provinces (Almería, Granada, Jaen, Malaga, Cordoba, Sevilla, Cadiz and Huelva). However, suitable habitats may be occupied or unusable for the establishment of the species. For example, artificial surfaces, water bodies, agricultural areas, flood zones, and productive forest systems (e.g., dehesas, timber plantations, etc.) can hardly be transferred to pinsapo fir forests. For this reason, in order to identify suitable areas for implementing ex situ conservation measures, it was necessary to constrain the projections obtained from SDMs by taking into account certain parameters, such as land cover and administrative boundaries. Thus, the suitable area was delineated exclusively within forest areas with legal protection status (Figure 1), thereby ensuring their long-term conservation and their efficacy as functional refuge habitats (e.g., National and Natural Parks, Biosphere Reserves or Natura 2000 European Ecological Network; ~28.364 km2). Land cover information was obtained from the Andalusian Regional Government database (https://portalrediam.cica.es/descargas/index.php/s/descargas?dir=/07_PATRIMONIO_NATURAL/01_ESPACIOS_PROTEGIDOS/01_RENPA (accessed on 8 March 2022)).

2.2. Methodological Framework

This article used several datasets and required the application of different statistical analysis. Therefore, a flowchart outlining the steps and relationships of each process is provided in Figure 2.

2.2.1. Occurrence Dataset

Our data on A. pinsapo occurrence is based on the Andalusian Regional Government database (https://portalrediam.cica.es/descargas/index.php/s/descargas?dir=/04_RECURSOS_NATURALES/01_BIODIVERSIDAD (accessed on 8 March 2022)). New occurrences of the species were identified through field surveys to ensure complete coverage of the species’ distribution in Andalusia (Figure 1). In total, we compiled 655 presence records from natural pinsapo fir forests located within the protected areas of Sierra de las Nieves, Sierra de Grazalema, and Los Reales–Sierra Bermeja (hereafter SN, SG, and SB; Figure 1). Additionally, 650 pseudoabsence points were randomly distributed across Andalusia to represent a broad spectrum of local environmental conditions, capturing variation in temperature, precipitation, soil properties, aspect, and topography.

2.2.2. Environmental Data Compilation and Pre-Processing

The selected predictor variables can be grouped into three categories of abiotic site characteristics—climate, soil, and topography—which jointly regulate water, energy, and nutrient availability, thereby influencing the growth and ultimately constraining the distribution of Abies pinsapo. These variables were selected for their biological relevance to the species, in consistency with previous studies [21,22,23,24]. We used several datasets to define the predictor variables. A description of each dataset is provided below, and a complete list and additional details of the variables are provided in Supplementary Material.
Climatic information was extracted from the SICMA platform (https://andalucia.sicma.red/ (accessed on 8 March 2022)). This platform is an e-GIS-WEB tool offers a map-based projection of 80 climate-related variables, allowing stakeholders to explore future scenarios related to climate, water balance, biodiversity, and thermal comfort [25]. We considered 53 climate variables at a resolution of 200 m2, averaged for the period of 1961–1990 (see variable description in Table S1 Supplementary Material). Outputs from the latest generation of climate simulations included in the Coupled Model Intercomparison Project Phase 6 (CMIP6, https://pcmdi.llnl.gov/CMIP6/, accessed on 26 November 2025) were used to evaluate projected future climatic conditions. The CMIP6 dataset provides climate projections from nine global climate models and across four Shared Socioeconomic Pathways (SSPs), offering a comprehensive basis for assessing potential future scenarios.
Soil data was downloaded from the Andalusian Regional Government database (https://portalrediam.cica.es/descargas/index.php/s/descargas?dir=/10_SISTEMAS_PRODUCTIVOS/03_RECURSOS_FORESTALES/Var_EdaficasBiomasa (accessed on 8 March 2022)). We considered 14 edaphic variables at a resolution of 100 m2, encompassing both physical and chemical soil parameters (Table S1 Supplementary Material) [26].
Topographic variables were calculated from the Digital Elevation Model (DEM) available from the national centre for geographic information (https://centrodedescargas.cnig.es/CentroDescargas/catalogo.do?Serie=MDT02 (accessed on 8 March 2022)). The DEM data were used to generate slope and aspect. We also included solar incidence variables available from the Andalusia Environmental Data Network at a resolution of 100 m2 (https://portalrediam.cica.es/descargas/index.php/s/descargas?dir=/10_SISTEMAS_PRODUCTIVOS/03_RECURSOS_FORESTALES/Var_OrograficasBiomasa (accessed on 8 March 2022)).
Considering the large number of variables included (n = 71; see Table 1), the Mean Decrease Gini (MDG) values was used to assess and rank the importance of the variables used [27]. This index is a default output of random forest that quantifies each variable’s contribution to node impurity reduction. This approach facilitates the selection of the most relevant predictors for modelling [27]. Higher MDG values indicate that the corresponding predictor variable has a greater impact on the classification process [28]. Then, a collinearity analysis was carried out to reduce uncertainty in the models’ predictions. For this, a variance inflation analysis (VIF) was applied [29]. All variables with a VIF value > 10 were eliminated [30]. In this way, the dimensionality was reduced, without significant loss of information.

2.2.3. Species Distribution Modelling

Ensemble model was fitted using BIOMOD (‘biomod2’ R package v3.5.1) [31]. We selected six statistical algorithms commonly used in SDM: generalized linear models (GLM) with linear and quadratic terms for each predictor [32], generalized boosting models (GBM) [33], generalized additive models (GAM) with a maximum of four degrees of freedom per variable [32], artificial neural networks (ANN) [34], random forests (RF) [35], and Maxent [36]. To assess the predictive performance of the models, we performed a tenfold cross-validation after dividing our dataset into training (70%) and evaluation (30%) subsets. Hyperparameters were optimized through internal biomod2 tuning, except for RF (500 trees) and GBM (1500 trees; learning rate 0.01).
The SDMs performance was assessed by the Area Under the Curve (AUC) criterion and True Skill Statistic (TSS) [37] in which the AUC values of the 10 iterations for each model were averaged [38]. Then, an ensemble (i.e., consensus) model was built using weighted averaging proportional to TSS scores.

2.2.4. Model Projections and Outputs

The ensemble model calibrated under current climatic conditions was projected onto future scenarios using the ACCESS-CM2 global climate model. To provide plausible and contrasting pathways for management and conservation planning, we selected two divergent SSP scenarios (SSP126 and SSP585) and four future periods (2031–2060, 2041–2070, 2051–2080 and 2071–2100). For all projections, we assumed that land-cover characteristics and administrative boundaries remained unchanged from present-day conditions (2025).
Continuous suitability outputs were subsequently transformed into discrete classes using threshold-based intervals. Values between 0 and 0.25 were classified as “low suitability”, 0.25–0.50 as “medium suitability”, 0.50–0.75 as “high suitability”, and values exceeding 0.75 as “optimum suitability”. This classification approach provided a standardized discretization of model results, ensuring reproducibility while facilitating a clearer interpretation of spatial suitability patterns and their relevance for conservation decision-making.

3. Results

3.1. Environmental Predictors and Performance of the SDMS

The variable selection procedure reduced the initial set of 71 predictors (Table 1) to five key environmental drivers that best describe the current habitat of Abies pinsapo Boiss. in Andalusia: precipitation sum from June to August (prc_summer; mm), autumn aridity index (ai_autumn; dimensionless), number of days in winter with minimum temperatures < 0 °C (nfd_winter; units), mean temperature from December to March (tmean_winter; °C) and altitude (dem; m.a.s.l.).
Climatic variables overwhelmingly dominated predictor importance, whereas soil and topographic variables—except altitude—played comparatively minor roles. Among the climatic factors, the highest scores were associated with the seasonal and annual aridity indices (ai_*) as well as precipitation (prc_*). This highlights the strong climatic dependency of the species and its sensitivity to water availability and winter thermal regimes.
The range of projected values for these selected variables (Figure 3) indicates substantial shifts under future climate scenarios, particularly in summer precipitation and seasonal aridity, which are expected to intensify under SSP585 scenario. These changes are consistent with the broader climatic trends forecasted for Mediterranean mountain systems.
A complete parametric characterization based on 71 environmental predictors for the realized niche of Abies pinsapo in SN, SG, and SB is provided in the Supplementary Material (Tables S2–S4).

3.2. Performance of the SDMs

All algorithms showed high predictive performance, particularly RF (random forest) and GBM (generalized boosted regression models), which consistently ranked highest across evaluation metrics (Table 2), reflecting their strong ability to capture nonlinear relationships between environmental gradients and species presence. In contrast, MaxEnt exhibited more variable performance, reinforcing the value of the ensemble approach for narrowly distributed species with strong climatic constraints.

3.3. Changes in Habitat Suitability

Ensemble projections revealed marked spatial and temporal shifts in habitat suitability (Figure 4). Within the current distribution areas (SN, SG, SB), suitability is projected to decline consistently across scenarios, reflecting increased aridity and warming. Losses are uneven among regions: SN retains comparatively more suitable conditions, while SG and SB are projected to undergo pronounced contraction of highly suitable areas. By late century, only SN will keep some areas of optimal suitability under the low-emission scenario (SSP126), whereas SG and SB are reduced to medium or low suitability classes.
Outside the current range, several protected mountain areas emerge as potential climatic refugia. In Sierra Nevada National and Natural parks, suitability increases during mid-century, with a transition from medium to high or optimal suitability classes, suggesting that this area may temporarily provide suitable conditions for the species. Although suitability eventually declines after 2070, some high-elevation areas continue to offer relatively stable future habitat. For other hand, Sierras de Cazorla, Segura y las Villas Natural Park exhibits a more limited but still notable increase in suitable habitat during mid-century, although these gains diminish by 2100, with no areas retaining high or optimal suitability (Figure 4).
Under the low-emission scenario (SSP126), contractions proceed progressively: an initial phase of moderate reduction is followed by pronounced declines after mid-century. By late century, habitat fragmentation becomes evident in all native areas except SN (Figure 5). In Sierra Nevada, areas initially classified as “medium suitability” are projected to shift to “high” and “optimum” suitability in the medium term (Figure 5d). From 2070 onwards, a general decline in suitability is observed; however, areas classified as “optimum suitability” are expected to be maintained in the long term (2100). In contrast, in the Sierras de Cazorla, Segura y Las Villas (Figure 5e), there is a slight increase in areas of “high suitability”, but this follows a marked decline relative to the reference climate period (1990–2020). Unlike Sierra Nevada, this region does not retain areas classified as “high” or “optimum” suitability in the long term.
The comparison of current and future suitability maps shows that ecologically meaningful differences arise not only from the magnitude of habitat change but also from the spatial configuration of suitable areas. Although statistical significance is intrinsic to the SDM outputs, the ecological significance stems from the persistence of high-elevation, mesic microclimates that buffer against regional aridification—conditions that are more common in the eastern Andalusian mountain ranges than in most parts of the species’ current distribution (with the exception of SN).

4. Discussion

Species distribution models (SDMs) have become indispensable tools for assessing the vulnerability of mountain coniferous species, which are often confined to narrow climatic envelopes and thus highly sensitive to the impacts of global change [39]. On this study, SDMs have been applied to assess the current and future distribution of Abies pinsapo. The model projections consistently indicate that the species will experience a profound reduction in climatically suitable habitat across its native range over the course of the twenty-first century, underscoring its high sensitivity to warming and increasing aridity (see Figure 4 and Figure 5). Even under the low-emissions scenario (SSP126), highly suitable areas contract substantially in SN and virtually disappear from SG and SB by late century. These declines are matched by only limited and spatially fragmented gains in suitability outside the current distribution. Overall, the projections portray a trajectory of severe long-term vulnerability, in which short-term stability masks the eventual collapse of optimum habitat conditions.

4.1. Current Habitat of Abies pinsapo and Environmental Drivers

Our results demonstrate that climatic variables, particularly summer precipitation, autumn aridity, and winter temperature extremes, are the dominant factors shaping the current and future distribution of this species. This is consistent with the known ecological requirements of Mediterranean firs, which depend heavily on cool and humid microclimates to withstand the strong summer drought typical of southern Iberia [22,40]. The inclusion of elevation as a significant predictor further highlights the role of topography in buffering against extreme climatic stress, by providing microrefugia where soil moisture and lower temperatures may mitigate broader warming and aridification trends.
The strong performance of most species distribution models (SDMs), particularly Random Forest (RF) and Generalized Boosted Models (GBM), underscores the robustness of our projections [41,42]. These algorithms captured the niche of A. pinsapo with high accuracy, reflecting the clear climatic constraints that define the limited distribution of A. pinsapo in its natural distribution. By contrast, the weaker performance of MaxEnt suggests limitations when modelling species with narrow ranges and strong dependence on a few environmental axes. Similar findings have been reported for other narrowly distributed conifers, where ensemble approaches typically outperform single algorithms [43]. Previous distribution models for A. pinsapo have also demonstrated the critical role of climatic factors, particularly precipitation and temperature seasonality, in shaping its distribution [40,44].
Our results improve upon these earlier approaches by incorporating multiple algorithms and rigorous variable selection, which reduce overprediction and yield projections that better match the observed ecological limits of the species. Similarly, studies on other Mediterranean firs, such as Abies alba and Abies marocana, have shown that machine-learning methods like RF and GBM provide more ecologically realistic predictions than MaxEnt alone, especially when dealing with highly restricted distributions [22,45].
Another important advance in our modelling approach is the ensemble framework, which integrates the strengths of different algorithms and reduces uncertainty associated with individual model biases [46]. Previous single-model applications for A. pinsapo have often struggled with transferability to future climate scenarios, particularly when relying solely on correlative methods. In contrast, the high AUC and TSS values obtained in this study suggest that ensemble projections are more reliable for informing long-term conservation planning. This methodological improvement is especially relevant for species with small, fragmented populations like A. pinsapo, where even minor modelling inaccuracies can lead to misleading conclusions about persistence or extinction risk [47]. In this context, our results highlight the importance of integrating multiple algorithms and carefully selected environmental predictors to improve simpler approaches, offering a more robust basis for forecasting climate-driven range shifts in Mediterranean mountain conifers. These findings reinforce the need to apply ensemble SDMs in conservation assessments of narrow-range species, where the stakes of inaccurate predictions are particularly high.
However, it is essential to acknowledge that the models describe only the abiotic, or Grinnellian, niche [48]. They do not incorporate biotic interactions such as competition, soil microbial mutualists, herbivores or pathogens. These interactions, which belong to the Eltonian niche, can critically influence establishment success and long-term persistence [48]. Consequently, while the models accurately identify areas with suitable climatic conditions, they cannot fully predict the realized ecological suitability of new sites, particularly when assessing the viability of assisted migration strategies.

4.2. Future Distribution of Abies pinsapo: Opportunities and Risks

Projected changes in climate are expected to substantially reduce the extent of suitable habitats for A. pinsapo [49]. The temporal pattern—initial stability or modest gains in suitable habitat followed by sharp declines—suggests that short-term resilience may mask long-term vulnerability. Such a trajectory has been observed for this species [23] and for other Mediterranean montane species [22,24], where early-century climatic variability can create transient windows of habitat expansion before the cumulative impacts of warming and aridification dominate. The projected decline underlines the vulnerability of A. pinsapo as a paleoendemic species restricted to small, fragmented populations [44]. Its dependence on high water availability, coupled with the increasing frequency of droughts and heatwaves in the Mediterranean Basin, makes it particularly sensitive to ongoing climate change [44,50]. This vulnerability is compounded by additional stressors not explicitly modelled here, such as wildfire risk, pathogen outbreaks, and anthropogenic pressures, which could accelerate declines beyond those projected by climate models alone [51].
At the same time, several high-elevation areas outside the current range, particularly in Sierra Nevada and Sierras de Cazorla, Segura y Las Villas, are projected to offer climates analogous to those historically occupied by A. pinsapo. These protected areas are characterized by lower temperatures, higher moisture availability, and microclimatic stability associated with complex mountain topography [20]. In both areas, suitability increases during mid-century, and some areas maintain favourable conditions into late-century projections, suggesting their potential role as future climate refugia if assisted migration programmes are implemented [52]. This highlights a potential role for active management, such as translocations or assisted gene flow, to maintain the species within climatically favourable habitats. However, the interpretation of these patterns must consider uncertainty related to model structure and the lack of biotic information [53].

4.3. Conservation Implications

From a conservation perspective, the findings of this study emphasize the urgency of proactive strategies [54,55]. The predicted collapse of suitable habitat within current distribution areas of A. pinsapo suggests that in situ conservation alone will not be sufficient. Integrating dynamic management approaches, such as identifying and safeguarding climate refugia, fostering ex situ conservation programmes, and considering assisted migration to climatically suitable areas, will be critical to ensure the persistence of A. pinsapo [56]. Moreover, maintaining genetic diversity across fragmented populations will be essential to enhance adaptive potential under rapid environmental change [57]. Existing populations of A. pinsapo are small, fragmented and potentially genetically differentiated [58]. Identifying appropriate source populations will require integrating genetic markers, provenance trials and climate-matching approaches [59].
For other hand, while high-elevation protected areas such as Sierra Nevada appear promising as future refugia, several limitations must be addressed before any intervention. Biotic interactions not considered in the modelling framework may strongly condition the success of translocated populations [24]. Moreover, logistical, legal and economic constraints also influence the feasibility of assisted migration [60]. Protected areas operate under conservation frameworks that may restrict the introduction of new taxa, even if they are native to the broader region. Establishing new A. pinsapo stands would require long-term resources for planting, irrigation during establishment, protection from herbivory and continuous monitoring, all of which demand coordination with park authorities and stable financial support.
Taken together, these considerations suggest that assisted migration should proceed cautiously and incrementally, beginning with small-scale experimental trials and accompanied by robust monitoring. Finally, adaptive responses, including potential phenotypic plasticity or local adaptation, were not considered but could alter species’ resilience.

5. Conclusions

This study shows that the distribution of Abies pinsapo is tightly constrained by climatic variables, with precipitation, aridity, winter cold, and altitude emerging as the strongest predictors of its niche. While our models reveal short-term stability or even potential expansion of suitable habitats into adjacent mountain ranges, long-term projections under climate change scenarios consistently indicate dramatic reductions in suitable areas, with near-total loss of optimum habitats by the end of the century. These findings highlight the vulnerability of A. pinsapo, a paleoendemic conifer of high conservation value, to the combined pressures of warming, aridification, and limited dispersal capacity. Conservation strategies should therefore move beyond static protection of current habitats and embrace dynamic approaches, including the identification of climatic refugia, assisted migration to newly suitable areas, and reinforcement of genetic diversity through ex situ measures. However, the feasibility of assisted migration deserves careful, evidence-based evaluation. This includes assessing genetic suitability of source populations, evaluating soil and biotic compatibility at potential recipient sites, and designing small-scale experimental trials supported by long-term monitoring. Therefore, future research should integrate genetic analyses, field-based transplant experiments, and models that explicitly incorporate biotic interactions to complement the climatic suitability framework presented here. Advancing these lines of inquiry will be essential for informing practical, forward-looking conservation actions capable of enhancing the long-term persistence of A. pinsapo under accelerating climate change.

Supplementary Materials

The following supporting information can be downloaded at https://www.mdpi.com/article/10.3390/f16121805/s1: Table S1: Description of the environmental variables selected for this study. Table S2: Habitat parametric characterization of Abies pinsapo Boiss. in Sierra de las Nieves National Park (SN). Minimum (min), 10th percentile (P10), 50th percentile (P50), 90th percentile (P90), maximum (max), mean ( x ¯ ), standard deviation (sd), and variation coefficient (cv). Table S3: Habitat parametric characterization of Abies pinsapo Boiss. in Sierra de Grazalema Natural Park (SG). Minimum (min), 10th percentile (P10), 50th percentile (P50), 90th percentile (P90), maximum (max), mean ( x ¯ ), standard deviation (sd), and variation coefficient (cv). Table S4: Habitat parametric characterization of Abies pinsapo Boiss. in Sierras Bermeja y Real Natural Site (SB). Minimum (min), 10th percentile (P10), 50th percentile (P50), 90th percentile (P90), maximum (max), mean ( x ¯ ), standard deviation (sd), and variation coefficient (cv).

Author Contributions

A.J.A.-S.: Conceptualization, Methodology, Formal analysis, Investigation, Data curation, Writing—original draft, Writing—review and editing, and Visualization. P.G.-M.: Investigation, Writing—original draft, Writing—review and editing, Visualization, and Supervision. J.B.L.-Q.: Conceptualization, Investigation, Writing—original draft, Writing—review and editing, Visualization, Supervision, Project administration, and Funding acquisition. R.M.N.-C.: Conceptualization, Formal analysis, Investigation, Resources, Writing—original draft, Writing—review and editing, Visualization, Supervision, Project administration, and Funding acquisition. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by Consejería de Sostenibilidad y Medio Ambiente (Junta de Andalucía) and the University of Cordoba.

Data Availability Statement

The original contributions presented in this study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

Acknowledgments

The authors would like to acknowledge the staff of the Consejería de Sostenibilidad y Medio Ambiente (Junta de Andalucía) for their contributions to the field data collection and for their technical and logistic support during the study, in particular Francisco Gómez, José Luís Sánchez Vallejo and Fernando Ríos Gálvez. Also, we knowledge the support of Juan José Guerrero Álvarez, Juan Carlos Cava Martínez, and Antonio Rivas Rangel (Agencia Andaluza del Medio Ambiente y Agua).

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Pearson, R.G. Climate Change and the Migration Capacity of Species. Trends Ecol. Evol. 2006, 21, 111–113. [Google Scholar] [CrossRef]
  2. Grandcolas, P.; Nattier, R.; Trewick, S. Relict Species: A Relict Concept? Trends Ecol. Evol. 2014, 29, 655–663. [Google Scholar] [CrossRef] [PubMed]
  3. Linares, J.C. Biogeography and Evolution of Abies (Pinaceae) in the Mediterranean Basin: The Roles of Long-Term Climatic Change and Glacial Refugia. J. Biogeogr. 2011, 38, 619–630. [Google Scholar] [CrossRef]
  4. Willis, K.J.; McElwain, J.C. The Evolution of Plants; Oxford University Press: Oxford, UK, 2014; pp. 678–679. [Google Scholar]
  5. Williams, M.I.; Dumroese, R.K. Preparing for Climate Change: Forestry and Assisted Migration. J. For. 2013, 111, 287–297. [Google Scholar] [CrossRef]
  6. Leech, S.M.; Almuedo, P.L.; O’Neill, G. Assisted Migration: Adapting Forest Management to a Changing Climate. J. Ecosyst. Manag. 2011, 12. Available online: https://jem-online.org/index.php/jem/article/view/91 (accessed on 22 November 2011). [CrossRef]
  7. McLachlan, J.S.; Hellmann, J.J.; Schwartz, M.W. A Framework for Debate of Assisted Migration in an Era of Climate Change. Conserv. Biol. 2007, 21, 297–302. [Google Scholar] [CrossRef]
  8. Lundgren, E.J.; Wallach, A.D.; Svenning, J.; Schlaepfer, M.A.; Andersson, A.L.A.; Ramp, D. Preventing Extinction in an Age of Species Migration and Planetary Change. Conserv. Biol. 2024, 38, e14270. [Google Scholar] [CrossRef]
  9. Rushton, S.P.; Ormerod, S.J.; Kerby, G. New Paradigms for Modelling Species Distributions? J. Appl. Ecol. 2004, 41, 193–200. [Google Scholar] [CrossRef]
  10. Elith, J.; Leathwick, J.R. Species Distribution Models: Ecological Explanation and Prediction Across Space and Time. Annu. Rev. Ecol. Evol. Syst. 2009, 40, 677–697. [Google Scholar] [CrossRef]
  11. Dormann, C.F.; Schymanski, S.J.; Cabral, J.; Chuine, I.; Graham, C.; Hartig, F.; Kearney, M.; Morin, X.; Römermann, C.; Schröder, B.; et al. Correlation and Process in Species Distribution Models: Bridging a Dichotomy. J. Biogeogr. 2012, 39, 2119–2131. [Google Scholar] [CrossRef]
  12. Falk, W.; Mellert, K.H. Species Distribution Models as a Tool for Forest Management Planning under Climate Change: Risk Evaluation of Abies Alba in Bavaria. J. Veg. Sci. 2011, 22, 621–634. [Google Scholar] [CrossRef]
  13. Ariza-Salamanca, A.J.; González-Moreno, P.; López-Quintanilla, J.B.; Navarro-Cerrillo, R.M. Large-Scale Mapping of Complex Forest Typologies Using Multispectral Imagery and Low-Density Airborne LiDAR: A Case Study in Pinsapo Fir Forests. Remote Sens. 2024, 16, 3182. [Google Scholar] [CrossRef]
  14. Alba-Sánchez, F.; López-Sáez, J.A.; Abel-Schaad, D.; Sabariego Ruiz, S.; Pérez-Díaz, S.; González-Hernández, A.; Linares, J.C. The Impact of Climate and Land-Use Changes on the Most Southerly Fir Forests (Abies pinsapo) in Europe. Holocene 2019, 29, 1176–1188. [Google Scholar] [CrossRef]
  15. Arista, A.; Alaoui, M.L.; Knees, S.; Gardner, M. Abies pinsapo. The IUCN Red List of Threatened Species. Available online: https://threatenedconifers.rbge.org.uk/conifers/abies-pinsapo (accessed on 26 November 2025).
  16. Linares, J.C.; Delgado-Huertas, A.; Carreira, J.A. Climatic Trends and Different Drought Adaptive Capacity and Vulnerability in a Mixed Abies pinsapo–Pinus halepensis Forest. Clim. Change 2011, 105, 67–90. [Google Scholar] [CrossRef]
  17. Cortés-Molino, Á.; Linares, J.C.; Viñegla, B.; Lechuga, V.; Salvo-Tierra, A.E.; Flores-Moya, A.; Fernández-Luque, I.; Carreira, J.A. Unexpected Resilience in Relict Abies pinsapo Boiss Forests to Dieback and Mortality Induced by Climate Change. Front. Plant Sci. 2022, 13, 991720. [Google Scholar] [CrossRef]
  18. Santini, F.; Shestakova, T.A.; Dashevskaya, S.; Notivol, E.; Voltas, J. Dendroecological and Genetic Insights for Future Management of an Old-Planted Forest of the Endangered Mediterranean Fir Abies pinsapo. Dendrochronologia 2020, 63, 125754. [Google Scholar] [CrossRef]
  19. Sancho-Knapik, D.; Peguero-Pina, J.J.; Flexas, J.; Herbette, S.; Cochard, H.; Niinemets, Ü.; Gil-Pelegrín, E. Coping with Low Light under High Atmospheric Dryness: Shade Acclimation in a Mediterranean Conifer (Abies pinsapo Boiss.). Tree Physiol. 2014, 34, 1321–1333. [Google Scholar] [CrossRef]
  20. González-Hernández, A.; Nieto-Lugilde, D.; Peñas, J.; Alba-Sánchez, F. Lean Pattern in an Altitude Range Shift of a Tree Species: Abies pinsapo Boiss. Forests 2021, 12, 1451. [Google Scholar] [CrossRef]
  21. Cerrillo, R.M.N.; Fernández, A.L.; Martinez, C.C.; Alvarez, J.G.; Salguero, R.S.; Oyonarte, P.B.; Quintanilla, J.L.; Cancio, A.F. Aproximación a la Definición del Hábitat Fisiográfico del Abies pinsapo Boiss. En Andalucía. For. Syst. 2006, 15, 137–152. [Google Scholar]
  22. López-Tirado, J.; Hidalgo, P.J. A High Resolution Predictive Model for Relict Trees in the Mediterranean-Mountain Forests (Pinus sylvestris L., P. nigra Arnold and Abies pinsapo Boiss.) from the South of Spain: A Reliable Management Tool for Reforestation. For. Ecol. Manag. 2014, 330, 105–114. [Google Scholar] [CrossRef]
  23. Navarro Cerrillo, R.M.; Duque-Lazo, J.; Rios-Gil, N.; Guerrero-Alvarez, J.J.; Lopez-Quintanilla, J.; Palacios-Rodriguez, G. Can Habitat Prediction Models Contribute to the Restoration and Conservation of the Threatened Tree Abies pinsapo Boiss. in Southern Spain? New For. 2021, 52, 89–112. [Google Scholar] [CrossRef]
  24. Blanco-Cano, L.; Navarro-Cerrillo, R.M.; González-Moreno, P. Biotic and Abiotic Effects Determining the Resilience of Conifer Mountain Forests: The Case Study of the Endangered Spanish Fir. For. Ecol. Manag. 2022, 520, 120356. [Google Scholar] [CrossRef]
  25. Chioncel, M. Transformative Innovation for Better Climate Change Adaptation—Case Study: Granada—Andalucia, Spain. Available online: https://publications.jrc.ec.europa.eu/repository/handle/JRC137324 (accessed on 19 March 2025).
  26. Álvarez, J.R.G.; Troncoso, J.V.; Rengel, A.S.; Almazán, M.L.S.; Álvarez, J.A.R.; Álvarez, J.J.G.; Rodríguez, J.J.L.; González, P.G.; Morales, J.I.; Marín, G. Biomasa Forestal en Andalucía. 1. Modelo de Existencias, Crecimiento y Producción. Coníferas. Available online: https://www.researchgate.net/publication/301748011_Biomasa_forestal_en_Andalucia_1_Modelos_de_existencias_crecimiento_y_produccion_Coniferas (accessed on 26 November 2025).
  27. Han, H.; Guo, X.; Yu, H. Variable Selection Using Mean Decrease Accuracy and Mean Decrease Gini Based on Random Forest. In Proceedings of the 2016 7th IEEE international Conference on Software Engineering and Service Science (ICSESS), Beijing, China, 26–28 August 2016; pp. 219–224. [Google Scholar]
  28. Hermosilla, T.; Bastyr, A.; Coops, N.C.; White, J.C.; Wulder, M.A. Mapping the Presence and Distribution of Tree Species in Canada’s Forested Ecosystems. Remote Sens. Environ. 2022, 282, 113276. [Google Scholar] [CrossRef]
  29. Shrestha, N. Detecting Multicollinearity in Regression Analysis. Am. J. Appl. Math. Stat. 2020, 8, 39–42. [Google Scholar] [CrossRef]
  30. Graham, M.H. Confronting Multicollinearity in Ecological Multiple Regression. Ecology 2003, 84, 2809–2815. [Google Scholar] [CrossRef]
  31. Thuiller, W.; Georges, D.; Gueguen, M.; Engler, R.; Breiner, F.; Lafourcade, B.; Patin, R.; Blancheteau, H. Package biomod2: Ensemble Platform for Species Distribution Modeling, Version R 4.2-6-2; R Foundation for Statistical Computing: Vienna, Austria, 2016; Available online: https://cran.r-project.org/web/packages/biomod2/index.html (accessed on 26 November 2025).
  32. Guisan, A.; Edwards, T.C., Jr.; Hastie, T. Generalized Linear and Generalized Additive Models in Studies of Species Distributions: Setting the Scene. Ecol. Model. 2002, 157, 89–100. [Google Scholar] [CrossRef]
  33. Ridgeway, G.; Edwards, D.; Kriegler, B.; Schroedl, S.; Southworth, H.; Greenwell, B.; Boehmke, B.; Cunningham, J.; GBM Developers. Gbm: Generalized Boosted Regression Models, Version 2.2.2; R Foundation for Statistical Computing: Vienna, Austria, 2019; Available online: https://cran.r-project.org/web/packages/gbm/gbm.pdf (accessed on 22 July 2025).
  34. Lek, S.; Guégan, J.-F. Artificial Neural Networks as a Tool in Ecological Modelling, an Introduction. Ecol. Model. 1999, 120, 65–73. [Google Scholar] [CrossRef]
  35. Breiman, L. Random Forests. Mach. Learn. 2001, 45, 5–32. [Google Scholar] [CrossRef]
  36. Phillips, S.J.; Dudík, M. Modeling of Species Distributions with Maxent: New Extensions and a Comprehensive Evaluation. Ecography 2008, 31, 161–175. [Google Scholar] [CrossRef]
  37. Stewart, S.B.; Fedrigo, M.; Kasel, S.; Roxburgh, S.H.; Choden, K.; Tenzin, K.; Allen, K.; Nitschke, C.R. Predicting Plant Species Distributions Using Climate-based Model Ensembles with Corresponding Measures of Congruence and Uncertainty. Divers. Distrib. 2022, 28, 1105–1122. [Google Scholar] [CrossRef]
  38. Jiménez-Valverde, A. Insights into the Area under the Receiver Operating Characteristic Curve (AUC) as a Discrimination Measure in Species Distribution Modelling. Glob. Ecol. Biogeogr. 2012, 21, 498–507. [Google Scholar] [CrossRef]
  39. Oke, O.A.; Thompson, K.A. Distribution Models for Mountain Plant Species: The Value of Elevation. Ecol. Model. 2015, 301, 72–77. [Google Scholar] [CrossRef]
  40. Fernández-Cancio, A.; NavarroCerrillo, R.M.; Fernández, R.F.; Hernández, P.G.; Meneéndez, E.M.; Martínez, C.C. Climate Classification of Abies pinsapo Boiss. Forests in Southern Spain. For. Syst. 2007, 16, 222–229. [Google Scholar] [CrossRef]
  41. Valavi, R.; Elith, J.; Lahoz-Monfort, J.J.; Guillera-Arroita, G. Modelling Species Presence-only Data with Random Forests. Ecography 2021, 44, 1731–1742. [Google Scholar] [CrossRef]
  42. Ahmadi, K.; Mahmoodi, S.; Pal, S.C.; Saha, A.; Chowdhuri, I.; Nguyen, T.T.; Jarvie, S.; Szostak, M.; Socha, J.; Thai, V.N. Improving Species Distribution Models for Dominant Trees in Climate Data-Poor Forests Using High-Resolution Remote Sensing. Ecol. Model. 2023, 475, 110190. [Google Scholar] [CrossRef]
  43. Lissovsky, A.A.; Dudov, S.V. Species-Distribution Modeling: Advantages and Limitations of Its Application. 2. MaxEnt. Biol. Bull. Rev. 2021, 11, 265–275. [Google Scholar] [CrossRef]
  44. Alba-Sánchez, F.; López-Sáez, J.A.; Pando, B.B.; Linares, J.C.; Nieto-Lugilde, D.; López-Merino, L. Past and Present Potential Distribution of the Iberian Abies Species: A Phytogeographic Approach Using Fossil Pollen Data and Species Distribution Models. Divers. Distrib. 2010, 16, 214–228. [Google Scholar] [CrossRef]
  45. Macias, M.; Andreu, L.; Bosch, O.; Camarero, J.J.; Gutiérrez, E. Increasing Aridity Is Enhancing Silver Fir Abies alba Mill.) Water Stress in Its South-Western Distribution Limit. Clim. Chang. 2006, 79, 289–313. [Google Scholar] [CrossRef]
  46. Araújo, M.B.; New, M. Ensemble Forecasting of Species Distributions. Trends Ecol. Evol. 2007, 22, 42–47. [Google Scholar] [CrossRef]
  47. Ramirez-Reyes, C.; Nazeri, M.; Street, G.; Jones-Farrand, D.T.; Vilella, F.J.; Evans, K.O. Embracing Ensemble Species Distribution Models to Inform At-Risk Species Status Assessments. J. Fish Wildl. Manag. 2021, 12, 98–111. [Google Scholar] [CrossRef]
  48. Soberón, J. Grinnellian and Eltonian Niches and Geographic Distributions of Species. Ecol. Lett. 2007, 10, 1115–1123. [Google Scholar] [CrossRef] [PubMed]
  49. Urban, M.C.; Bocedi, G.; Hendry, A.P.; Mihoub, J.-B.; Pe’er, G.; Singer, A.; Bridle, J.R.; Crozier, L.G.; De Meester, L.; Godsoe, W.; et al. Improving the Forecast for Biodiversity under Climate Change. Science 2016, 353, aad8466. [Google Scholar] [CrossRef] [PubMed]
  50. Sánchez-Salguero, R.; Ortíz, C.; Covelo, F.; Ochoa, V.; García-Ruíz, R.; Seco, J.I.; Carreira, J.A.; Merino, J.Á.; Linares, J.C. Regulation of Water Use in the Southernmost European Fir (Abies pinsapo Boiss.): Drought Avoidance Matters. Forests 2015, 6, 2241–2260. [Google Scholar] [CrossRef]
  51. Álvarez, J.R.G.; Catalina, M.A.; Cerrillo, R.M.N.; Quintanilla, J.L.; Sánchez-Salguero, R. Los Paisajes del Pinsapo a Través del Tiempo. In Proceedings of the Los Pinsapares en Andalucía (Abies pinsapo Boiss.): Conservación y Sostenibilidad en el siglo XXI, Córdoba, Spain, 14 June 2013; pp. 111–149. [Google Scholar]
  52. Vitt, P.; Havens, K.; Kramer, A.T.; Sollenberger, D.; Yates, E. Assisted Migration of Plants: Changes in Latitudes, Changes in Attitudes. Biol. Conserv. 2010, 143, 18–27. [Google Scholar] [CrossRef]
  53. Boiffin, J.; Badeau, V.; Bréda, N. Species Distribution Models May Misdirect Assisted Migration: Insights from the Introduction of Douglas-Fir to Europe. Ecol. Appl. 2017, 27, 446–457. [Google Scholar] [CrossRef]
  54. Resco De Dios, V.; Fischer, C.; Colinas, C. Climate Change Effects on Mediterranean Forests and Preventive Measures. New For. 2007, 33, 29–40. [Google Scholar] [CrossRef]
  55. Ruiz-Labourdette, D.; Schmitz, M.F.; Pineda, F.D. Changes in Tree Species Composition in Mediterranean Mountains under Climate Change: Indicators for Conservation Planning. Ecol. Indic. 2013, 24, 310–323. [Google Scholar] [CrossRef]
  56. Sáenz-Romero, C.; Mendoza-Maya, E.; Gómez-Pineda, E.; Blanco-García, A.; Endara-Agramont, A.R.; Lindig-Cisneros, R.; López-Upton, J.; Trejo-Ramírez, O.; Wehenkel, C.; Cibrián-Tovar, D.; et al. Recent Evidence of Mexican Temperate Forest Decline and the Need for Ex Situ Conservation, Assisted Migration, and Translocation of Species Ensembles as Adaptive Management to Face Projected Climatic Change Impacts in a Megadiverse Country. Can. J. For. Res. 2020, 50, 843–854. [Google Scholar] [CrossRef]
  57. Frascella, A.; Della Rocca, G.; Barberini, S.; Emiliani, G.; Secci, S.; Lambardi, M.; Benelli, C.; Tarraf, W.; Izgu, T.; Schicchi, R. Innovative in Situ and Ex Situ Conservation Strategies of the Madonie Fir Abies Nebrodensis. Sustainability 2022, 14, 12643. [Google Scholar] [CrossRef]
  58. Cobo-Simón, I.; Méndez-Cea, B.; Jump, A.S.; Seco, J.; Gallego, F.J.; Linares, J.C. Understanding Genetic Diversity of Relict Forests. Linking Long-Term Isolation Legacies and Current Habitat Fragmentation in Abies pinsapo Boiss. For. Ecol. Manag. 2020, 461, 117947. [Google Scholar] [CrossRef]
  59. Chen, Z.; Grossfurthner, L.; Loxterman, J.L.; Masingale, J.; Richardson, B.A.; Seaborn, T.; Smith, B.; Waits, L.P.; Narum, S.R. Applying Genomics in Assisted Migration under Climate Change: Framework, Empirical Applications, and Case Studies. Evol. Appl. 2022, 15, 3–21. [Google Scholar] [CrossRef]
  60. Hewitt, N.; Klenk, N.; Smith, A.L.; Bazely, D.R.; Yan, N.; Wood, S.; MacLellan, J.I.; Lipsig-Mumme, C.; Henriques, I. Taking Stock of the Assisted Migration Debate. Biol. Conserv. 2011, 144, 2560–2572. [Google Scholar] [CrossRef]
Forests 16 01805 g001
Figure 1. Location of the study area. Graphs were generated by QGIS 3.26.3 (https://www.qgis.org (accessed on 8 March 2022)) with vector data from the GADM database (https://gadm.org) and the Andalusian Regional Government database (https://portalrediam.cica.es/descargas/index.php/s/descargas?dir=/07_PATRIMONIO_NATURAL/01_ESPACIOS_PROTEGIDOS/01_RENPA (accessed on 8 March 2022)).
Figure 1. Location of the study area. Graphs were generated by QGIS 3.26.3 (https://www.qgis.org (accessed on 8 March 2022)) with vector data from the GADM database (https://gadm.org) and the Andalusian Regional Government database (https://portalrediam.cica.es/descargas/index.php/s/descargas?dir=/07_PATRIMONIO_NATURAL/01_ESPACIOS_PROTEGIDOS/01_RENPA (accessed on 8 March 2022)).
Forests 16 01805 g001
Forests 16 01805 g002
Figure 2. Flowchart of used methodology.
Figure 2. Flowchart of used methodology.
Forests 16 01805 g002
Forests 16 01805 g003
Figure 3. Boxplots of the selected environmental predictors for the realized niche of Abies pinsapo in SN, SG, and SB under two SSPs and four time periods (2031–2060, 2041–2070, 2051–2080, and 2071–2100) projected by the ACCES-CM2 global circulation model.
Figure 3. Boxplots of the selected environmental predictors for the realized niche of Abies pinsapo in SN, SG, and SB under two SSPs and four time periods (2031–2060, 2041–2070, 2051–2080, and 2071–2100) projected by the ACCES-CM2 global circulation model.
Forests 16 01805 g003
Forests 16 01805 g004
Figure 4. Relative habitat suitability for Abies pinsapo Boiss. under current conditions (top), and projected shifts in suitability by 2100 in Andalusia under the low-emissions scenario (SSP126). Red squares indicate the current distribution of Abies pinsapo in SN, SG, and SB. Graphs were generated by QGIS 3.26.3 (https://www.qgis.org) with the outputs of the suitability model.
Figure 4. Relative habitat suitability for Abies pinsapo Boiss. under current conditions (top), and projected shifts in suitability by 2100 in Andalusia under the low-emissions scenario (SSP126). Red squares indicate the current distribution of Abies pinsapo in SN, SG, and SB. Graphs were generated by QGIS 3.26.3 (https://www.qgis.org) with the outputs of the suitability model.
Forests 16 01805 g004
Forests 16 01805 g005
Figure 5. Changes in suitable area for Abies pinsapo Boiss. under the low-emission scenario (SSP126) across all time periods, within the realized niche of the species in SG, SB, and SN ((a–c), respectively), and in other protected areas beyond its natural range, i.e., Sierra Nevada and Sierras de Cazorla, Segura y las Villas ((d,e), respectively). Percentages represent the proportion of pixels in each class relative to the total number of pixels.
Figure 5. Changes in suitable area for Abies pinsapo Boiss. under the low-emission scenario (SSP126) across all time periods, within the realized niche of the species in SG, SB, and SN ((a–c), respectively), and in other protected areas beyond its natural range, i.e., Sierra Nevada and Sierras de Cazorla, Segura y las Villas ((d,e), respectively). Percentages represent the proportion of pixels in each class relative to the total number of pixels.
Forests 16 01805 g005
Table 1. Ranking of predictor variables based on Mean Decrease Gini (MDG) values. In italics, variables removed from VIF analysis. In bold, selected variables after the variable selection. For variable information, see Table S1 in the Supplementary Material.
Table 1. Ranking of predictor variables based on Mean Decrease Gini (MDG) values. In italics, variables removed from VIF analysis. In bold, selected variables after the variable selection. For variable information, see Table S1 in the Supplementary Material.
VariableMean Decrease Gini
prc_summer58.45wb_summer8.08tmean_autumn2.52
ai_autumn48.12tmax_spring7.91tmaxwm_annual2.26
ai_winter39.92tmin_autumn7.74fef2.18
ai_summer31.39wb_autumn6.79wd_autumn1.96
ai_annual29.28wb_spring6.71sand1.84
nfd_winter24.67tmin_annual6.18wd_summer1.79
dp23.42nfd_autumn6.01tmax_summer1.79
prc_autumn23.06sur_winter5.68tmean_annual1.74
wd_spring2.88silt4.89ttr_annual1.69
wi_annual22.00sur_autumn4.38slope1.53
ai_spring18.63tmean_spring3.70tmean_summer1.49
tmean_winter18.42trange_annual3.47wrc1.44
tmin_summer15.32act_lim3.46tmax_annual1.34
tmax_winter15.31sr_annual3.31tmeanwm_annual1.20
sur_spring14.06soilpH3.12curvature1.09
nfd_spring12.86cod_hid3.08nfd_summer1.02
prc_winter12.85tmax_autumn2.98sd0.48
dem11.06top_nitr2.89wd_winter0.03
wb_winter10.87som2.83sur_summer0.00
nfd_annual10.78tmin_winter2.81
wb_annual10.59clay2.75
prc_annual10.49wd_annual2.74
tmin_spring8.89om2.70
prc_spring8.51sur_annual2.69
tmeancm_annual8.47tmincm_annual2.55
cec8.32pbs2.52
Table 2. Model performance.
Table 2. Model performance.
ModelP1P2P3P4P5P6P7P8P9P10 x ¯
AUC
GLM0.9830.9880.9790.9620.9810.980.9780.9790.970.9740.98
GBM0.9920.9880.9910.9720.9890.9880.9880.9890.9870.9870.99
GAM0.9450.9530.9390.8980.9410.960.9350.9360.910.960.94
ANN0.9830.9780.9720.9590.9730.9730.970.9670.9590.970.97
RF0.9920.990.9940.9730.9880.990.9880.9880.990.9870.99
MaxEnt0.9370.9190.9120.9420.5410.970.9060.8790.8920.9310.89
TSS
GLM0.9130.9130.9080.8820.9080.9230.9080.9130.9030.8920.91
GBM0.9030.9180.9130.8720.9180.9340.9230.9030.9180.9030.91
GAM0.8320.8170.8270.7710.8170.8320.8220.8270.7760.8370.82
ANN0.8930.8670.8360.8210.8420.8620.8620.8470.8310.8360.85
RF0.9180.9390.9340.8930.9290.9490.9440.9180.9290.9130.93
MaxEnt0.8420.8120.7660.7760.810.8520.7710.7350.7350.8060.79
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.

Share and Cite

MDPI and ACS Style

Ariza-Salamanca, A.J.; González-Moreno, P.; López-Quintanilla, J.B.; Navarro-Cerrillo, R.M. Predicting Optimal Sites for Ecosystem Restoration and Assisted Migration of Abies pinsapo Boiss. Using Species Distribution Modelling. Forests 2025, 16, 1805. https://doi.org/10.3390/f16121805

AMA Style

Ariza-Salamanca AJ, González-Moreno P, López-Quintanilla JB, Navarro-Cerrillo RM. Predicting Optimal Sites for Ecosystem Restoration and Assisted Migration of Abies pinsapo Boiss. Using Species Distribution Modelling. Forests. 2025; 16(12):1805. https://doi.org/10.3390/f16121805

Chicago/Turabian Style

Ariza-Salamanca, Antonio Jesús, Pablo González-Moreno, José Benedicto López-Quintanilla, and Rafael María Navarro-Cerrillo. 2025. "Predicting Optimal Sites for Ecosystem Restoration and Assisted Migration of Abies pinsapo Boiss. Using Species Distribution Modelling" Forests 16, no. 12: 1805. https://doi.org/10.3390/f16121805

APA Style

Ariza-Salamanca, A. J., González-Moreno, P., López-Quintanilla, J. B., & Navarro-Cerrillo, R. M. (2025). Predicting Optimal Sites for Ecosystem Restoration and Assisted Migration of Abies pinsapo Boiss. Using Species Distribution Modelling. Forests, 16(12), 1805. https://doi.org/10.3390/f16121805

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop