Next Article in Journal
Preliminary Study on Mechanism of Cold Stress on Dendroctonus valens Larvae
Previous Article in Journal
Molecular Crosstalk Underlying Pre-Colonization Signaling and Recognition in Ectomycorrhizal Symbiosis
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Forests
Volume 17
Issue 1
10.3390/f17010135
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

From Establishment to Expansion: Changing Drivers of Acacia spp. Invasion in Mainland Central Portugal

by
Matilde Salgueiro
1,*,
Carla Mora
1,2 and
César Capinha
1,2
1
Centre of Geographical Studies, Institute of Geography and Spatial Planning, University of Lisbon, Rua Branca Edmée Marques, 1600-276 Lisboa, Portugal
2
Associate Laboratory TERRA, Tapada da Ajuda, 1349-017 Lisbon, Portugal
*
Author to whom correspondence should be addressed.
Forests 2026, 17(1), 135; https://doi.org/10.3390/f17010135
Submission received: 28 November 2025 / Revised: 8 January 2026 / Accepted: 13 January 2026 / Published: 19 January 2026
(This article belongs to the Section Forest Ecology and Management)
Browse Figures
Versions Notes

Abstract

Land abandonment and recurrent wildfires are major drivers of landscape transformation in Mediterranean Europe, creating favorable conditions for the spread of non-native invasive woody species. Among these, Australian wattles (genus Acacia) are particularly widespread and problematic in Portugal. This work analyzed the spatiotemporal dynamics of Acacia spp. in two municipalities of central Portugal (Sertã and Pedrógão-Grande) by combining multitemporal photointerpretation of aerial imagery (2004–2021), generalized additive models (GAMs), and local perception surveys. Results reveal a 417% increase in occupied area over the last two decades. Modeling outcomes indicate a temporal shift in invasion drivers: from an establishment phase (2004–2010), mainly constrained by altitude and proximity to primary introduction sites, to a disturbance-driven expansion phase (2010–2021), influenced by fire recurrence, slope, and land-use context. Spatial clustering persisted throughout, underscoring the role of founder populations. Surveys confirmed high public awareness of Acacia invasiveness and identified abandonment and wildfire as the main perceived triggers of spread. By integrating ecological and social dimensions, this study provides a socioecological perspective on Acacia spp. expansion in Mediterranean rural landscapes and highlights the urgent need for integrated, landscape-scale management strategies.

1. Introduction

The Mediterranean rural landscapes are among the most dynamic socioecological systems in Europe [1,2]. Shaped by centuries of agroforestry use, they have been profoundly transformed in recent decades by rural depopulation, farmland abandonment, and recurrent wildfires [3,4,5,6]. These processes have weakened traditional land management and created favorable conditions for the spread of invasive woody species. In Portugal, Australian wattles (Acacia spp.) stand out as some of the most widespread [7] and ecologically disruptive invaders, forming dense patches that alter soil properties through their allelopathic capacity [8,9], inhibit native regeneration [10,11], and interact strongly with fire regimes [12,13,14].
The central interior of Portugal is particularly vulnerable to this phenomenon. Large-scale land abandonment, combined with severe and recurrent wildfires, has accelerated ecological succession and opened new ecological niches for invasive species. Initially confined to disturbed areas and riparian zones [15,16,17], Acacia spp. has rapidly expanded into abandoned terraces, shrublands, and post-fire areas [18,19,20], forming a persistent mosaic that challenges restoration and management efforts [21,22]. These invasions reflect not only the species’ biological adaptability, such as prolific seed banks [23,24] and post-fire regeneration capacity, but also the profound structural transformations in Mediterranean rural landscapes. These include the persistent fragmentation of agricultural land and the declining economic viability of smallholdings under European agricultural and inheritance frameworks, which together have accelerated land abandonment and reduced management cohesion [25,26,27,28].
Despite growing recognition of these trends, few quantitative studies have integrated both environmental and social dimensions of wattle invasion [29,30,31]. Most invasion studies have traditionally focused on species-specific traits or local ecological patterns, often overlooking how invasion drivers change over time as landscapes and ecological pressures evolve [32,33,34]. This knowledge gap is exacerbated by the limited understanding of how historical landscape changes influence current invasion processes and the lack of standardized methods for detecting and quantifying impacts. Adopting a socioecological perspective is therefore essential to disentangle how environmental gradients, disturbance regimes, and human-driven processes interact to shape invasion trajectories. This study addresses that gap by combining spatial modelling, multi-temporal mapping, and local perception data to examine Acacia spp. expansion in two municipalities of central Portugal: Sertã and Pedrógão-Grande. In line with invasion-stage frameworks, we distinguish an early establishment phase (2004–2010) from a subsequent expansion phase (2010–2021), based on the marked increase in newly mapped stands. Specifically, the study aims to: (i) map the spatial extent and expansions of Acacia stands from 2004 to 2021 through detailed photointerpretation of aerial imagery; (ii) identify the main environmental, land-use, and disturbance factors explaining species expansion using spatially informed Generalized Additive Models (GAMs); and (iii) explore local perceptions to contextualize ecological findings within a socioecological framework. We hypothesize that expansion is more likely in disturbed, low-management landscapes (e.g., shorter time since fire and land abandonment) and near dispersal pathways and propagule sources (e.g., roads, power lines, and existing stands), and that the relative importance of these drivers shifts through time, with early spread being more constrained by legacy introduction patterns, and later spread increasingly driven by disturbance and land-use context. By integrating ecological modelling with community-based insights, this research provides a comprehensive understanding of the mechanisms driving Acacia invasion in Mediterranean landscapes and supports the development of more effective, landscape-scale management strategies.

2. Materials and Methods

2.1. Study Area

The study was carried out in the municipalities of Sertã (446 km2) and Pedrógão- Grande (128 km2), located in the Pinhal Interior region of central mainland Portugal (Figure 1). According to the Köppen–Geiger climate classification [35] the area spans two closely related climatic types: Csb in Pedrógão-Grande (temperate with dry, mild summers) and Csa in Sertã (temperate with dry, hot summers). Despite this distinction, both municipalities share very similar conditions, with mean annual temperatures of approximately 22 °C.
Both municipalities are defined by a topography dominated by moderate slopes, interspersed with locally steeper areas, with elevations ranging ~120 m to 1080 m a.s.l., and the landscape is a heterogeneous mosaic composed mainly of forest plantations (Pinus pinaster, Eucalyptus globulus), semi-natural shrublands, and widespread patches of abandoned agricultural land.
The study area has undergone a pronounced socio-economic contraction since the mid-20th century. According to national census data, employment in the primary sector declined by more than 90% between 1960 and 2021, from approximately 7500 to 500 workers in Sertã, and from nearly 1800 to under 100 in Pedrógão-Grande (Table A1, Appendix A.1). This decline mirrors the long-term rural depopulation and structural transformation of Mediterranean agricultural systems through mechanization, modernization, and persistent land abandonment [21,36,37,38,39]. These processes have progressively reduced management intensity, promoted forest expansion, and increased the region’s ecological vulnerability to disturbance and biological invasion. In this context, Serta and Pedrógão-Grande provide a representative setting for examining the socioecological drivers of woody plant invasions in Mediterranean landscapes undergoing land-use transition.

2.2. Mapping of Acacia Stands

In the study area, invasions are dominated by Acacia dealbata, although other taxa (e.g., A. melanoxylon) also occur. Because species-level discrimination is not consistently reliable in orthophotos, stands were mapped at the genus level (Acacia spp.). To assess the distribution and expansion of Acacia stands, we obtained high-resolution orthophotos covering the study area for 2004, 2010, and 2021. These datasets were provided by the Portuguese Directorate-General for the Territory [40] and were available in natural-color (RGB) and false-color composites, with a spatial resolution of 50 cm and 25 cm, respectively.
All visible stands of Acacia spp. in the three periods, whether pure or mixed with other tree species, were manually digitized in QGIS v.3.36 (QGIS Association, Gossau, ZH, Switzerland), using both RGB and false-color composites to improve species discrimination. Polygons with uncertain identification were excluded to minimize false positives. The overall classification protocol is illustrated in Figure 2.
To evaluate the accuracy of the classification, ground-truthing campaigns were performed between February and March 2025. A total of 50 sites (Figure A1), distributed across the study area were visited to confirm the identity of the mapped polygons. Ground-truthing results indicated a high accuracy rate (98%), with only one misclassified site, indicating robust map reliability.

2.3. Predictors of Spatial Expansion

To identify drivers of expansion of Acacia stands, we considered a set of spatial variables characterizing land-use patterns and environmental conditions in the study area.
Land-use characterization was based on the Portuguese Land Use and Land Cover Cartography [41], using the 1995 and 2018 editions reclassified into Level 1 categories (Agriculture, Artificial surfaces, Shrubland, Pastures, and Forests). Due to the absence of intermediate land-use cartography, the 1995 and 2018 maps were used as the closest available representations for the 2004–2010 and 2010–2021 modelling periods. We therefore assumed that land-use classes remained relatively stable within each interval, acknowledging that this temporal disparity may introduce some uncertainty in the estimation of land-use effects, particularly for the first period. Additional predictor variables included altitude (m), slope (°), direct solar radiation (kWh/m2), distance to water bodies (m), distance to roads (m), distance to high-voltage power lines (m), and time since the last wildfire (years). Altitude and slope were derived from a digital elevation model, using digitized contour lines and spot heights extracted from the Portuguese Military Map (scale 1:25,000). The elevation data was interpolated to a final spatial resolution of 10 m. Direct solar radiation was estimated in SAGA GIS (version 9.8.1, SAGA GIS Development Team, Göttingen, Germany), using the Potential Incoming Solar Radiation module, with the spring equinox as the reference date. Spatial layers for roads, power lines, and water bodies were obtained from Open Street Map [42] and downloaded using the QuickOSM plugin (v. 2.4.1), in QGIS. These layers, originally in vector format, were converted to raster, reclassified into binary presence/absence variables, and used to calculate Euclidean distances in GIS. Fire history was derived from the national burned-area dataset from ICNF [43] and represented as time since last fire (years) and quantified at the pixel level using official wildfire perimeter data. For each raster cell, the year of the most recent recorded fire was identified, and the number of years elapsed since that event was calculated. This variable could only be included for the 2010–2021 period, as wildfire-year data are not available prior to 2009. These variables acted as proxies for anthropogenic disturbance and potential dispersal pathways. All spatial variables were standardized to the same coordinate system (ETRS89/Portugal TM06, EPSG:3763) and resampled to 10 m spatial resolution to ensure comparability.

2.4. Model Fitting and Analysis

To assess the expansion of Acacia, the mapped polygons were rasterized into a binary variable (presence = 1; absence = 0). Expansion was quantified by comparing the 2004 and 2010 maps (7-year interval) and the 2010 and 2021 maps (12-year interval), resulting in two periods of analysis: 2004–2010 and 2010–2021. We treat 2004–2010 as an establishment phase (early invasion from a limited baseline) and 2010–2021 as an expansion phase (spread from a larger invaded baseline), based on the marked increase in newly mapped stands and occupied area after 2010.
To contrast these conditions with those available across the study area, we also extracted predictor values from a random sample of cells distributed across the study area. For each period, 20,000 cells were randomly selected for both new Acacia stands and background (random) locations. This sample size was chosen to ensure broad coverage of environmental and spatial gradients across the study area, and it is consistent with common ecological modelling practice of using large background/pseudo-absence samples (often ≥10,000) in regression models (e.g., [44]).
Prior to model fitting, predictors were screened for multicollinearity using the Variance Inflation Factor (VIF). Variables with VIF > 10 were excluded [45,46], resulting in the exclusion of one predictor in the 2010–2021 period (Forests).
To assess the relationships between Acacia expansion and environmental and land-use predictors, we used a Generalized Additive Model (GAM) with a binomial error distribution and a logit link function. GAMs were selected because they provide a semi-parametric and flexible framework that can account for spatial autocorrelation in the data while maintaining statistical interpretability [47,48,49].
A two-dimensional smooth term based on the geographical coordinates of each sample point [50] was used to capture spatial autocorrelation and unmeasured spatial processes, while environmental and land-use predictors were included as linear parametric effects. This approach allows the model to separate large-scale spatial dependence from the specific influence of measured environmental variables.
Two independent GAMs were fitted to capture distinct invasion phases: (i) 2004–2010, and (ii) 2010–2021. This temporal partition allowed the identification of potential shifts in invasion drivers through time.
Model performance was evaluated through five-fold cross-validation. For each fold, the model was trained on 80% of the data and tested on the remaining 20%. Given that independently collected validation data were not available, and that the model is intended for application within the same study region and time period (i.e., not for transfer to other regions or future periods), five-fold cross-validation provides a standard and robust performance assessment. Predictive ability was quantified using the Area Under the Receiver Operating Characteristic Curve (AUC) [51,52]. All analyses were performed in R (version 4.4.1), using the mgcv package [53].

2.5. Survey to Local Population

We initially conducted a survey of the local population, aiming to gain contextual insights into perceptions of Acacia spp. invasions and their drivers. The survey, targeting residents in the study area, was developed in Google Forms and disseminated through social networks and local associations using a snowball sampling approach. It included both closed- and open-ended questions addressing residents’ knowledge of Acacia spp., perceptions of their invasiveness, land-use history, and management practices. Examples of survey questions included ‘Are you aware of various species of wattles in the region?’ and ‘What are your general opinions on the expansion of wattles in this region?’. The full version is available in Appendix A.3 (Table A2).
The survey was reviewed and approved by the IGOT’s Ethics Committee, and all procedures complied with institutional data privacy and ethical guidelines.
Responses were exported to a spreadsheet, with incomplete entries discarded. Quantitative questions were analyzed using descriptive statistics, while qualitative responses were grouped thematically to capture dominant narratives regarding the drivers and impacts of invasion.

3. Results

3.1. Mapping of Acacia spp. Stands

Results from orthophoto photointerpretation showed substantial expansion of Acacia spp. stands (Figure 3). A total of 79 patches (164 ha; 0.3% of the study area) were detected in 2004, increasing to 315 patches (360 ha; 0.6%) in 2010 and 779 patches (848 ha; 1.5%) in 2021. This corresponds to a cumulative increase of 417% in mapped area between 2004 and 2021 (Figure 4).

3.2. Predictive Modelling (GAM)

The results of the Generalized Additive Models (GAMs) revealed distinct patterns in the factors influencing Acacia spp. expansion between the establishment (2004–2010) and expansion (2010–2021) phases (Table 1).
During the establishment phase (2004–2010), Altitude was the only predictor with a significant negative effect (Estimate = −0.0192; p < 0.001), indicating a lower likelihood of expansion at higher elevations. Slope, Agricultural Land, Time Since Fire, Distance to Roads, Distance to Water, Direct Solar Radiation, and Distance to Power Lines did not show statistically significant effects (p > 0.05). However, the spatial smooth term was highly significant (edf = 17.5; p < 0.001), indicating substantial residual spatial structure in expansion probability after accounting for the measured environmental and land-use predictors. In other words, expansion probability remains spatially patterned beyond what can be explained by the included covariates, with some areas showing consistently higher or lower values than expected based on the measured variables alone.
In the expansion phase (2010–2021), a greater number of variables became significant. Altitude maintained its significant negative influence (Estimate = −0.0193; p < 0.001). Furthermore, Slope (Estimate = −0.0336; p = 0.0286), Agricultural Land (Estimate = −1.5710; p = 0.0175), and Time Since Fire (Estimate = −0. 0010; p = 0.0347) also exhibited significant negative relationships with expansion. This indicates that during this phase, expansion was favoured in lower-lying terrain, with gentler slopes, in non-agricultural areas, and in locations with a more recent fire history. The spatial structure remained a crucial component of the model, with the spatial smooth term again being highly significant (edf = 22.13; p < 0.001).
Regarding model performance, both models showed moderate explanatory power (R2 ≈ 0.39; 2004–2010: 0.398; 2010–2021: 0.389) and good discrimination (AUC > 0.88; 2004–2010: 0.885; 2010–2021: 0.884) across the two phases.
The spatial projection of the 2010–2021 model outputs illustrate these relationships across the landscape (Figure 5). The predicted probability of Acacia expansion shows a clear gradient, with the highest values concentrated in areas of moderate slope, in proximity to watercourses, and in zones that had experienced recent wildfires. Conversely, lower probabilities are consistently associated with higher elevations and less disturbed areas. A more detailed visualization of the predictions further shows how the modelled inhibitory effects of steep terrain and managed agricultural land, as reflected in the estimated parametric coefficients, manifest spatially (Figure 6).

3.3. Local Perceptions of Acacia Invasion and Expansion

A total of 45 individuals participated in the survey. Most participants (93%, n = 42) reported recognizing the most common species in the region. Among them, 88.9% were aware of its invasive behavior, although some respondents (n = 5) were unfamiliar with its ecological implications. Only 6.7% reported no prior knowledge of the species.
Regarding private land ownership, 65% of respondents confirmed the presence of Acacia spp. on their properties, with most of these lands having been abandoned for five or more years (65%). In terms of land-use types, 47% of respondents reported that invaded parcels were initially forest plantations (mainly Eucalyptus or Pinus pinaster), followed by permanent crops (24.4%), dryland/irrigated fields (18%), and other uses (11%).
Based on 24 (53%) participants, qualitative responses revealed three dominant narratives regarding perception of ongoing Acacia invasion in the region: (i) abandonment of agricultural and forest land as the main driver (38% of these respondents); (ii) perception of the phenomenon as being ‘natural’ and ‘inevitable’ (46%); and (iii) mixed causes, including wildfires, lack of institutional support, and climate change (17%).
Reported management actions were limited, with only 12 respondents indicating that they had attempted control measures. Most of these efforts were described as unsuccessful due to high costs, physical constraints, or limited effectiveness. Only two respondents reported successful outcomes, both using debarking techniques.

4. Discussion

This work provides a comprehensive assessment of the spatiotemporal dynamics and expansion drivers of Acacia spp. invasion in central Portugal, integrating high-resolution multitemporal mapping, ecological modelling, and perceptions of residents. Results revealed a marked expansion of the species between 2004 and 2021, with a cumulative increase of 417% in occupied area. The modelling results indicated a clear temporal shift in the factors driving expansion. During 2004–2010, Acacia spread was limited and primarily associated with altitude and spatial proximity to founder populations. In contrast, the 2010–2021 period corresponded to a disturbance-driven expansion phase, still affected by proximity to initial foci but increasingly influenced by fire recurrence, slope, and land-use patterns. Australian wattles exemplify this pattern, as their ecological and spatial impacts extend well beyond the canopy, exerting measurable functional effects on adjacent areas. Ref. [54] showed that such effects are strongest during early invasion stages and along patch edges, where secondary or satellite foci emerge from primary stands. These foci act as launch points for new nuclei, reinforcing the fragmented and nucleated spatial pattern typical of early invasion phases. Consistent spatial trends were also reported by [55,56], who found that invasion probability increases near existing populations and along roads or watercourses: areas of high propagule pressure and landscape connectivity that facilitate the formation of spatially connected invasion nuclei. Together, these findings support the idea that invasion foci act as persistent propagule sources sustaining local expansion even when environmental gradients remain stable.
Topography emerged as a persistent constraint shaping invasion pathways. During the early establishment phase, Acacia colonized predominantly low to mid-elevation zones, where climatic and edaphic conditions are more favorable and agricultural activity has historically concentrated. However, altitudinal variation in the study area is limited. Lower elevations typically correspond to valley bottoms characterized by higher soil moisture and proximity to watercourses. As a result, the altitudinal gradient does not represent a true climatic or hydric contrast; instead, establishment at lower elevations likely reflects local edaphic and hydrological conditions rather than altitude per se. Supporting this interpretation, ref. [57] highlighted the affinity of Acacia species for humid habitats, particularly riparian zones and valley bottoms, where soil moisture and nutrient availability enhance establishment success. Although proximity to roads and watercourses has often been identified as a key facilitator of Acacia spread [58,59,60], these variables showed no significant effect in our models. One possible interpretation is that this reflects the maturity of the invasion process: once Acacia populations are well established, further expansion may become less dependent on linear dispersal corridors and increasingly driven by local propagule pressure (e.g., seed rain, resprouting, and propagule accumulation around existing stands). Under this scenario, accessibility may play a secondary role relative to disturbance and land abandonment. However, these non-significant results may also arise from model-related factors, such as how corridor influence is represented at the analysis resolution, or because part of the relevant variation is expressed through the model’s two-dimensional spatial smooth, making independent effects harder to detect. Accessibility plays a secondary role compared with disturbance and land abandonment. Over time, recurrent fires and continued land abandonment reduced these constraints, enabling upward expansion into more rugged terrain. Slope exhibited a negative effect, limiting establishment on steep areas but favoring colonization on gentler slopes that retain moisture and organic matter. These findings indicate that topography operates not as a fixed barrier but as a dynamic gradient interacting with disturbance regimes to shape invasion trajectories across Mediterranean mountain landscapes.
The increasing influence of fire recurrence and slope in the later period confirms a transition toward disturbance-mediated invasion. The positive association with recently burned areas highlights the synergy between Acacia’s regenerative capacity and fire occurrence, which is well-documented for this genus [7,61]. Interestingly, older burned areas are not associated with similar levels of expansion, suggesting that post-fire recruitment is strongest immediately after disturbance and declines as native vegetation regenerates or management resumes. This pattern aligns with studies showing that Acacia spp. benefit from the window of opportunity created after fire-reduced competition, open soil, and increased light, while their competitive advantage diminishes as ecosystems recover [62].
Agricultural areas were negatively associated with the establishment of new Acacia stands. This pattern is also evident in the predictive surface which, when overlaid with agricultural patches, shows that low-suitability zones largely coincide with cultivated land. This suggests that persistent agricultural management is a strong barrier to invasion. This interpretation is consistent with [20], who analyzed the transboundary Gerês–Xurés Biosphere Reserve (NW Iberian Peninsula) and found that actively managed agricultural areas significantly limit Acacia expansion by reducing suitable habitat availability. Their analyses demonstrated that sustainable farming systems and High Nature Value farmlands act as barriers to invasion, whereas fire-exclusion or agricultural abandonment enhances seed production and establishment opportunities.
The patterns observed illustrate two interconnected feedbacks: ecological feedback, in which recurrent fires enhance Acacia regeneration and fuel accumulation, thereby increasing future fire risk; and a socioecological feedback, in which land abandonment facilitates invasion while the resulting spread further discourages agricultural recovery (Figure A2). Together, these reinforcing loops perpetuate invasion pressure and landscape degradation, strengthening the coupling between ecological disturbance and socio-economic decline in Mediterranean rural systems.
The ecological transformation observed is tightly linked to the socio-economic trajectory of the region. Between 1960 and 2021, employment in the primary sector declined by more than 90% in both municipalities, and ageing indices now approach 400 in Pedrógão-Grande and 300 in Sertã. This long-term depopulation and decline in land management capacity fostered widespread land abandonment, facilitating the establishment of pioneer woody species such as Acacia spp. and enhancing their post-fire regeneration. Local surveys, although based on a moderate sample size and potentially subject to some non-independence due to snowball recruitment, corroborated this interpretation. Residents most frequently identified wildfire and land abandonment as the primary drivers of invasion, consistent with the key predictors highlighted by the models. While the survey is unlikely to be representative of the full local population, the convergence between ecological and social evidence reinforces the view that invasion dynamics in Mediterranean rural regions are inherently socioecological, shaped by interactions between biophysical feedbacks and human-driven land-use change. This convergence of land abandonment, demographic decline, and recurrent wildfires has created favourable conditions for Acacia spp. expansion in central Portugal, reinforcing feedbacks between socio-economic change, disturbance regimes, and invasive species dynamics, and highlighting the role of long-term structural drivers beyond local environmental controls
The persistence of a strong spatial signal in both modelling periods, captured by the significant spatial terms, reinforces that Acacia expansion is largely spatially clustered. Distinct invasion hotspots coincide with zones of recurrent disturbance and favorable site conditions, acting as continuous propagule sources. This spatial autocorrelation demonstrates that invasion is not driven solely by environmental suitability but also by self-reinforcing spatial feedbacks, whereby established populations promote nearby colonization through short-distance dispersal. Over time, these scattered foci tend to merge into extensive, homogeneous stands: a hallmark of advanced invasion phases in woody Fabaceae [63].
The evidence that recurrent wildfires accelerate Acacia spread calls for a re-evaluation of current fire-management strategies, which remain largely reactive and disconnected from invasion dynamics. Integrating fuel management, post-fire restoration, and targeted invasive-control actions could help disrupt the fire–invasion feedback loop, provided that fuel breaks and high-risk areas are consistently maintained [64,65,66]. However, ecological interventions alone are insufficient, as the persistence of Acacia nuclei can ultimately reflect broader socio-economic drivers. While active agriculture contributes to stabilizing land-use mosaics, the region’s geological and edaphic constraints limit the feasibility of restoring traditional farming systems. The schist–greywacke soils typical of central Portugal are inherently poor and erosion-prone, which reduces the long-term productivity of abandoned land, even under CAP incentives. Given these limitations, extensive and multifunctional systems, such as silvopastoral and agroforestry models, are likely to represent the most viable land-use alternatives. Studies from Mediterranean contexts, including Italy [67,68], France [69], Greece [70] and Portugal [71], demonstrate that such systems can sustain soil fertility and biodiversity while maintaining rural livelihoods, thereby mitigating the socioecological feedbacks that reinforce abandonment and invasion.
At the governance level, the persistence of clustered invasion foci highlights the need for coordinated management across fragmented landscapes rather than isolated local interventions. In central Portugal, where landholdings are small and highly fragmented, collaborative initiatives such as Forest Intervention Zones (ZIFs) provide an institutional framework for integrated landscape management. ZIFs are legally established cooperatives that group multiple private forest owners under a shared management plan, aiming to coordinate actions related to wildfire prevention, ecological restoration, and land-use planning. By promoting cooperation among landholders and enabling joint decision-making, ZIFs enhance ecological connectivity and socioecological resilience [72,73,74,75]. Strengthening such collective frameworks could help overcome the structural constraints of small-scale ownership and reduce the fragmentation that underlies both wildfire recurrence and invasive expansion.

5. Conclusions

Overall, this work demonstrates that Acacia spp. invasions in central Portugal have undergone a rapid transition from a phase of topographically constrained establishment to a disturbance-driven expansion, strongly reinforced by recurrent wildfires and rural land abandonment. By combining multitemporal mapping, ecological modelling, and local perception surveys, this integrated approach clarifies how ecological processes and human dynamics interact to accelerate invasion trajectories.
The persistence of spatial clustering highlights the long-term influence of initial invasion nuclei, confirming that early establishment areas continue to shape expansion patterns over time. In contrast, the negative association with agricultural land use underscores the protective role of active land management in limiting invasion risk, reinforcing the importance of maintaining agroforestry practices in vulnerable rural landscapes.
Importantly, there is a strong convergence between the drivers identified by the models and those recognized by local communities. Residents consistently pointed to fire recurrence and land abandonment as key triggers of Acacia expansion, even though management responses to these processes remain limited. This alignment between scientific evidence and local perceptions strengthens the credibility of the identified drivers and highlights the value of integrating local knowledge into invasion assessments.
From a management perspective, this socio-ecological perspective suggests that effective conservation strategies should move beyond sector-specific approaches. Instead, they should integrate scientific evidence with community engagement and improved governance frameworks to address the abandonment–fire–invasion feedback loop. Targeting persistent invasion nuclei, promoting active land use, and incorporating local actors into management planning emerge as critical steps for mitigating future invasion risk in Mediterranean rural systems.

Author Contributions

Conceptualization, C.C., C.M. and M.S.; methodology, C.C. and M.S.; supervision, C.C. and C.M.; writing—original draft preparation, writing, M.S.; review and editing, visualization, C.C. and C.M. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by Fundação para a Ciência e a Tecnologia (FCT), through institutional funding (Ref. UID/00295/2025; https://doi.org/10.54499/UID/00295/2025).

Data Availability Statement

The data and code supporting this work will be made available in a public repository upon the publication of the work.

Acknowledgments

During the preparation of the graphical abstract, the authors used generative artificial intelligence tools for the creation of illustrative visual elements. The authors reviewed and edited the output and take full responsibility for the content of this publication. This work was supported by Fundação para a Ciência e a Tecnologia (FCT), through institutional funding (Ref. UID/00295/2025; https://doi.org/10.54499/UID/00295/2025).

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
GAMsGeneralized Additive Models
AUCArea Under the Receiver Operating Characteristic Curve
CAPCommon Agricultural Policy
ZIFForest Intervention Zones

Appendix A

Appendix A.1

Table A1. Temporal evolution of primary-sector employment in Sertã and Pedrógão-Grande, between 1960 and 2021: Source: INE.
Table A1. Temporal evolution of primary-sector employment in Sertã and Pedrógão-Grande, between 1960 and 2021: Source: INE.
YearSertãPedrógão-GrandeVariation (%) SertãVariation (%) Pedrógão
196074871787––
198137311203−50.2−32.7
19912526369−32.3−69.3
20011067118−57.8−68.0
201156384−47.2−28.8
2021~500~70−10.0−16.7
1960–2021––−93.3%−96.1%

Appendix A.2

Forests 17 00135 g0a1
Figure A1. Ground-truth sampling locations used for response variable validation across Sertã and Pedrógão Grande municipalities.
Figure A1. Ground-truth sampling locations used for response variable validation across Sertã and Pedrógão Grande municipalities.
Forests 17 00135 g0a1

Appendix A.3

Table A2. Structure of the questionnaire applied to assess perceptions, experiences, and management of Acacia spp. in the study area.
Table A2. Structure of the questionnaire applied to assess perceptions, experiences, and management of Acacia spp. in the study area.
Thematic AxisExample QuestionVariable TypePurpose of Inclusion
Knowlegde and perception“Do you recognise various species of wattles in the region?”Closed (Yes/No)Acess the general awareness about the species
“Are you aware of this species invasiveness?”
Land-use history“When was the last agricultural activity carried out in your property?”Closed (ordinal)Relate abandonment time with presence of Acacia
“What is the current land use type of your property?”Closed (categorical)Characterize land-use context of invaded areas
Experience of invasion“Are wattles present on your parcel?”Closed (Yes/No)Distinguish between direct and indirect experience
“In your opinion, what explains In your opinion, their presence on your property ?”OpenCapture perceived causes of invasion
Management and attitudes“Have you attempted to control or remove the species?”Closed (Yes/No)Assess management practices
“If yes, what method did you use and was it sucessful?”OpenIdentify strategies and barriers to success

Appendix A.4

Forests 17 00135 g0a2
Figure A2. Conceptual synthesis illustrating the socioecological feedback loops driving Acacia spp. invasion in central Portugal. The left section represents initial ecological filters (altitude, slope, spatial clustering); the centre depicts disturbance and socio-economic drivers (wildfire recurrence, agricultural abandonment); and the right section shows invasion outcomes (expansion, dominance, ecosystem degradation). Circular arrows (in green), highlight the positive feedbacks linking fire with regeneration and land abandonment with invasion.
Figure A2. Conceptual synthesis illustrating the socioecological feedback loops driving Acacia spp. invasion in central Portugal. The left section represents initial ecological filters (altitude, slope, spatial clustering); the centre depicts disturbance and socio-economic drivers (wildfire recurrence, agricultural abandonment); and the right section shows invasion outcomes (expansion, dominance, ecosystem degradation). Circular arrows (in green), highlight the positive feedbacks linking fire with regeneration and land abandonment with invasion.
Forests 17 00135 g0a2

References

  1. Barbera, G.; Cullotta, S. The Halaesa landscape (III B. C.) as ancient example of the complex and bio-diverse traditional Mediterranean polycultural landscape. Landsc. Hist. 2014, 35, 53–66. [Google Scholar] [CrossRef]
  2. Rescia, A.J.; Schmitz, M.F.; de Aranzabal, I.; Pineda, F.D. Biocultural Diversity Conservation: Management and Recovery of Rural Mediterranean Landscapes. In Cultural Diversity: Issues, Challenges and Perspectives; Nova Science Publishers, Inc.: Hauppauge, NY, USA, 2010; pp. 189–201. [Google Scholar]
  3. Plieninger, T. Habitat loss, fragmentation, and alteration: Quantifying the impact of land-use changes on a Spanish dehesa landscape by use of aerial photography and GIS. Landsc. Ecol. 2006, 21, 91–105. [Google Scholar] [CrossRef]
  4. Plieninger, T.; Hui, C.; Gaertner, M.; Huntsinger, L. The impact of land abandonment on species richness and abundance in the Mediterranean Basin: A meta-analysis. PLoS ONE 2014, 9, e98355. [Google Scholar] [CrossRef]
  5. Sil, Â.; Azevedo, J.C.; Fernandes, P.M.; Honrado, J.P. Will fire-smart landscape management buffer the effects of climate and land-use changes on fire regimes? Ecol. Process. 2024, 13, 57. [Google Scholar] [CrossRef]
  6. Moreira, F.; Viedma, O.; Arianoutsou, M.; Curt, T.; Koutsias, N.; Rigolot, E.; Barbati, A.; Corona, P.; Vaz, P.; Xanthopoulos, G.; et al. Landscape–wildfire interactions in southern Europe: Implications for landscape management. J. Environ. Manag. 2011, 92, 2389–2402. [Google Scholar] [CrossRef] [PubMed]
  7. Vieites-Blanco, C.; González-Prieto, S.J. Invasiveness, ecological impacts and control of acacias in southwestern Europe: A review. Web Ecol. 2020, 20, 33–51. [Google Scholar] [CrossRef]
  8. Lorenzo, P.; Palomera-Pérez, A.; Reigosa, M.J.; González, L. Allelopathic interference of invasive Acacia dealbata Link on the physiological parameters of native understory species. Plant Ecol. 2011, 212, 403–412. [Google Scholar] [CrossRef]
  9. Rodríguez-Echeverría, S.; Afonso, C.; Correia, M.; Lorenzo, P.; Roiloa, S.R. The Effect of Soil Legacy on Competition and Invasion by Acacia dealbata Link. Plant Ecol. 2013, 214, 1139–1146. [Google Scholar] [CrossRef]
  10. Rascher, K.G.; Große-Stoltenberg, A.; Máguas, C.; Werner, C. Acacia longifolia invasion impacts vegetation structure and regeneration dynamics in open dunes and pine forests. Biol. Invasions 2011, 13, 1099–1113. [Google Scholar] [CrossRef]
  11. Marchante, E.; Kjoller, A.; Struwe, S.; Freitas, H. Short- and long-term impacts of Acacia longifolia invasion on the belowground processes of a Mediterranean coastal dune ecosystem. Appl. Soil Ecol. 2008, 40, 210–217. [Google Scholar] [CrossRef]
  12. Nsikani, M.M.; van Wilgen, B.W.; Bacher, S.; Gaertner, M. Re-establishment of Protea repens after clearing invasive Acacia saligna: Consequences of soil legacy effects and a native nitrophilic weedy species. S. Afr. J. Bot. 2018, 116, 103–109. [Google Scholar] [CrossRef]
  13. Le Maitre, D.C.; Máguas, C.; Ulm, F.; Marchante, H. Ecological impacts and changes in ecosystem services and disservices mediated by invasive Australian acacia species. In Wattles: Australian Acacia Species Around the World; CABI International: Wallingford, UK, 2023; pp. 342–358. [Google Scholar] [CrossRef]
  14. Nunes, L.J.R.; Raposo, M.A.M.; Meireles, C.I.R.; Gomes, C.J.P.; Ribeiro, N.M.C.A. The impact of rural fires on the development of invasive species: Analysis of a case study with Acacia dealbata Link. in Casal do Rei (Seia, Portugal). Environments 2021, 8, 44. [Google Scholar] [CrossRef]
  15. Fuentes-Ramírez, A.; Pauchard, A.; Cavieres, L.A.; García, R.A. Survival and growth of Acacia dealbata vs. native trees across an invasion front in south-central Chile. For. Ecol. Manag. 2011, 261, 1003–1009. [Google Scholar] [CrossRef]
  16. Richardson, D.M.; Rejmánek, M. Trees and shrubs as invasive alien species–a global review. Divers. Distrib. 2011, 17, 788–809. [Google Scholar] [CrossRef]
  17. Pabst, R.; Dias, F.S.; Borda-de-Água, L.; Rodriguez-González, P.M.; Capinha, C. Assessing and Predicting the Distribution of Riparian Invasive Plants in Continental Portugal. Front. Ecol. Evol. 2022, 10, 875578. [Google Scholar] [CrossRef]
  18. Oliveira-Costa, J.L.P. A View of Biological Invasions at the Landscape Scale: A Case Study of Two Australian Acacia Species in Portugal. In Earth Observation for Monitoring and Modeling Land Use; Elsevier: Amsterdam, The Netherlands, 2024; pp. 151–181. [Google Scholar] [CrossRef]
  19. Rodríguez, J.; Lorenzo, P.; González, L. Different growth strategies to invade undisturbed plant communities by Acacia dealbata Link. For. Ecol. Manag. 2017, 399, 47–53. [Google Scholar] [CrossRef]
  20. Lima, C.G.; Campos, J.C.; Regos, A.; Honrado, J.P.; Fernandes, P.M. Fire suppression and land-use strategies drive future dynamics of an invasive plant in a fire-prone mountain area under climate change. J. Environ. Manag. 2024, 359, 120997. [Google Scholar] [CrossRef]
  21. Nunes, L.J.R.; Meireles, C.I.R.; Gomes, C.J.P.; Ribeiro, N.M.C.A. Acacia dealbata Link. Aboveground biomass assessment: Sustainability of control and eradication actions to reduce rural fires risk. Fire 2022, 5, 7. [Google Scholar] [CrossRef]
  22. Colaço, M.C.; Sequeira, A.C.; Skulska, I. Genus Acacia in mainland Portugal: Knowledge and experience of stakeholders in their management. Land 2023, 12, 2026. [Google Scholar] [CrossRef]
  23. Correia, M.; Montesinos, D.; French, K.; Rodríguez-Echeverría, S. Evidence for enemy release and increased seed production and size for two invasive Australian acacias. J. Ecol. 2016, 104, 1831–1841. [Google Scholar] [CrossRef]
  24. Passos, I.; Marchante, H.; Pinho, R.; Marchante, E.; Kjøller, A. What we don’t seed: The role of long-lived seed banks as hidden legacies of invasive plants. Plant Ecol. 2017, 218, 1313–1324. [Google Scholar] [CrossRef]
  25. Salvioni, C.; Papadopoulou, E.; Dos Santos, M. Small farm survival in Greece, Italy and Portugal. EuroChoices 2014, 13, 52–57. [Google Scholar] [CrossRef]
  26. Gomes, E.; Banos, A.; Abrantes, P.; Rocha, J.; Kristensen, S.B.P.; Busck, A. Agricultural land fragmentation analysis in a peri-urban context: From the past into the future. Ecol. Indic. 2019, 97, 380–388. [Google Scholar] [CrossRef]
  27. Sil, Â.; Fernandes, P.M.; Azevedo, J.C. Farmland abandonment decreases the fire regulation capacity and the fire protection ecosystem service in mountain landscapes. Ecosyst. Serv. 2019, 36, 100908. [Google Scholar] [CrossRef]
  28. Xavier, A.; Costa Freitas, M.B.; Antunes, C. A compromise programming approach for assessing territorial biophysical suitability: A case study. Land 2025, 14, 569. [Google Scholar] [CrossRef]
  29. Bortolus, A.; Schwindt, E. Biological invasions and human dimensions: We still need to work hard on our social perspectives. Ecol. Austral 2022, 32, 767–783. [Google Scholar] [CrossRef]
  30. Kull, C.A.; Shackleton, C.M.; Cunningham, P.J.; Ducatillion, C.; Dufour-Dror, J.-M.; Esler, K.J.; Friday, J.B.; Gouveia, A.C.; Griffin, A.R.; Marchante, E.; et al. Adoption, use, and perception of Australian acacias around the world. Divers. Distrib. 2011, 17, 822–836. [Google Scholar] [CrossRef]
  31. Rogers, P.; Nunan, F.; Fentie, A.A. Reimagining invasions: The social and cultural impacts of Prosopis on pastoralists in southern Afar, Ethiopia. Pastor. Res. Policy Pract. 2017, 7, 22. [Google Scholar] [CrossRef]
  32. Horvitz, C.C.; Qin, H.; Wang, H. Pervasive human-mediated large-scale invasion: Analysis of spread patterns and their underlying mechanisms in 17 of China’s worst invasive plants. J. Ecol. 2017, 105, 1611–1622. [Google Scholar] [CrossRef]
  33. Núñez-Tobajas, Z.; Senar, J.C.; Rodríguez-Pastor, R.; Carrillo-Ortiz, J.G.; Cardador, L. Niche shifts over spread of a biological invasion: Unveiling the role of changing habitat preference and density dependence. Divers. Distrib. 2023, 30, 4–12. [Google Scholar] [CrossRef]
  34. Gruntman, M.; Segev, U. Trait evolution in invasive plants with residence time: A meta-analysis. NeoBiota 2024, 93, 1–17. [Google Scholar] [CrossRef]
  35. Rubel, F.; Kottek, M. World Maps of Köppen–Geiger Climate Classification. University of Veterinary Medicine Vienna. Available online: https://koeppen-geiger.vu-wien.ac.at/ (accessed on 6 January 2026).
  36. Pinto-Correia, T.; Vos, W. Multifunctionality in Mediterranean Landscapes: Past and future. In The New Dimensions of the European Landscape; Jongman, R.H.G., Ed.; Springer: Dordrecht, The Netherlands, 2004; pp. 135–164. [Google Scholar]
  37. Nunes, A.N.; Coelho, C.O.A.; De Almeida, A.C.; Figueiredo, A. Soil erosion and hydrological response to land abandonment in a central inland area of Portugal. Land Degrad. Dev. 2010, 21, 260–273. [Google Scholar] [CrossRef]
  38. Lourenço, L.; Nunes, A.; Bento-Gonçalves, A.; Vieira, A. Environmental concerns of Portuguese mountains: A case study in the central mountains. In Mountains: Geology, Topography and Environmental Concerns; Johnson, R.L., Jones, J.F., Eds.; Nova Science Publishers: Hauppauge, NY, USA, 2014; pp. 195–212. [Google Scholar]
  39. Tonini, M.; Parente, J.; Pereira, M.G. Global assessment of rural–urban interface in Portugal related to land cover changes. Nat. Hazards Earth Syst. Sci. 2018, 18, 1647–1664. [Google Scholar] [CrossRef]
  40. Direção-Geral do Território. Ortofotos Digitais. 2025. Available online: https://www.dgterritorio.gov.pt/cartografia/cartografia-topografica/ortofotos/ortofotos-digitais (accessed on 16 April 2024).
  41. Direção-Geral do Território. Carta de Uso e Ocupação do Solo (COS) (COS2018). SNIG–Sistema Nacional de Informação Geográfica. 2018. Available online: https://snig.dgterritorio.gov.pt/rndg/srv/por/catalog.search#/metadata/e7d91649-71dd-4a3a-ae38-e680fc54be88 (accessed on 13 November 2024).
  42. OpenStreetMap Contributors. 2025. Available online: https://www.openstreetmap.org (accessed on 7 April 2025).
  43. Instituto da Conservação da Natureza e das Florestas. Cartografia de Áreas Ardidas. 2025. Available online: https://geocatalogo.icnf.pt/catalogo_tema5.html (accessed on 13 November 2024).
  44. Barbet-Massin, M.; Jiguet, F.; Albert, C.H.; Thuiller, W. Selecting pseudo-absences for species distribution models: How, where and how many? Methods Ecol. Evol. 2012, 3, 327–338. [Google Scholar] [CrossRef]
  45. Salmerón, R.; García, C.; García, J. Variance Inflation Factor and Condition Number in multiple linear regression. J. Stat. Comput. Simul. 2018, 88, 2365–2384. [Google Scholar] [CrossRef]
  46. Kyriazos, T.; Poga, M. Dealing with Multicollinearity in Factor Analysis: The Problem, Detections, and Solutions. Open J. Stat. 2023, 13, 404–424. [Google Scholar] [CrossRef]
  47. Comber, A.; Harris, P.; Brunsdon, C. Multiscale spatially varying coefficient modelling using a Geographical Gaussian Process GAM. Int. J. Geogr. Inf. Sci. 2023, 38, 27–47. [Google Scholar] [CrossRef]
  48. Biju, V. Non linear generalized additive models using likelihood estimations with laplace and newton approximations. J. Mech. Contin. Math. Sci. 2020, 15, 7. [Google Scholar] [CrossRef]
  49. Simpson, G. Modelling Palaeoecological Time Series Using Generalised Additive Models. Front. Ecol. Evol. 2018, 6, 149. [Google Scholar] [CrossRef]
  50. Wood, S.N. Thin plate regression splines. J. R. Stat. Soc. Ser. B Stat. Methodol. 2003, 65, 95–114. [Google Scholar] [CrossRef]
  51. Carrington, A.M.; Manuel, D.G.; Fieguth, P.W.; Ramsay, T.; Osmani, V.; Wernly, B.; Bennett, C.; Hawken, S.; Magwood, O.; Sheikh, Y.; et al. Deep ROC analysis and AUC as balanced average accuracy, for improved classifier selection, audit and explanation. IEEE Trans. Pattern Anal. Mach. Intell. 2023, 45, 329–345. [Google Scholar] [CrossRef]
  52. Moisen, G.G.; Freeman, E.A.; Blackard, J.A.; Frescino, T.S.; Zimmermann, N.E.; Edwards, T.C., Jr. Predicting tree species presence and basal area in Utah: A comparison of stochastic gradient boosting, generalized additive models, and tree-based methods. Ecol. Model. 2006, 199, 176–187. [Google Scholar] [CrossRef]
  53. Wood, S.N. Generalized Additive Modles: An Introduction with R, 2nd ed.; Chapman and Hall/CRC: Boca Raton, FL, USA, 2017; pp. 1–496. [Google Scholar]
  54. Große-Stoltenberg, A.; Werner, C.; Hellmann, C.; Oldeland, J.; Thiele, J. Quantifying the spatial impact of an invasive Acacia on ecosystem functioning using remote sensing. Ecol. Indic. 2025, 170, 112928. [Google Scholar] [CrossRef]
  55. Hernández, L.; Martínez- Fernández, J.; Cañellas, I.; de la Cueva, A.V. Assessing spatio-temporal rates, patterns and determinants of biological invasions in forest ecosystems. The case of Acacia species in NW Spain. For. Ecol. Manag. 2014, 329, 206–213. [Google Scholar] [CrossRef]
  56. Lozano, V.; Lai, R.; Cabras, S.; Campesi, G.; Bacchetta, G. Modeling the spatial distribution of the invasive species Acacia saligna (Labill.) Wendl. in Sardinia (Italy) to define management priorities. Ecol. Indic. 2020, 113, 106207. [Google Scholar] [CrossRef]
  57. Pedro, S.I.; Antunes, C.A.L.; Horta, C.; Pitacas, I.; Gonçalves, J.; Gominho, J.; Gallardo, E.; Anjos, O. Characterization of mineral composition and nutritional value of Acacia green pods. Plants 2023, 12, 1853. [Google Scholar] [CrossRef] [PubMed]
  58. Amal, I.; Shah, A.; Idrissi, M.; Asyhari, A.; Zaini, M.; Alizan, M.; Lai, D.; Slik, F. Digital Mapping of Invasive Acacia Mangium Willd. Trees Along Telisai-Lumut Highway Along the Andulau Forest Reserve. In Proceedings of the International Conference of Recent Trends in Environmental Science and Engineering, Virtual, 1–2 June 2023. [Google Scholar] [CrossRef]
  59. Zamorano, D.; Labra, F.; Vila, I.; Meier, C. Rivers as a potential dispersing agent of the invasive tree Acacia dealbata. Rev. Chil. Hist. Nat. 2022, 95, 1–14. [Google Scholar] [CrossRef]
  60. Heringer, G.; Thiele, J.; Amaral, C.; Meira-Neto, J.; Matos, F.; Lehmann, J.; Buttschardt, T.; Neri, A. Acacia invasion is facilitated by landscape permeability: The role of habitat degradation and road networks. Appl. Veg. Sci. 2020, 23, 598–609. [Google Scholar] [CrossRef]
  61. Gordon, C.E.; Price, O.F.; Tasker, E.M.; Denham, A.J. Acacia shrubs respond positively to high-severity wildfire: Implications for conservation and fuel hazard management. Sci. Total Environ. 2017, 575, 858–868. [Google Scholar] [CrossRef]
  62. Richardson, D.M.; Pyšek, P.; Rejmánek, M.; Barbour, M.G.; Panetta, F.D.; West, C.J. Naturalization and invasion of alien plants: Concepts and definitions. Divers. Distrib. 2000, 6, 93–107. [Google Scholar] [CrossRef]
  63. Shoko, C.; Mutanga, O.; Dube, T. Remotely sensed characterization of Acacia longifolia invasive plants in the Cape Floristic region of the Western Cape, South Africa. J. Appl. Remote Sens. 2020, 14, 044511. [Google Scholar] [CrossRef]
  64. United States Fish Wildlife Service (USFWS). United States Department of Agriculture (USDA), Fire Management Invasive Plants: A handbook; U.S. Department of the Interior: Arlington, VA, USA, 2005.
  65. Erickson, H.E.; White, R.H. Invasive Plant Species and the Joint Fire Science Program: A Summary of Ongoing Research, USDA Forest Service, General Technical Report PNW-GTR-707; Pacific Northwest Research Station: Portland, OR, USA, 2007. [CrossRef]
  66. Oliveras Menor, I.; Pausas, J.G.; Bond, W.J.; Bradstock, R.A.; Keeley, J.E. Integrated fire management as an adaptation and mitigation strategy to altered fire regimes. Commun. Earth Environ. 2025, 6, 278. [Google Scholar] [CrossRef]
  67. Cappai, C.; Kemanian, A.R.; Lagomarsino, A.; Roggero, P.P.; Lai, R.; Agnelli, A.E.; Seddaiu, G. Small-scale spatial variation of soil organic matter pools generated by cork oak trees in Mediterranean agro-silvo-pastoral systems. Geoderma 2017, 304, 59–67. [Google Scholar] [CrossRef]
  68. Badalamenti, E.; Scalenghe, R.; La Mantia, T.; Bueno, R.S.; Sala, G.; Pizzurro, G.M.; Giaimo, A.; Pasta, S. The cork oak in the Mountains of Palermo (Italy): Ecological insights from the south-eastern edge of its distribution range. iForest 2020, 13, 336–344. [Google Scholar] [CrossRef]
  69. Orgeas, J.; Ourcival, J.; Bonin, G. Seasonal and spatial patterns of foliar nutrients in cork oak (Quercus suber L.) growing on siliceous soils in Provence (France). Plant Ecol. 2003, 164, 201–211. [Google Scholar] [CrossRef]
  70. Kokkora, M.I.; Vrahnakis, M.; Kleftoyanni, V. Soil quality characteristics of traditional agroforestry systems in Mouzaki area, central Greece. Agroforest Syst 2022, 96, 857–871. [Google Scholar] [CrossRef]
  71. Guimarães, H.; Pinto-Correia, T.; De Belém Costa Freitas, M.; Ferraz-De-Oliveira, I.; Sales-Baptista, E.; Da Veiga, J.; Marques, T.; Pinto-Cruz, C.; Godinho, C.; Belo, A. Farming for nature in the Montado: The application of ecosystem services in a results-based model. Ecosyst. Serv. 2023, 61, 101524. [Google Scholar] [CrossRef]
  72. Marto, M.; Reynolds, K.; Borges, J.; Bushenkov, V.; Marques, S. Combining Decision Support Approaches for Optimizing the Selection of Bundles of Ecosystem Services. Forests 2018, 9, 438. [Google Scholar] [CrossRef]
  73. Valente, S.; Coelho, C.; Ribeiro, C.; Soares, J. Forest intervention areas (ZIF): A new approach for non-industrial private forest management in Portugal. Silva Lusit. 2013, 21, 137–161. [Google Scholar]
  74. Marques, M.; Juerges, N.; Borges, J. Appraisal framework for actor interest and power analysis in forest management–Insights from Northern Portugal. For. Policy Econ. 2020, 111, 102049. [Google Scholar] [CrossRef]
  75. Ferreira, L.; Baptista, A.; Constantino, M.; Marques, S.; Martins, I.; Borges, J. Integrating wildfire resistance and environmental concerns into a sustainable forest ecosystem management approach. Front. For. Glob. Change 2023, 6, 1177698. [Google Scholar] [CrossRef]
Forests 17 00135 g001
Figure 1. Study area and its location. (a) Location of the study area in mainland Portugal. (b) Detailed view of Pedrógão-Grande and Sertã municipalities. The inset in the panel represents the location of Portugal (grey) the European continent (white).
Figure 1. Study area and its location. (a) Location of the study area in mainland Portugal. (b) Detailed view of Pedrógão-Grande and Sertã municipalities. The inset in the panel represents the location of Portugal (grey) the European continent (white).
Forests 17 00135 g001
Forests 17 00135 g002
Figure 2. Workflow for Acacia patch vectorization based on orthomosaic interpretation and data sources.
Figure 2. Workflow for Acacia patch vectorization based on orthomosaic interpretation and data sources.
Forests 17 00135 g002
Forests 17 00135 g003
Figure 3. Spatiotemporal distribution of Acacia spp. stands in the municipalities of Sertã and Pedrógão-Grande for the years 2004, 2010, and 2021. Data derived from multitemporal orthophoto interpretation.
Figure 3. Spatiotemporal distribution of Acacia spp. stands in the municipalities of Sertã and Pedrógão-Grande for the years 2004, 2010, and 2021. Data derived from multitemporal orthophoto interpretation.
Forests 17 00135 g003
Forests 17 00135 g004
Figure 4. Total area occupied by Acacia spp. stands from 2004 to 2021. The invaded area expanded from 164 hectares to 848 hectares, representing a 417% increase over the 17-year period.
Figure 4. Total area occupied by Acacia spp. stands from 2004 to 2021. The invaded area expanded from 164 hectares to 848 hectares, representing a 417% increase over the 17-year period.
Forests 17 00135 g004
Forests 17 00135 g005
Figure 5. Predicted probability of Acacia spp. expansion across the study area, derived from the generalized additive model (GAM) fitted to data from 2010–2021.
Figure 5. Predicted probability of Acacia spp. expansion across the study area, derived from the generalized additive model (GAM) fitted to data from 2010–2021.
Forests 17 00135 g005
Forests 17 00135 g006
Figure 6. Multi-scale visualization of Acacia spp. predicted suitability in the study area. (a) Predictive map derived from the 2010–2021 model, showing spatial variation in invasion probability (0–0.996). (b) Enlarged section highlighting the spatial pattern of low-suitability zones. (c) Overlay of the predictive surface with agricultural land patches (in red), illustrating how actively managed farmland tends to coincide with areas of lower predicted suitability for Acacia spp. expansion.
Figure 6. Multi-scale visualization of Acacia spp. predicted suitability in the study area. (a) Predictive map derived from the 2010–2021 model, showing spatial variation in invasion probability (0–0.996). (b) Enlarged section highlighting the spatial pattern of low-suitability zones. (c) Overlay of the predictive surface with agricultural land patches (in red), illustrating how actively managed farmland tends to coincide with areas of lower predicted suitability for Acacia spp. expansion.
Forests 17 00135 g006
Table 1. Results of the Generalized Additive Models (GAMs) showing the parametric coefficients for Acacia spp. expansion during the establishment (2004–2010) and expansion (2010–2021) phases.
Table 1. Results of the Generalized Additive Models (GAMs) showing the parametric coefficients for Acacia spp. expansion during the establishment (2004–2010) and expansion (2010–2021) phases.
Predictor2004–20102010–2021
EstimateStd. Errorp-valueEstimateStd. Errorp-value
(Intercept)5.61202.15300.00926.03801.28502.64 × 10−6
Altitude−0.01920.0048<0.001−0.01930.0031<0.001
Slope−0.04160.02500.0953−0.03360.01530.0286
Agricultural Land−1.5331.0790.1555−1.57100.66100.0175
Time Since Fire--0.3816−0.00100.00480.0347
Distance to Roads−0.00140.00190.4527−0.00110.01000.9121
Distance to Water0.000470.001370.7318−0.00050.00080.5089
Direct Solar Radiation0.3310.3530.34860.33400.19550.0876
Distance to Power Lines−0.000030.000250.8997−0.00070.00170.6798
Spatial Smooth (edf)17.5-<2 × 10−1622.13-<2 × 10−16
Model Performance
R2 (explained deviance)0.398 0.389
AUC0.885 0.884
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

Salgueiro, M.; Mora, C.; Capinha, C. From Establishment to Expansion: Changing Drivers of Acacia spp. Invasion in Mainland Central Portugal. Forests 2026, 17, 135. https://doi.org/10.3390/f17010135

AMA Style

Salgueiro M, Mora C, Capinha C. From Establishment to Expansion: Changing Drivers of Acacia spp. Invasion in Mainland Central Portugal. Forests. 2026; 17(1):135. https://doi.org/10.3390/f17010135

Chicago/Turabian Style

Salgueiro, Matilde, Carla Mora, and César Capinha. 2026. "From Establishment to Expansion: Changing Drivers of Acacia spp. Invasion in Mainland Central Portugal" Forests 17, no. 1: 135. https://doi.org/10.3390/f17010135

APA Style

Salgueiro, M., Mora, C., & Capinha, C. (2026). From Establishment to Expansion: Changing Drivers of Acacia spp. Invasion in Mainland Central Portugal. Forests, 17(1), 135. https://doi.org/10.3390/f17010135

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