Multitaxon biodiversity and functional recovery in restored native forests in a biosphere reserve

Multitaxon biodiversity and functional recovery in restored native forests in a biosphere reserve Unai Ortega-Barrueta 1,* , Unai Sertutxa 2 , Ibone Ametzaga-Arregi 3 , Markel Monta˜ no-Lekue 4 , Lorena Pe˜ na 5 Department of Plant Biology and Ecology, University of the Basque Country (UPV/EHU), Apdo 644, Bilbao 48080, Spain ARTICLE INFO Key words: Exotic plantations Eucalyptus sp. Pinus radiata Plants Birds Bacteria Fungi ABSTRACT Forest landscapes in Europe have historically undergone extensive anthropogenic transformation, leading to the widespread replacement of native forests with pastures, exotic plantations, and urban developments, ultimately reducing biodiversity. Despite recent natural forest expansion and restoration efforts, this has not always been accompanied by ecological recovery. In the Basque Country (northern Iberian Peninsula), Pinus radiata and Eucalyptus plantations have become dominant, raising ecological concerns due to their low species diversity and impacts on soil processes. In response, the EU Nature Restoration Law (2024) has established binding targets to recover degraded ecosystems, promoting active restoration strategies that incorporate trait-based and multitaxon approaches. This study evaluates whether native forest restorations in the Urdaibai Biosphere Reserve support higher biodiversity and ecosystem functioning compared to the most abundant forest systems in the area. Four forest types (restored native forests, Pinus radiata plantations, Eucalyptus plantations, and reference native forest) were compared across 36 plots. We assessed vascular plants, birds, bacteria, and fungi using field surveys, soil DNA sequencing, and functional trait assignments. Diversity metrics and both taxonomic and functional composition were compared among systems. Restored forests showed significantly higher plant and bird diversity than exotic plantations, with Eucalyptus consistently exhibiting lowest values across most indicators. Species and functional composition analyses revealed that restored forests closely resembled native systems, while Eucalyptus was associated with early-successional and stress-tolerant traits. Pine plantations exhibited intermediate patterns, particularly in microbial communities. Overall, active restoration with native species proved effective in recovering biodiversity and ecosystem function, though full convergence with native forests may require longer timescales. 1. Introduction European forest landscapes have undergone centuries of anthropogenic transformation. The replacement or fragmentation of native forests by agricultural land, pastures, forestry monoculture systems or urban developments (Kaplan et al., 2009) has led to a decline in biodiversity (Maes et al., 2023). Natural expansion of native forests, largely driven by rural abandonment, has contributed to increased forest cover in recent times. However, large-scale afforestation and reforestation programmes implemented during the twentieth century to restore forest cover has mostly been aimed at supplying timber markets (Mather, 2001; Chazdon, 2008; Vadell et al., 2016). Furthermore, in recent years, this trend has intensified under the umbrella of climate policy (CO 2 removal), incentivising the expansion of fast-growing exotic forest plantations. This phenomenon has been particularly evident in the Iberian Peninsula and Scandinavia, reflecting a growing pressure on * Corresponding author. E-mail addresses: [email protected] (U. Ortega-Barrueta), [email protected] (U. Sertutxa), [email protected] (I. Ametzaga-Arregi), [email protected] (M. Monta˜ no-Lekue), [email protected] (L. Pe˜ na). 1 ORCID: 0000–0001-7773–7973 2 ORCID: 0000–0003-1961–5649 3 ORCID: 0000–0003-0683–1190 4 ORCID: 0009–0008-5203–505X 5 ORCID: 0000–0001-8329–9869 Contents lists available at ScienceDirect Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco https://doi.org/10.1016/j.foreco.2025.123057 Received 2 June 2025; Received in revised form 29 July 2025; Accepted 31 July 2025 Forest Ecology and Management 596 (2025) 123057 Available online 14 August 2025 0378-1127/© 2025 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/bync-nd/4.0/ ). forest systems linked to intensify management practices (Ceccherini et al., 2020). However, maximizing short-term carbon sequestration through monoculture plantations may lead to trade-offs with biodiversity, especially in temperate forests (Sabatini et al., 2019, Smith et al., 2024). The northern Iberian Peninsula provides a clear example of the ecological impact of these policies, where intensive plantation forestry has long shaped the landscape (Vadell et al., 2016). Concretely in the Basque Country, over the past century, Pinus radiata monocultures have spread rapidly across the landscape due to their rapid growth and timber yield (Michel Rodríguez, 2006). However, in recent decades, the decline in pine profitability, driven by market instability and widespread outbreaks of red and brown needle blight (Monteiro et al., 2022; Ortu˜ no et al., 2022), has led to an increasing substitution with Eucalyptus sp. plantations (Gobierno Vasco, 2024; Vadell et al., 2022). This change in land use raises ecological concerns, as eucalyptus plantations exhibit low species diversity and alter soil conditions, thereby affecting key ecosystem processes such as nutrient cycling and litter decomposition (Calvi˜ no-Cancela et al., 2012; Goded et al., 2019; Mallen-Cooper et al., 2022; Sertutxa et al., 2025a). These impacts are associated not only with their intensive management practices, such as their short rotation cycles, compared to pine plantations, but also with intrinsic characteristics of eucalyptus, such as its slow-decomposing litter, production of allelopathic compounds and a high water use (Amazonas et al., 2018; Elosegi et al., 2020). In response to increasing environmental pressures, in 2024 the EU Nature Restoration Law was established in the European Union as part of the European Green Deal and the EU Biodiversity Strategy for 2030. The Regulation establishes legally binding restoration targets for a wide range of ecosystems, with the aim of restoring 20 % of the EU’s degraded ecosystems by 2030 and all ecosystems in need of restoration by 2050 (Regulation 2024/1991). Indeed, ecological restoration has emerged as a pivotal strategy for reversing biodiversity loss, with various organisations employing land stewardships to carry out restoration activities in different ecosystems. The goal of restoration ecology is to increase vegetation cover and restore ecological processes, species interactions, and community structures that resemble those observed in reference ecosystems (Chazdon, 2008; Gatica-Saavedra et al., 2017). In this context, active restoration measures, encompassing the planting of native species, the eradication of invasive alien species, and the protection of regenerating vegetation, are frequently imperative in highly degraded or fragmented landscapes (Crouzeilles et al., 2017; Meli et al., 2017). Nevertheless, assessing the efficacy of restoration endeavours remains a challenge, as conventional metrics, such as species richness and canopy cover, only offer a limited perspective on ecological recovery. Consequently, restoration success is increasingly being evaluated using biodiversity metrics that include taxonomic richness and composition, as well as functional diversity, defined as the range of traits influencing ecosystem functioning (Cadotte et al., 2011; Gatica-Saavedra et al., 2017). The incorporation of functional traits enables a deeper understanding of how restored communities resemble to reference systems and how they regulate key ecological processes over time. The application of functional traits has been demonstrated to facilitate the design of more resilient and adaptive restoration strategies. This is achieved by aligning species selection and management strategies with specific environmental concerns and desired ecosystem functions (Loureiro et al., 2023). Furthermore, there is a growing consensus that a multitaxon approach is crucial for accurately capturing restoration outcomes across various trophic levels (Seibold et al., 2018). This is particularly relevant given that different taxonomic groups respond asynchronously to environmental change and restoration interventions, in terms of intensity and direction of change. For this reason, methodological approaches based on a single taxon have the potential to yield misleading conclusions (Crouzeilles et al., 2016; Doerfler et al., 2020). The inclusion of both aboveground and belowground organisms is therefore essential in order to achieve a comprehensive evaluation of the restoration outcomes, given the sensitivity of vascular plants, birds, bacteria, and fungi to alterations resulting from forest management practices (Bardgett and van der Putten, 2014). Indeed, vascular plants and birds are frequently utilised as bioindicators due to their sensitivity to habitat structure, vegetation complexity, and landscape composition. Furthermore, they fulfil pivotal functions in ecosystem processes, including primary production, seed dispersal and trophic regulation (Massi et al., 2022; Jung et al., 2024). However, the majority of these assessments demonstrate a paucity of consideration with regard to microbial diversity and functioning. Soil microbial communities, comprising bacteria and fungi, are fundamental for nutrient cycling and decomposition. Their pronounced response to alterations in litter quality and soil chemistry renders them increasingly valuable as indicators of ecological recovery (Sun et al., 2017; Sansupa et al., 2021; Huang et al., 2023). The integration of soil microbial communities with aboveground taxa has been demonstrated to enhance the ability to detect restoration outcomes and assess ecosystem resilience in a more comprehensive manner (Hou et al., 2024). Nevertheless, empirical assessments applying this multitaxon, functional perspective remain scarce, particularly in biosphere reserves characterised by degraded or fragmented landscapes, such as the Urdaibai Biosphere Reserve (UBR). These areas have been designated for the promotion of sustainable development, with a view to enhancing the well-being of the local population (Ishwaran et al., 2008) and achieving conservation objectives. For this reason, in recent years, in UBR restoration efforts led by organisations such as Fundaci´ on Lurgaia have focused on actively transforming monocultures into native forests and restoring degraded and fragmented native forests. These restorations have involved planting of native species to improve biodiversity and landscape connectivity (Ortega et al., 2023). The aim of this study is to evaluate whether the restoration of native forests in the UBR is more effective at supporting biodiversity and key ecosystem functioning compared to the most abundant forest systems. In order to achieve this objective, we conducted a comparative analysis of the species and functional diversity and composition of four forest systems across multiple taxonomic groups (vascular plants, birds, bacteria and fungi). Concretely, restored native forests were compared to Pinus radiata and Eucalyptus sp. plantations of similar age, as well as to native forests that serve as reference ecosystems in the region. This multitaxon comparison allowed us to examine whether restored forests exhibited signs of convergence with reference forest or with forest plantations. In addition, we identified the main ecological functions associated with each forest type in order to assess whether restored forests are functionally closer to native forests or to exotic plantations. 2. Methods 2.1. Study area The UBR, located in the Basque Country in northern Iberian Peninsula (Fig. 1), encompasses approximately 22,000 ha and was designated a UNESCO Biosphere Reserve in 1984 and subsequently integrated into the Natura 2000 network. This designation has promoted the sustainable use of natural resources while fostering ecosystem protection, giving rise to a socio-ecological system characterised by the coexistence between conservation goals and economic activities (Onaindia et al., 2013b, Castillo-Eguskitza et al., 2017). The UBR landscape is characterised by a heterogeneous mosaic shaped by historical land transformations and current anthropogenic pressures. The region has a humid Atlantic climate, with mean annual temperatures exceeding 13◦C and annual precipitation ranging from 1200 to 1700 mm, distributed relatively evenly throughout the year. The altitudinal gradient ranges from sea level to 776 m, though most of the reserve lies below 400 m. The lithology of the area is composed of cretaceous formations, primarily flysch, marl and sandstone, that contribute to the development of acidic, oligotrophic soils (pH 3.5–6.0) that support acidophilous forest U. Ortega-Barrueta et al. Forest Ecology and Management 596 (2025) 123057 2 communities (Loidi et al., 2011; Orrantia et al., 2019). The Cantabrian green oak forest is considered one of the most highly valued natural forests of the reserve; however, mixed-oak forests dominated by Quercus robur L. are the dominant potential vegetation in the UBR (Onaindia et al., 2013a). Most of these mixed-oak forests are constituted of small stands, typically between 35 and 70 years, with minimal or virtually no silvicultural intervention. Currently, they occupy less than 9 % of the reserve, while monocultures of fast-growing exotic species for silvicultural purposes, including Pinus radiata and Eucalyptus sp., cover over 50 % of the landscape (Gobierno Vasco, 2024). Concretely, Pinus radiata, also known as Monterey pine, is native to a restricted coastal range in California and northern Baja California (Michel Rodríguez, 2006), where it plays a relevant role in maintaining ecosystem dynamics through rapid growth and high disturbance resilience. Regarding eucalyptus plantations, the main species planted in the region are Eucalyptus globulus Labill. and Eucalyptus nitens H. Deane & Maiden, which are native to southeastern Australia and constitute a fundamental element of temperate sclerophyll forests in that region. The establishment of pine plantations in the Basque Country, which are predominantly situated on privately owned land that previously functioned as pastures, marginal agricultural croplands and native forests, can be attributed to the mid-20th century afforestation policies. However, area of eucalyptus plantations has doubled from 2005 to 2024 due to substitution of pine with eucalyptus due to the brown and red band disease that have recently affected pine plantations in the region. These changes in the landscape are a response to socio-economic changes and declining profitability of traditional agricultural practices and forestry practices (Sertutxa et al., 2025a). In the Basque Country, the process of establishing forest plantations is characterised by a sequence of steps, beginning with soil preparation followed by tree planting. In the case of pine plantations, subsequently two to three thinning operations are typically conducted during the stand’s development, with rotation cycles of 30–40 years. In the case of eucalyptus plantations, rotation cycles are of 10–15 years and subsequent to harvesting, the stumps are permitted to resprout. These plantations are often situated in areas with challenging topography, leading to ecological fragmentation and loss of functional connectivity. Despite these challenges, the region harbours habitats of high conservation value, including riparian forests, Cantabrian holm oak groves, and estuarine marshes. These habitats contribute to the area’s elevated floristic and faunal diversity (Castillo-Eguskitza et al., 2017). 2.2. Stand selection Firstly, 6 sampling points were established randomly in the two oldest forest restoration areas that are managed by Fundaci´ on Lurgaia, 3 sampling points in each one with a minimum distance of 200 m (Barlow et al., 2007). The number of sampling points per site was determined based on the size of the restoration areas (5 and 7 ha) and the minimum distance required between sampling points to reduce spatial autocorrelation. These areas have undergone ecological restoration for approximately 15–20 years and are referred to herein as “Restoration”. The restoration management strategy encompassed the planting of native tree species, including native trees such as Quercus robur, Corylus avellana, and Fraxinus excelsior. Protective barriers were placed around each tree, and the area was subjected to monitoring to prevent the proliferation of invasive species. Subsequently, 10 Eucalyptus sp. plantations stands (hereafter referred to as “Eucalyptus”) and 10 Pinus radiata plantations stands (hereafter referred to as “Pine”) were selected randomly; both were approximately 20 years old. Finally, 10 mixed oak forests stands older than 35 years (hereafter referred to as “Native”) were selected randomly (Fig. 1). Orthophotos were used to verify the age Fig. 1. Map showing the geographic location of the sampling points for Native forests, Restoration sites, Pine plantations, and Eucalyptus plantations in the UBR located in the Basque Country. U. Ortega-Barrueta et al. Forest Ecology and Management 596 (2025) 123057 3 of the forest systems. In each stand, a sampling point was established at the centre to avoid edge effects. A summary table with key site descriptors (elevation, slope, forest structure metrics, soil pH, and lithology) for each sampling point is provided in Appendix B (Table B.1). 2.3. Sampling design Field surveys were performed between April and June in 2023. In order to sample the vascular plant species, from each sample point, two perpendicular transects of 50 m were established, forming a cross. A total of five subplots, each measuring 10 m² (2 m ×5 m), were established at regular intervals along the transects. The central subplot was shared by both transects, resulting in a total sampled area of 90 m² per plot. In each subplot, the percentage of cover for each plant species was visually estimated. The mean values for each species per plot were obtained by averaging the cover estimates across subplots (Brower and Zar, 1978; Onaindia et al., 2004). Plant species and their occurrences are in Table B.3. Bird sampling was carried out by point count methods at the same central subplot established for vegetation sampling. Sampling was conducted during the spring season, which coincides with the period of maximum breeding activity, from dawn until two hours after sunrise and without adverse weather conditions. Samplings were carried out by the same person. One plot of each forest system was sampled each day, changing the order every day to avoid effects of time of day (Goded et al., 2019). At each sampling point, a 15-minute observation period was conducted, during which all bird species observed or heard were recorded (Proença et al., 2010). Bird species and their occurrences are in Table B.4. Soil sampling for microbial community analysis was conducted at three randomly selected subplot within each sampling point in order to capture spatial heterogeneity. At each centroid, composite soil samples were collected by combining three soil cores taken from a depth of 20 cm. The sampling depth was set at 20 cm, as this level encompasses the biologically active surface layer prior to the transition zone in soil microbial communities (Guo et al., 2022). Soil samples were first homogenized and sieved to remove coarse debris and roots. From each homogenized sample, 250 mg of soil were used for total DNA extraction using the DNeasy PowerSoil Pro Kit (Qiagen), in accordance with the manufacturer’s protocol, to ensure the procurement of high-quality nucleic acid for downstream applications in sequencing. Extracted DNA was subsequently dispatched to Novogene (Cambridge, UK) for amplicon-based sequencing of bacterial and fungal communities. The 16S rRNA gene (V3 - V4 region) was targeted for bacterial identification, while the ITS2 region was amplified for fungal identification (Bu´ ee et al., 2009; Christensen et al., 2023). The PCR amplification was performed using 15 µL of Phusion® High-Fidelity PCR Master Mix (New England Biolabs), with a concentration of 0.2 µM for each forward and reverse primer and approximately 10 ng of template DNA. The sequencing was conducted on a paired-end Illumina platform, generating 250 bp paired-end reads. The raw sequences were then subjected to stringent quality control measures, which included primer trimming and quality filtering. The denoising and chimera removal processes were conducted using the DADA2 pipeline, thereby yielding in high-resolution amplicon sequence variants (ASVs). Finally, an average value of ASV abundance for each plot was calculated considering the three centroids. 2.4. Functional traits assignment The functional traits of plants, birds, bacteria, and fungal species were characterised to evaluate the functional diversity and composition of community functions. The selection of these traits was based on their ecological relevance and their potential to influence ecosystem functioning. The description of the functional traits for each taxon appear in Appendix A. The assignment of plant functional traits was conducted at the species level using the FloraVeg.eu database. The following traits were assigned: life form (tree, shrub, herbaceous, and fern); dispersal mode (endozoochory, myrmecochory, anemochory, dyszoochory, and local dispersal); dispersal distance (categorised into six distance classes, ranging from short (1) to long distances (6)); Specific Leaf Area (SLA) (low, medium, and high); shade tolerance (low, medium, and high); and nutrient requirement (oligotrophic, mesotrophic, eutrophic). The categorisation of SLA and shade tolerance was conducted through the implementation of the Jenks natural breaks classification method. In the case of shade tolerance, the data were derived from the Ellenberg-type light indicator value. The functional traits exhibited by bird species were assigned based on ecological behaviour and morphological characteristics. The data from Pedley et al. (2019) and Seobirdlife (https://seo.org/) were used to assign the following traits: body mass (small and medium); foraging site (ground, vegetation, and mixed); nesting site (ground, cavities, tree, shrub, and mixed); habitat type (forest species and non-forest species); and diet (insectivorous and mixed-diet). The mean body mass for each species was categorised into the two groups using the Jenks natural breaks classification method. In the case of bacteria, firstly a taxonomic assignment was conducted using the SILVA 138.1 database for bacterial 16S sequences, employing a 97 % identity threshold (Quast et al., 2013). Subsequently, their functional characteristics were determined, by means of the FAPROTAX database (Louca et al., 2016), a system that enables the allocation of functional roles based on taxonomic classification. In our dataset, only 21.4 % of bacterial ASVs could be assigned to at least one functional group, while 78.6 % remained unassigned. This is a common limitation in bacterial functional annotation, as many soil taxa remain poorly characterized, which may lead to underestimation of certain functional groups. Functions that were present in a minimum of five stands were selected. The following functional groups were selected: carbon metabolism (chemoheterotrophy, photoheterotrophy, cellulolysis, aromatic hydrocarbon degradation, fermentation, and methylotrophy); nitrogen metabolism (nitrogen fixation, nitrification, denitrification, nitrate respiration, and ureolysis); iron respiration; manganese oxidation; human-associated bacteria; and animal parasites or symbionts. In the case of fungi, the taxonomic assignment was conducted using and the UNITE database for fungal ITS2 sequences, employing a 97 % identity threshold (Abarenkov et al., 2023). Subsequently, the assignment of fungal functional traits was conducted using the FUNGuild database (Nguyen et al., 2016), which associates fungal taxa with ecological guilds. In our dataset, 34 % of fungal ASVs were successfully assigned to ecological guilds, whereas 65 % could not be linked to specific functional categories. The low assignment rate reflects known gaps in fungal guild databases, potentially limiting the detection of less-studied ecological functions. It is important to note that the study only encompassed guild assignments characterised by a high level of confidence (high probable and probable) and guilds that were present in a minimum of five stands. The following guilds were considered: saprotrophs (wood, plant, dung, and undefined); mycorrhizal fungi (ectomycorrhizal and ericoid mycorrhizal); pathogens (plant and animal); parasites (plant and animal); and endophytes. It is important to note that functional annotations derived from both FAPROTAX and FUNGuild reflect potential metabolic or ecological functions inferred from taxonomic identities, rather than directly measured activities. This incomplete functional coverage must be considered when interpreting the ecological roles of microbial communities in our study. 2.5. Species diversity and composition Plant and bird species richness and Shannon diversity index (Shannnon and Weaver, 1949) were calculated for each stand. In the case of plants, the percentage of cover for each species was used, whereas for birds, the bird abundance was considered. Concurrently, U. Ortega-Barrueta et al. Forest Ecology and Management 596 (2025) 123057 4 species composition was evaluated using the Bray-Curtis dissimilarity index. In the case of bacteria and fungi, alpha diversity metrics were calculated after rarefaction to standard sequencing depth. Specifically, the number of observed features (ASV richness) and the Shannon diversity index were computed. For composition, a phylogenetic tree was constructed using FastTree based on multiple sequence alignments of ASV representative sequences. This tree, together with ASV abundance data, was used to compute pairwise community dissimilarities employing the Weighted UniFrac metric for each stand. This metric incorporates both phylogenetic relatedness and relative abundances of taxa. 2.6. Functional diversity and composition The calculation of functional diversity for the four taxonomic groups was conducted using Rao’s quadratic entropy, a metric that accounts for both species abundances and functional trait dissimilarities within a community. The functional distances between species were computed using Gower’s similarity coefficient, and to improve the accuracy of the results a square root correction was applied (Mammola et al., 2021). Although we did not explicitly test for multicollinearity among functional traits, the calculation of Rao’s Q inherently accounts for trait redundancy through the use of a distance matrix, which reduces the influence of overlapping trait information. Functional composition was assessed using the proportional representation of each functional trait within the community. These relative abundances were used to calculate Bray-Curtis dissimilarity matrices between stands. Collinearity between traits was not addressed prior to NMDS or PERMANOVA, as these are exploratory, non-parametric analyses. Trait correlations were expected due to the categorical and mutually exclusive nature of certain traits (e.g., SLA or shade tolerance), but all traits were retained to preserve ecological interpretability. 2.7. Statistical analysis To compare species diversity (richness and Shannon index) and functional diversity among the four forest systems, we first assessed the normality of the data using the Shapiro-Wilk test. Homogeneity of variances was then evaluated with Levene’s test. Based on the outcomes of these preliminary tests, different statistical approaches were performed: when data met both normality and homoscedasticity assumptions, we used one-way ANOVA followed by Tukey’s Honest Significant Difference test for post-hoc pairwise comparisons. If the data were normally distributed but showed unequal variances, Welch’s ANOVA was performed, with pairwise differences evaluated using the Games-Howell test. In cases of non-normal but homoscedastic data, the KruskalWallis test was used, followed by Dunn’s test with appropriate correction for multiple comparisons. Finally, when neither assumptions were met, a bootstrap-based procedure was employed to assess pairwise differences robustly. In the case of species and functional composition, we computed a Principal Coordinates Analysis (PCoA) and a Non-Metric Multidimensional Scaling (NMDS) analysis, respectively, to visualize the difference patterns of species composition. Subsequently, for both cases, a Permutational Multivariate Analysis of Variance (PERMANOVA) was conducted to examine the statistic differences among forest systems. Finally, in order to identify the functional traits that were most indicative of each forest system, we conducted an Indicator Species Analysis at functional level. This facilitated the identification of specific functional attributes that characterised each system. All statistical tests were performed using R software (v 4.2.2; R Core Team, 2024) in R Studio platform (RStudio Team, 2023). 3. Results 3.1. Species and functional diversity Plants species richness and Shannon diversity index were found to be significantly higher in Native and Restoration than in forest plantations. The Eucalyptus exhibited the lowest values for both diversity metrics. In the context of plant functional diversity, Eucalyptus exhibited significantly lower values in comparison to the other forest systems, with the exception of Pine. Furthermore, the richness of bird species was found to be significantly higher in Native than in forest plantations, and significantly higher in Restoration than in Eucalyptus. In the case of Shannon diversity, Native and Restoration demonstrated significantly higher values than Eucalyptus. However, bird functional diversity did not demonstrate significant differences among any forest systems (Fig. 2; Table B.2). With regard to bacterial species richness and Shannon diversity, no significant differences were observed among forest systems; however, functional diversity was higher in Native and Restoration compared to Eucalyptus. Ultimately, the richness of fungal species did not demonstrate a significant difference among forest systems. In contrast, Shannon diversity was significantly lower in Eucalyptus than in Native and Pine, while, functional diversity was only lower in Eucalyptus than in Native (Fig. 2; Table B.2). 3.2. Species composition In the context of plant species, Native was distinguished from Eucalyptus and Pine, as evidenced by the results of the PCoA (Fig. 3). This result was confirmed by significant differences between Native and Pine, and Native and Eucalyptus in the PERMANOVA results. In a similar manner, Restoration differed significantly from Pine and Eucalyptus, yet not from Native, indicating a recovery of plant community composition in restored sites. Finally, Pine and Eucalyptus also exhibited significant differences in species composition between both (Table 1). In the case of birds, the ordination exhibited a less clearly defined separation among forest systems in comparison to that observed in plants (Fig. 3). However, significant differences were detected between the Native and Pine, and Native and Eucalyptus. In the case of Restoration, no significant differences were observed with Native, Pine or Eucalyptus (Table 1). For bacterial communities, the PCoA demonstrated considerable overlap among all forest systems, though some differentiation was evident between Eucalyptus and Restoration (Fig. 3). Indeed, Eucalyptus showed significant differences with Native, Pine, and Restoration in PERMANOVA analysis. Moreover, a significant difference was observed between the Pine and Restoration, in contrast to the non-significant differences between the Pine and Native, and Native and Restoration (Table 1). In the case of fungal communities, a moderate separation was exhibited between Eucalyptus and the other forest systems (Fig. 3). This finding was corroborated by the PERMANOVA results, which revealed significant differences between Eucalyptus and the remaining forest systems (Table 1). 3.3. Functional composition In plants, Eucalyptus and Pine exhibited a greater divergence from both Native and Restoration, as evidenced by the NMDS analysis (Fig. 4). However, the PERMANOVA analysis revealed significant differences among all forest systems (Table 2). Furthermore, the following traits were found to be more associated with Native and Restoration: life forms (trees, and herbs), animal dispersal (endozoochorous and dyszoochorous), dispersal distance (low), SLA (high), shade tolerant (high), and nutrient requirement (eutrophic and mesotrophic). In contrast, Pine and Eucalyptus were distributed in proximity to the following traits: life U. Ortega-Barrueta et al. Forest Ecology and Management 596 (2025) 123057 5 forms (ferns), dispersal mode (anemochorous and myrmecochorous), dispersal distance (high), shade tolerance (low and medium), and nutrient requirement (oligotrophic). In birds, the NMDS analysis indicated that Eucalyptus were slightly separated from the other forest systems (Fig. 4). Indeed, its functional composition differed significantly with both Native and Pine, yet did no differ from Restoration (Table 2). Traits’ distribution is not clearly associated with forest systems but Eucalyptus seemed to be more associated with the following traits: foraging site (vegetation), habitat type (non-forest species), and diet (insectivorous). In contrast, Native, Pine and Restoration were relatively more closely linked to the following traits: body mass (medium), foraging site (ground and mixed), nesting site (tree and mixed), habitat type (forest species), and diet (mixed-diet). In bacteria, the NMDS exhibited a distinct separation between Fig. 2. Boxplots of species richness, Shannon diversity, and functional diversity for plants, birds, bacteria, and fungi across the four forest systems: Native, Pine, Eucalyptus, and Restoration. The central line of each box represents the median, while the lower and upper edges correspond to the 25th and 75th percentiles, respectively. Whiskers extend to 1.5 times the interquartile range, and dots represent outliers. Statistical differences among forest systems were tested using different methods depending on data normality and homoscedasticity (Table B.2). Different letters indicate significant differences between groups (p <0.05). U. Ortega-Barrueta et al. Forest Ecology and Management 596 (2025) 123057 6 Eucalyptus and Restoration (Fig. 4). Indeed, the results of PERMANOVA indicated significant differences between these forest systems, as well as between Eucalyptus and Native, and Restoration and Pine (Table 2). The distribution of functions related to nitrogen metabolism and manganese oxidation were distributed near Restoration, Native and Pine plots, while those functions associated with carbon metabolism, iron respiration and human-associated bacteria were distributed in the proximity of Eucalyptus. In fungi, Eucalyptus exhibited distinctive functional composition that set it slightly apart from other forest systems (Fig. 4). A detailed analysis revealed that its functional composition differed significantly from that of the other systems (Table 2). Specifically, the ericoid mycorrhizal guild was identified associated with Eucalyptus systems. However, it is evident that Native, Restoration and Pine were more Fig. 3. PCoA plots representing community composition for plants, birds, bacteria and fungi across the four forest systems (Native, Restoration, Pine, and Eucalyptus). For plants and birds, Bray-Curtis distances were used, while for bacteria and fungi Weighted UniFrac distances were applied. Axes represent the first two principal coordinates, which explain the highest proportion of variance in community dissimilarity. Small points indicate the stands and the triangles indicate the centroid of each forest system. Table 1 Results of PERMANOVA analysis for species composition for plants, birds, bacteria and fungi among the pairs of the four forest systems (Native, Restoration, Pine, and Eucalyptus). Significance: ***: p ≤0.001; **: p ≤0.01; *: p ≤0.05. Aboveground taxon Comparison between forest systems p-value Belowground taxon Comparison between forest systems p-value Plants Native - Pine 0.002 ** Bacteria Native - Pine 0.21 Native - Eucalyptus 0.001 *** Native - Eucalyptus 0.017 * Native - Restoration 0.072 Native - Restoration 0.097 Pine - Eucalyptus 0.001 *** Pine - Eucalyptus 0.025 * Pine - Restoration 0.001 *** Pine - Restoration 0.016 * Eucalyptus - Restoration 0.001 *** Eucalyptus - Restoration 0.001 *** Birds Native - Pine 0.019 * Fungi Native - Pine 0.48 Native - Eucalyptus 0.011 * Native - Eucalyptus 0.002 ** Native - Restoration 0.603 Native - Restoration 0.345 Pine - Eucalyptus 0.067 Pine - Eucalyptus 0.034 * Pine - Restoration 0.306 Pine - Restoration 0.096 Eucalyptus - Restoration 0.052 Eucalyptus - Restoration 0.001 *** U. Ortega-Barrueta et al. Forest Ecology and Management 596 (2025) 123057 7 Fig. 4. NMDS plots representing the functional composition of plant, bird, bacterial, and fungal communities across the four forest systems (Native, Restoration, Pine, and Eucalyptus). Bray-Curtis distances based on relative abundances of functional groups were used to compute dissimilarities between stands. See the meaning of plant and bird traits acronyms in Table A.1 and A.2. Table 2 Results of PERMANOVA analysis for functional composition for plants, birds, bacteria and fungi among the pairs of the four forest systems (Native, Restoration, Pine, and Eucalyptus). Significance: ***: p ≤0.001; **: p ≤0.01; *: p ≤0.05. Aboveground taxon Comparison between forest systems p-value Belowground taxon Comparison between forest systems p-value Plants Native - Pine 0.001 *** Bacteria Native - Pine 0.363 Native - Eucalyptus 0.002 *** Native - Eucalyptus 0.021 * Native - Restoration 0.016 * Native - Restoration 0.226 Pine - Eucalyptus 0.033 * Pine - Eucalyptus 0.061 Pine - Restoration 0.001 *** Pine - Restoration 0.017 * Eucalyptus - Restoration 0.001 *** Eucalyptus - Restoration 0.001 *** Birds Native - Pine 0.161 Fungi Native - Pine 0.274 Native - Eucalyptus 0.011 * Native - Eucalyptus 0.001 *** Native - Restoration 0.717 Native - Restoration 0.366 Pine - Eucalyptus 0.016 * Pine - Eucalyptus 0.02 * Pine - Restoration 0.912 Pine - Restoration 0.211 Eucalyptus - Restoration 0.268 Eucalyptus - Restoration 0.002 ** U. Ortega-Barrueta et al. Forest Ecology and Management 596 (2025) 123057 8 closely associated with the following guilds: saprotrophs (wood, plant and dung) and pathogens (plant and animal). 3.4. Functional indicators In plants, the traits indicative of Native and Restoration were as follows: dispersal mode (dyszoochorus) (p =0.010), dispersal distance (class 4) (p =0.001), and shade tolerance (high) (p =0.008). Meanwhile, the following traits were indicative of Native, Restoration and Pine: SLA (high), and dispersal mode (local dispersal) and distance (class 2) (p =0.001 for all of them). Concurrently, the trait of birds nesting site (tree) (p =0.009) and bacterial groups related to nitrogen metabolism (denitrification (p =0.019) and nitrate respiration (p =0.031)) were identified as indicators of them as well. In the case of fungi, no significant indicators were identified. 4. Discussion 4.1. Ecological impacts of forest plantations Despite the long-standing presence and spatial dominance of monoculture plantations in the UBR, their biodiversity remains markedly inferior to that of native forests. Indeed, the results obtained demonstrate that plantations of Pinus radiata and Eucalyptus sp. support significantly lower species and functional diversity, and a distinct composition of aboveand belowground organisms, compared to both native and restored forests. Concretely, plant species and functional diversity in monoculture plantations were found to be significantly lower than in native forests, as these plantations tend to simplify forest structure, reduce niche availability, and constrain ecological processes (Liu et al., 2018; Wang et al., 2022; Vu Ho et al., 2023). Species and function compositional analyses further revealed that both species and functional assemblages in plantations were significantly distinct from those in native forests. Indeed, traits associated with early-successional or stress-tolerant strategies, such as low shade tolerance, long-distance dispersal, oligotrophic affinity, and prevalence of anemochorous syndrome or myrmecochorous syndrome, have been linked to pine and eucalyptus stands. These trait profiles are typical of communities that are established in simplified forest structures and under resource-limited conditions, which are characteristic of intensively managed forests (Brockerhoff et al., 2003; Meers et al., 2010; Calvi˜ no-Cancela et al., 2012). In contrast, native and restored forests were associated with zoochorous dispersal, high SLA, eutrophic conditions, and higher shade tolerance, reflecting more advanced successional stages and increased ecological complexity (Suganuma et al., 2014; Blandino et al., 2021). Canopy architecture could contribute to these ecological contrasts, as the vertical leaf orientation of eucalyptus and the generally lower crown closure of young pine plantations allow higher light penetration, thereby promoting understorey dominance by heliophilous and generalist species (Alday et al., 2017; Elosegi et al., 2020). Such species include the fern Pteridium aquilinum and the shrubs like Rubus sp. and Ulex sp. that have been observed to aggressively colonise open forest floors and can dominate the understorey in coniferous or eucalyptus plantations contributing to understorey homogenization (Onaindia et al., 2013a; Sertutxa et al., 2024). The most marked divergence was observed in Eucalyptus plantations for all taxonomic groups. In the case of birds, Eucalyptus plantations supported a significantly lower species diversity than Restored and Native, and a distinct species composition. Furthermore, the functional composition was significantly different from both the Native and Pine systems, thus, being less associated with traits such as forest specialists. These findings are consistent with previous studies in northern Spain, which have shown that eucalyptus stands harbour only a nested subset of forest bird assemblages, leading to biotic homogenisation and the loss of forest specialists (Goded et al., 2019). The quality of avian habitat is constrained by reduced shrub diversity, low canopy heterogeneity, and the lack of nesting substrates (e.g., cavities and deadwood). Furthermore, Eucalyptus litter and bark chemistry reduce epiphytic and invertebrate abundance, thereby diminishing food resources for birds (Calvi˜ no-Cancela.,2013; García-Fernandez et al., 2025). This trend was further confirmed by the analysis of soil microbial communities. Despite the absence of a significant difference in bacterial and fungal species richness across forest systems, a notable divergence in functional diversity and composition was observed, with Eucalyptus plantations exhibiting the most pronounced disparities. In these plantations, bacterial communities were dominated by functional groups associated with carbon metabolism, iron respiration, and human-related taxa, while functional groups associated with nitrogen metabolism were less represented. This shift in bacterial communities reflects altered edaphic conditions, including lower nutrient availability, more acidic soils, and reduced litter quality, which are recognised as influential concerning microbial metabolism (Qu et al., 2020; Guo et al., 2024). Fungal communities were similarly affected by eucalyptus plantations and showed a lower Shannon diversity of fungi than both Native forests and Pine plantations. In fact, eucalyptus plantations showed a clear association with ericoid mycorrhizal fungi that is probably related to the availability of Ericaceae in the understory of these plantations, where high light availability usually favours understory development, in contrast to deciduous forests. These species also tend to exhibit myrmecochorous dispersal strategies, which were more prevalent in eucalyptus plantations, further contributing to the differentiation of their functional composition (Calvi˜ no-Cancela.,2013). In addition, previous studies in the Iberian Peninsula have shown that Eucalyptus plantations support lower ectomycorrhizal diversity compared to native forests and often harbour introduced fungal taxa from the tree’s native range (Díez, 2005; Santolamazza-Carbone et al., 2019). These introduced taxa may have contributed to the distinct fungal community composition observed in Eucalyptus plantations. Beyond the patterns of taxonomic and functional differentiation, the observed traits and microbial functions in Eucalyptus plantations indicate disruptions in key ecological processes. One of the most evident is the alteration of nutrient cycling, particularly the nitrogen cycle. In this context, Eucalyptus plantations produce nutrient-poor litter rich in polyphenols and lignified compounds, which inhibit microbial activity and suppress nitrification, thereby reducing soil nitrate availability and impairing nitrogen cycling (Zancada et al., 2003; Castro-Díez et al., 2021; Mallen-Cooper et al., 2022). In parallel, the prevalence of early-successional plant traits such as anemochory, myrmecochory, and low shade tolerance suggests a limited capacity for structural development and functional maturation, potentially constraining successional progression and forest regeneration (Calvi˜ no-Cancela et al., 2012). Moreover, the underrepresentation of forest specialist birds and lower bird diversity and richness may further limit biotic interactions such as seed dispersal, which are essential for ecosystem feedbacks and plant recruitment (García-Fern´ andez et al., 2025; Calvi˜ no-Cancela.,2013). Importantly, such disruptions are likely driven not only by the intrinsic biological traits of Eucalyptus plantations, but also by the intensive silvicultural practices commonly associated with its cultivation and harvesting. Finally, it is important to acknowledge that Eucalyptus plantations were established primarily for timber production, not conservation. However, in multifunctional landscapes such as biosphere reserves, understanding their ecological role remains relevant. In this context, our results indicate that these plantations contribute little to biodiversity conservation, highlighting the need to carefully balance production goals with ecological values in land-use planning. In contrast, Pine plantations exhibited an intermediate position across a considerable proportion of the obtained results. Despite a decline in plant richness and diversity, a decline in bird species richness and a different species composition when compared to both Native and Restored systems, pine plantations did not differ significantly in bird or microbial diversity when contrasted with native and restored forests. This pattern may be partly explained by the higher presence of native U. Ortega-Barrueta et al. Forest Ecology and Management 596 (2025) 123057 9