Hardwood riparian forests in northern Iberian Peninsula: classification and diversity patterns

Idoia Biurrun; Jokin Belmonte; Irati Sanz-Zubizarreta; Juan Antonio Campos

doi:10.3897/VCS.145406

Abstract

Aims: Temperate hardwood riparian forests constitute an endangered habitat occurring mostly on the floodplains of major European rivers, and included in habitat T13 of the European Nature Information System (EUNIS). In this study we aim to 1) confirm their presence in small rivers of northern Spain, 2) revise their classification at association level, and discuss it at alliance level, 3) evaluate their diversity patterns and ecological preferences. Study area: Inner temperate and submediterranean valleys in northern Spain. Methods: We sampled 72 nested series of vegetation plots with grain sizes ranging from 0.0001 m² to 200 m², where the presence of vascular plants was recorded in all grain sizes and their cover estimated in 100 m². 200 m² plots were submitted to the EUNIS Expert System (EUNIS-ESy) and to agglomerative hierarchical clustering. Non-metric dimensional scaling was used for their ordination in the multivariate space. Diagnostic species were assessed by means of modified phi index and differences among forest types regarding their diversity and ecological preferences through analyses of variance. Results: The EUNIS-ESy classified five plots in habitat T13, 19 were classified in other deciduous forest types, and 48 were classified only at level 1 (forests) or remained unclassified. Three main clusters emerging from the agglomerative clustering corresponded to three associations from the alliance Alnion incanae: Carici pendulae-Fraxinetum excelsioris, Galio laevigati-Fraxinetum excelsioris, and Viburno lantanae-Ulmetum minoris. The mean species richness per plot was 75.3, with the highest values in the western Pyrenean association Galio-Fraxinetum, while Shannon diversity was lowest in the submediterranean Viburno-Ulmetum. Conclusions: These riparian forests provide refuge for several endangered species and might mark the southwestern distribution of hardwood riparian forests in Europe. It is therefore crucial to clarify their classification at alliance level, for which a comprehensive analysis of European temperate riparian, alluvial and mesic oak forests is required.

Taxonomic reference: Euro+Med PlantBase (Euro+Med 2024).

Syntaxonomic reference: Mucina et al. (2016), updated in Chytrý et al. (2024), for high rank syntaxa; Biurrun et al. (2016) for associations and suballiances.

Abbreviations: CWM = Community Weighted Means; DIV = Disturbance Indicator Values for European plants; EDGG = Eurasian Dry Grassland Group; EIVE = Ecological Indicator Values for Europe; EUNIS = European Nature Information System; EUNIS-ESy = EUNIS Expert System; NMDS = Non-Metric Multidimensional Scaling; SAR = Species Area Relationship.

Keywords

Alnion incanae, alpha diversity, beta diversity, Fraxino-Quercion roboris, hardwood riparian forest, Iberian Peninsula, nested-plot, northern Spain, species-area relationship, syntaxonomy, z value

Introduction

European temperate hardwood riparian forests are one of the richest woodland types in Europe, with a diverse and structurally complex understory. They are typical of less-frequently flooded, well-aerated mineral soils on floodplains of the middle and lower reaches of major European rivers (Chytrý et al. 2020). The forest canopy is usually tall and multi-layered and combines several tree species, like ashes (Fraxinus excelsior, F. angustifolia), alder (Alnus glutinosa, A. incana), and elms (Ulmus glabra, U. minor, U. laevis), among others (Rodwell and Ssymank 2016). The forest understory includes small trees, shrubs and lianas, as well as a diverse herb layer with an admixture of vernal geophytes and hemicryptophytes. Many of these herbs also occur in mesic oak forests, while some, especially hemicryptophytes, are indicators of wet or moist soils.

These forests, included in the habitat T13 of the European Nature Information System (EUNIS; Chytrý et al. 2020), are categorized as Endangered habitat in the European Red List of Habitats (Janssen et al. 2016). Although Spain was not considered in the evaluation of this habitat, this is, in fact, included in the Spanish list of terrestrial habitats (MITECO 2024). Additionally, maps generated by the EUNIS Expert System (EUNIS-ESy) and provided by Chytrý et al. (2020) assign to habitat T13 many plots from northern Spain, mostly from the Pyrenees and adjacent temperate and submediterranean areas in the Basque Country and Catalonia. Therefore, even though there are no large rivers and floodplains in the temperate part of northern Iberian Peninsula, the presence of this habitat in Spain should be further investigated, as until now the vegetation plots mapped in Chytrý et al. (2020) are the only evidence about it. Actually, the fact sheet corresponding to this habitat produced during the evaluation of the European habitats (Janssen et al. 2016) indicates that although it is especially characteristic of the middle and lower reaches of major European rivers it also occurs throughout Europe as smaller stands in younger river valleys (Rodwell and Ssymank 2016).

Habitat T13 partly overlaps with the alliances Alnion incanae and Alno-Quercion roboris and it completely includes the alliance Fraxino-Quercion roboris (Chytrý et al. 2020). The alliance Alnion incanae has been traditionally recognized to be present in northern Spain (Rivas-Martínez 2011), but Alno-Quercion roboris and Fraxino-Quercion have never been documented (but see Mercadal and Vilar 2013), and are not indicated for Spain in the recent maps of European phytosociological alliances (Preislerová et al. 2022). The lack of documentation of the presence of Fraxino-Quercion and Alno-Quercion roboris in Spain also applies to the suballiance name Ulmenion Oberd. 1953, traditionally used to include the syntaxonomic content of Fraxino-Quercion and considered as part of the Alnion incanae (Mucina et al. 2016). The suballiance Ulmenion separates the elm and ash dominated riparian forests growing on the river floodplains only subject to episodic floodings, mostly in lowlands, from the alder dominated forests growing on the river edge, more frequent in mountains and hills (Wallnöfer et al. 1993). In Spain, the alliance Alnion incanae is found in the Pyrenean valleys and surrounding temperate and submediterranean areas, mostly in Catalonia and Navarre. It also occurs in the Atlantic region, although restricted to the inner subcantabrian valleys, i.e., the upper sections of the big Mediterranean rivers Duero and Ebro and their tributaries in the southern watershed of the Cantabrian Range and Basque-Cantabrian mountains (Biurrun et al. 2016). In the Cantabrian valleys, formed by the streams flowing to the Bay of Biscay, the Alnion incanae is replaced by the alliance Hyperico androsaemi-Alnion glutinosae (Biurrun et al. 2016).

The spatial co-occurrence of the habitat T13 (Chytrý et al. 2020) and the alliance Alnion incanae (Biurrun et al. 2016) in the Pyrenees and inner valleys of the Atlantic region suggests that northern Iberian hardwood riparian forests correspond, at least, to some of the six associations that Biurrun et al. (2016) included in this alliance, for which they described the new suballiance Buxo sempervirentis-Alnenion glutinosae. This suballiance gathers Pyrenean and Iberian alder and ash forests growing in mountain streams on calcareous substrata, under temperate euoceanic sometimes submediterranean climate. The six associations are distributed in the southern foothills of the Cantabrian Range (Euphorbio hybernae-Fraxinetum excelsioris), the subcantabrian (Carici pendulae-Fraxinetum excelsioris) and submediterranean (Viburno lantanae-Ulmetum minoris) valleys of the Basque Country and Navarre (Basque-Cantabrian mountains), the western Pyrenean valleys (Galio laevigati-Fraxinetum excelsioris), the Prepyrenean valleys in Aragón (Lathraeo clandestinae-Populetum nigrae) and the Central and Eastern Pyrenean valleys (Equiseto hyemalis-Alnetum glutinosae).

In this work, we focus on the riparian forests of the inner Basque-Cantabrian and western Pyrenean valleys to assess their current phytosociological classification and confirm their correspondence with EUNIS habitat T13, as well as to obtain further knowledge on their diversity patterns and ecology. With this aim, we sampled 72 new vegetation plots in the same or nearby locations to many plots already used by Biurrun et al. (2016), following a standardized sampling design. Specifically, our aims were 1) to apply the EUNIS-ESy to these new vegetation plots in order to obtain their habitat level classification and eventually confirm their correspondence with habitat T13; 2) to check if the phytosociological classification at association level emerging from a vegetation classification attempt at the scale of the Iberian Peninsula replicates at a regional scale; 3) to compare the diversity patterns and structural and ecologic features of the different hardwood riparian forests occurring in the region.

Study area

The study area includes the inner (subcantabrian and submediterranean) valleys in the Basque-Cantabrian mountains, as well as the westernmost Pyrenean valleys, belonging to the autonomous regions of Basque Country and Navarre. These regions, located in northern Spain, lay on the transition between the temperate and Mediterranean climate, and the whole area shares three biogeographic regions: Atlantic, Alpine and Mediterranean (EEA 2016). The northern part is typically temperate, very oceanic in the coast, with increasing continentality towards the inner subcantabrian valleys and the Pyrenees, the central part is submediterranean, and the southern part is typically Mediterranean (Loidi et al. 2011).

These subcantabrian, submediterranean and western Pyrenean valleys are formed by rivers and streams flowing southwards and eventually joining the Ebro river or one of its tributaries, i.e., all our study area is included in the Ebro basin. From west to east, we sampled riparian forests in the rivers Baia, Zadorra and its tributary Sta. Engracia, Basaburua and Ultzama (Arga basin) and several rivers in Irati basin: Erro, Urrobi, Irati and Zaraitzu (Figure 1). Baia and Zadorra are directly flowing into Ebro river, while Arga and Irati are joining the Pyrenean Aragón river, which is one of the main tributaries of the Ebro. These rivers have their sources in the mountains of the Atlantic-Mediterranean watershed, those from Irati basin in the western Pyrenees and the rest in the Basque-Cantabrian mountains.

Figure 1.

Map of the study area showing the location of the 72 plots of floodplain forests sampled in subcantabrian, western Pyrenean and submediterranean valleys of the Basque Country and Navarre.

The upper and medium sections of these rivers flow under temperate climate and belong to the Atlantic region (subcantabrian valleys) and the Alpine region (western Pyrenean valleys). However, climate in these valleys is drier and more continental than in the Cantabrian valleys with rivers flowing into the Bay of Biscay (Loidi et al. 2011). The lowest sections of most of these rivers opened gorges through the second chain of Basque-Cantabrian Mountains and the Prepyrenees (the so-called transition mountains) to enter into the Mediterranean region, where they flow under a submediterranean climate. These are the submediterranean valleys, the so-called “transition valleys” (Loidi et al. 2011) in the Basque Country and “central zone” in Navarre (Peralta et al. 2018). The largest submediterranean river sections are located in Irati river basin, where there is also an important gradient in its Pyrenean sections: the Erro and Urrobi rivers, the westernmost ones, are transitional between subcantabrian and Pyrenean valleys, while Zaraitzu river, the easternmost one, is a typical Pyrenean river where spring snow melting has an important role in the water regime. Snow is not an important component in the flowing regime of Atlantic rivers.

The landscape of these valleys is quite diverse, with different forest types as natural vegetation and meadows, scrubs and dry grasslands as semi-natural vegetation. Beech forests are spread in river sources and, depending on the river basin, several forest types can be found in the slopes and valleys. In the subcantabrian valleys (Baia, Zadorra, Sta. Engracia, Basaburua, Ultzama), mesic oak forests with Quercus robur occupy the valley bottoms, as well as gentle slopes in the wettest areas, while slopes are occupied by thermophilous deciduous forests (Quercus faginea, Q. pubescens, Q. pyrenaica) (Loidi et al. 2011). In the Pyrenean valleys (upper section of Irati basin) the riparian forest directly contacts with the Q. pubescens forests, as Q. robur is not occurring in the Pyrenean chain. Exceptionally, Irati and Urrobi rivers opened deep gorges in siliceous outcrops of western Pyrenees, and there the riparian forest contacts with Q. petraea forests. In the submediterranean valleys (lowest section of Irati basin) the riparian forests are neighbored by Q. faginea forests, although, given the intense agricultural use in these valleys, cereal and other crops have become the main contact of riparian forests. Hills and slopes, extensively grazed by sheep, horses, and cattle in the past, are now mostly abandoned (Peralta et al. 2018). In contrast, in the northern sections, the land use is mostly cattle breeding in the valleys and forestry in the mountain slopes.

The riparian vegetation is quite complex in the submediterranean valleys, with slow flowing sections rich in aquatic communities and river deposits colonized by tall-herb communities with Epilobium hirsutum, Mentha longifolia, Eupatorium cannabinum, etc., as well as willow communities with Salix purpurea. It is also common that a softwood forest with Salix alba, Alnus glutinosa and Populus nigra develops between the riverbank and the riparian hardwood forest (Peralta et al. 2018). This flooded forest corresponds to the association Humulo lupuli-Alnetum glutinosae, described by Biurrun et al. (1994) and included by Biurrun et al. (2016) in the alliance Populion albae. In the subcantabrian valleys there are also some slow flowing sections, but the riparian vegetation complex is not so developed. Tall-herb communities are more fragmentary, and the alder forest growing in the most frequently flooded banks occupies a very narrow space between the hardwood forest and the riverbank, often formed by one line of alder trees, which is absent in the smallest rivers, such as Basaburua river. The western Pyrenean rivers of Irati basin are more torrential, and it is almost impossible to distinguish two different riparian forests in the same river section. As a consequence, alder is a typical component of the hardwood forest, becoming abundant in the topographically most favorable riverbanks. However, alder is totally absent in Zaraitzu river, the most torrential one in the basin, and almost absent from Erro river, the smallest one and with a very low summer flow.

Methods

Vegetation data

The target riparian forests are well represented in the Vegetation-Plot Database of the University of the Basque Country (BIOVEG; EU-00-011, Biurrun et al. 2012), as well as in SIVIM Floodplain forests (EU-00-24; Biurrun and Font 2020), with several hundred plots also available in EVA (Chytrý et al. 2016) and sPlot (Bruelheide et al. 2019) databases. Nevertheless, we conducted a new standardized sampling in order to get a homogeneous dataset formed by high quality plots sharing identical sampling method. The sampling took place mostly in forest patches with historic plots already present in the mentioned databases. These historic plots were used both in the Iberian syntaxonomical revision of floodplain forests (Biurrun et al. 2016) and the EUNIS habitat classification (Chytrý et al. 2020).

We sampled 72 plots distributed across the study area, 18 in subcantabrian river sections (Baia, Zadorra, Sta. Engracia, Basaburua, Ultzama), 24 in the western Pyrenean sections and 30 in submediterranean sections of Irati basin. Irati basin is the largest and has the largest elevation gradient regarding plot location, from 404 m a.s.l. in the lowest section to 875 m a.s.l. in the upper section of Urrobi river. This elevation gradient is much smaller along the subcantabrian valleys, from 513 m a.s.l. in Zadorra river (near the Basque capital Vitoria-Gasteiz) to 606 m a.s.l. in the highest plot in Baia river.

We adapted to nested sampling in forests the standard EDGG sampling method (Dengler et al. 2016, 2021), which was originally created for the nested sampling of grasslands and other open vegetation types. Due to the specific linear shape of floodplain forests, the biggest plots of the nested series were not squares, but rectangles of 10 m × 20 m. It was not possible to set plots of 20 m × 20 m as many forest patches were not wide enough. Exceptionally, plots were 8 m wide and 25 m long. All 72 plots were set as permanent plots with four red-painted wooden stakes at each corner, along with their respective geographical coordinates. In our adaptation to the forest sampling, we set two contiguous 10 m × 10 m plots parallel to the riverbed trying to include homogeneous forest stands. The herb layer was sampled following the nested sampling design in two series, with subplots replicated once in each. The nested series in the downstream 100 m² started in the corner near the river, while the upstream series was located in the inner corner (Figure 2). This nested sampling included all herbaceous plant species as well as woody species seedlings and saplings up to 0.7 m high (Staubli et al. 2021). In the smallest subplots (0.0001 to 10 m²) species presence only was recorded. Cover (in %) of species in the herb layer was recorded in the two 100 m² plots. The tree and shrub layers were sampled in the 200 m² plot (i.e., the two 100 m² plots together), where species cover was also estimated as percentage cover. For the multivariate analyses we used floristic data of 200 m² plots, so we calculated the mean cover of the herb-layer species in both 100 m² plots and combined this table with the data of the tree and shrub layers sampled in the 200 m² plots, in order to get a complete composition table at the grain size of 200 m².

We estimated the cover of each vegetation layer (tree, shrub, herb and cryptogam) as well as cover of litter, dead wood and fractions of abiotic soil surface in each 200 m² plot. Measured parameters and other plot level data are available in Suppl. material 1. The main sampling took place in summer of the years 2022 and 2023, exceptionally in 2024, but we conducted an additional sampling in spring 2023 and 2024 not to dismiss the vernal geophytes. During these spring samplings we recorded the presence and cover in the 200 m² plot of vernal species that had not been not found in the previous sampling.

Figure 2.

Adaptation of the EDGG nested-plot design (Dengler et al. 2016) for the nested sampling of the herb layer of riparian forests.

Diversity variables

We calculated the species richness and the Shannon diversity index for each 200 m² plot and used these two variables as measures of alpha diversity. We also calculated the species richness of the herb layer for all subplot sizes and, additionally, we calculated Whittaker’s beta diversity for each forest type, as the quotient between its gamma diversity and the mean richness at plot level (Whittaker 1960). As a fine-scale beta diversity measure we used the slope parameter of the Species Area Relationship (SAR, Jurasinski et al. 2009). This slope parameter, the z value, represents the exponent of the power law function of the SAR (Dembicz et al. 2021a), which has been proposed as the best fitting function among the numerous proposed SAR functions, both in non-nested (as islands, Matthews et al. 2016) and nested (Dengler et al. 2020) sampling units. This z value is widely used to compare spatial species turnover between different ecological situations (Drakare et al. 2006; de Bello et al. 2007; Dembicz et al. 2021a) or community types (Dembicz et al. 2021b). As z value only exhibits a small amount of scale dependence (Zhang et al. 2021), we assumed it to be constant within our range of grain sizes. We calculated the z value for each 200 m² plot using the means of the two subplots of each size from 0.0001 m² to 100 m². All diversity measures at plot level are available in Suppl. material 1.

Ecological and disturbance indicator values

As proxies of the local-scale site conditions of the plot locations, we selected the Ecological Indicator Values for Europe (EIVE; Dengler et al. 2023) and the Disturbance Indicator Values for European plants (DIV; Midolo et al. 2023). As shown by Scherrer and Guisan (2019), mean ecological indicator values can improve the predictive power of models much compared to using only approximate site conditions from modelled databases via the coordinates. We calculated the plot-wise means for EIVE and DIV using the data of the entire community. The means of EIVE values were based on presence/absence data (unweighted), while those of DIV values were based on abundance data (community weighted means, CWM). Both calculations were performed with the “cwm” function from the R package weimea (Zelený 2020). The values of these indices in each plot are available in Suppl. material 1.

Climatic data

We retrieved climatic data from WorldClim at 30 arc sec resolution (Fick and Hijmans 2017). We selected five climatic variables from WorldClim which have shown to be meaningful in the context of the SW European forests (Biurrun et al. 2016; González-García et al. 2024): BIO1 (Annual mean temperature), BIO7 (Temperature annual range), BIO10 (Mean temperature of the warmest quarter), BIO12 (Annual precipitation), BIO18 (Precipitation of the warmest quarter) and additionally we calculated the ombrothermic index of the warmest quarter, i.e., the index Ios₃ of Rivas-Martínez, which is used to delimit regions under Mediterranean climate (Rivas-Martínez et al. 2011). This index is calculated dividing the sum of the precipitation of the warmest quarter by the sum of the mean monthly temperatures of the warmest quarter (BIO18/∑t_{summer quarter}). Generally speaking, Ios₃ < 2 indicates Mediterranean climate. The values of climatic variables for each plot are available in Suppl. material 1.

Statistical analyses

Expert system for automatic classification to EUNIS habitats

We used the composition data of the 200 m² grain size, where records of woody species in different layers were merged using the algorithm by Fischer (2015). The resulting matrix was formed by 72 plots and 456 taxa. We applied the EUNIS-ESy to this matrix in order to classify our plots to habitat types of the EUNIS Habitat classification using the R implementation of the EUNIS-ESy (Bruelheide et al. 2021). We used the latest version of the expert system (Chytrý et al. 2021).

Unsupervised classification, ordination and diagnostic species

We started with the same floristic matrix used for the EUNIS habitat classification, formed by 72 plots and 456 taxa. To reduce noise, we removed species that occurred only in one plot before conducting multivariate analyses (McCune and Grace 2002). This led to the removal of 116 taxa. No taxa were removed in 14 plots, and the rest of plots had few species removed (mostly 1–2, a few 4–8). The most extreme case was one plot of submediterranean riparian forest located near a campsite, with 13 singletons, all of them ruderal species. Bray-Curtis dissimilarity was calculated on the square-root transformed cover of species in plots, and the resulting distance matrix was used both for the classification and ordination analyses. Bray-Curtis dissimilarity is widely used in vegetation studies due to the relatively equal weighting it gives to both dominant and rare species in analyses (Bray and Curtis 1957), and is considered a robust measure of ecological distance (Faith et al. 1987).

We classified plots by means of agglomerative hierarchical clustering using the Beta flexible linkage method (with β = -0.25) in cluster package (R Core Team 2024). The structure of the resulting dendrogram was used to decide its partition in clusters, supported by the current classification of these forests in the study area. To test significant differences in composition and structure between clusters, we conducted permutational multivariate analysis of variance (PERMANOVA) of these groups using the “adonis2” function in the vegan R package with the calculated Bray-Curtis dissimilarity matrix and 999 permutations (Oksanen et al. 2024).

We determined diagnostic species of the resulting clusters using the phi coefficient of association (Chytrý et al. 2002) standardized to equal plot number per cluster (Tichý and Chytrý 2006) in the Ginkgo program of the VEGANA package (De Cáceres et al. 2007). Species with phi ≥ 0.3 were considered diagnostic, and those with phi ≥ 0.5 highly diagnostic. Additionally, for species to be considered diagnostic, we also required that their constancy in the target cluster was at least 20% and its phi-value at least 0.25 higher than in the cluster with the next-higher phi-value, as in García-Mijangos et al. (2021). Taxa identified only at genus level were removed from the list of diagnostic species. Some species that were highly diagnostic for one cluster and diagnostic for another are indicated as such, provided their phi value as highly diagnostic is ≥ 0.25 than their phi value as diagnostic.

To visualize the gradient of vascular plant species composition across the vegetation types, we visualized the clusters in multivariate space using non-metric multidimensional scaling (NMDS; McCune and Grace 2002) ordinations of our species matrix. We created all NMDS ordinations with the R package vegan function “metaMDS” using Bray-Curtis dissimilarity and 100 random starts (Oksanen et al. 2024). To assess which climatic gradients and structural parameters were correlated with the results of the NMDS ordination, we projected them onto our ordination using the R package vegan function “envfit” with 100 permutations (Oksanen et al. 2024). We projected 15 variables (6 climatic, 8 structural and elevation) but only significant variables (p-value < 0.05) were plotted: litter cover, gravel cover, stones and rocks cover, cryptogam cover and BIO7 (temperature annual range).

Analyses of differences between forest types

Differences among forest types regarding diversity values and structural and bioclimatic variables were analyzed by means of analyses of variance (ANOVAs) in the R programming language (R Core Team 2024). Tukey’s post-hoc test was applied following a significant ANOVA (p-value < 0.05). For some variables not meeting the assumption of normal distribution we performed the non-parametric Kruskal-Wallis test, as well as the post-hoc Dunn test with the Bonferroni correction for significant pairs. To reduce the pitfalls of using CWMs in vegetation analyses, which often provide overly optimistic results (Zelený and Schaffers 2012; Zelený 2018), the differences of EIVE and DIV values among the forest types were analyzed using modified permutation tests with 999 permutations in PRIMER 7 (Clarke et al. 2014).

Results

General floristic results

We identified 456 taxa across all vegetation plots, with only 48 taxa occurring in at least half of the plots (≥ 36), and three taxa, Acer campestre, Crataegus monogyna, and Hedera hibernica present in all plots. Among the other most frequent species, the five most common species in the herb layer were Alliaria petiolata, Brachypodium sylvaticum, Elymus caninus, Geum urbanum, and Viola reichenbachiana, and the five most common woody species Clematis vitalba, Cornus sanguinea, Euonymus europaeus, Ligustrum vulgare, and Rubus caesius.

Five of the species found were endangered species included in regional catalogues (Boletín Oficial del País Vasco 2011 and Boletín Oficial de Navarra 2023): Carex strigosa, Circaea alpina, Galanthus nivalis, Lathraea squamaria, and Narcissus pseudonarcissus subsp. nobilis. Galanthus nivalis is also listed, as Near Threatened, in the European red list of vascular plants (European Commission 2011), and Narcissus pseudonarcissus subsp. nobilis in the annex II of the EU Habitats Directive (European Union 1992) and the Spanish index of threatened species (Boletín Oficial del Estado 2011). In addition to threatened species, the sampled floodplain forests hosted several Cantabrian-Pyrenean endemics: Conopodium pyrenaeum, Crepis lampsanoides, Fritillaria pyrenaica, Helleborus viridis subsp. occidentalis, Meconopsis cambrica, Myosotis martini, Saxifraga hirsuta, Scilla lilio-hyacinthus, Scrophularia alpestris, Valeriana pyrenaica, and Veronica ponae. Other outstanding species, common in Central European forests and high mountains, but rare in the lowland forests of northern Spain, were Aconitum napellus subsp. vulgare, A. lycoctonum subsp. neapolitanum, Adenostyles alliariae, Anemone ranunculoides, and Valeriana officinalis. Pictures of some of these threatened and endemic species are provided in Figures 3, 4.

Mean species richness per 200 m² plot was 75.3, with a minimum of 39 plant species in a submediterranean ash forest dominated by Fraxinus angustifolia, with admixture of Salix alba, Alnus glutinosa, and Ulmus minor in the tree layer (IMB10), and a maximum of 124 species in a western Pyrenean ash forest dominated by Fraxinus excelsior with Corylus avellana, Alnus glutinosa, Fagus sylvatica, and Ulmus glabra (IE03).

Figure 3.

Photos of three taxa included in the red lists of the Basque Country and Navarre: A, Narcissus pseudonarcissus subsp. nobilis; B, Lathraea squamaria; C, Galanthus nivalis. Photos A and C by J. Belmonte, B by I. Biurrun.

Figure 4.

Photos of three species endemic to the Cantabrian-Pyrenean region: A, Fritillaria pyrenaica; B, Meconopsis cambrica; C, Scilla lilio-hyacinthus. Photos A and C by J. Belmonte, B by I. Biurrun.

Alien species (including archeophytes) were few, 28 species (percentage in the species pool: 6.1%), and none of them were present in more than half of the plots. The most frequent were Juglans regia (27 plots), Populus nigra var. italica (12) and Populus ×canadensis (6), with the following taxa occurring from one to three plots: Acer platanoides, Aesculus hippocastanum, Arundo donax, Asparagus officinalis, Avena sativa, Erigeron sumatrensis, Ficus carica, Fraxinus ornus, Helianthus tuberosus, Helianthus ×laetiflorus, Juglans nigra, Ligustrum ovalifolium, Malus pumila, Oxalis sp., Picea abies, Populus alba, Prunus armeniaca, P. laurocerasus, Quercus rubra, Solanum tuberosum, Triticum aestivum, Veronica persica, Vicia faba, Xanthium sp., and Yucca gloriosa. The mean invasion level at plot level was 1.6, with 26 plots without any alien species, and the highest invasion level (6.8) in a submediterranean plot in Irati river.

Expert system

Only five plots were classified by EUNIS-ESy in habitat T13 (Temperate hardwood riparian forest), nine were classified to habitat T1E (Carpinus and Quercus mesic deciduous forest), four to habitat T12 (Alnus glutinosa-Alnus incana forest on riparian and mineral soils), three to T1F (Ravine forest), two to T17 (Fagus forest on non-acid soils), and one to T1H (Broadleaved deciduous plantation of non-site-native trees). 39 plots were only classified at EUNIS level 1 as T (Forests and other wooded land), six plots had multiple assignments at the same hierarchical level, and three did not get a valid assignment to any vegetation type. Among the plots classified in T13, four were submediterranean forests of Viburno-Ulmetum minoris (in Irati and Zaraitzu rivers), and one was a subcantabrian forest of Carici-Fraxinetum excelsioris in Ultzama river.

Floristic differentiation in riparian forests

The agglomerative clustering yielded a dendrogram in which we distinguished three riparian forest types (PERMANOVA p < 0.001, R² = 0.30) (Figure 5). We interpreted these groups as the three associations of the alliance Alnion incanae recognized by Biurrun et al. (2016) in the study area: Carici pendulae-Fraxinetum excelsioris (subcantabrian riparian forest), Viburno lantanae-Ulmetum minoris (submediterranean riparian forest), and Galio laevigati-Fraxinetum excelsioris (western Pyrenean riparian forest). In all other results, tables and figures we refer to these three associations. A shortened synoptic table showing the diagnostic species of the three forest types is shown in Table 1, while the full synoptic table is provided in Suppl. material 2.

The western Pyrenean association Galio laevigati-Fraxinetum excelsioris included 60 diagnostic species, 16 of them highly diagnostic. The species with highest diagnostic value were common species in mesic forests, like Melica uniflora, Veronica chamaedrys, Fagus sylvatica and Moehringia trinervia. This association also included some diagnostic species that have a Pyrenean distribution in Spain such as Acer opalus, Cardamine heptaphylla, Galium laevigatum, Geranium phaeum, and Pulmonaria affinis. The other two associations included far fewer diagnostic species. Viburno lantanae-Ulmetum minoris had 23 diagnostic species, only six of them highly diagnostic. The most important was Fraxinus angustifolia, which only occurs in this submediterranean association within Iberian communities belonging to the Alnion incanae. Ulmus minor was also highly diagnostic, as well as some ruderal species as Lapsana communis, Melissa officinalis, and Torilis arvensis. The subcantabrian riparian forest of the association Carici pendulae-Fraxinetum excelsioris had 20 diagnostic species, nine of them highly diagnostic. Quercus robur stands out, as it is absent from both Pyrenean and submediterranean valleys. It is also remarkable the high diagnostic value of several geophytes and hemicryptophytes typical from wet grasslands and forests, such as Filipendula ulmaria, Narcissus pseudonarcissus subsp. nobilis, Ornithogalum pyrenaicum, Aconitum napellus subsp. vulgare, and A. lycoctonum subsp. neapolitanum, some of them indicated as diagnostic species of the alliance Alnion incanae by Biurrun et al. (2016). Fraxinus excelsior separates the subcantabrian and western Pyrenean forests from the submediterranean one, as it was moderately diagnostic in Carici-Fraxinetum and highly diagnostic in Galio-Fraxinetum. Both associations also shared other nemoral species such as Conopodium pyrenaeum, Helleborus viridis subsp. occidentalis, Isopyrum thalictroides, and Stellaria holostea, as well as Primula elatior, typical species of riparian forests. Pictures of these forests are provided in Figures 6, 7.

Table 1.

Download as

CSV

XLSX

Shortened synoptic table of the three associations of hardwood riparian forests in the study area. VU: Viburno lantanae-Ulmetum minoris, CF: Carici pendulae-Fraxinetum excelsioris, GF: Galio laevigati-Fraxinetum excelsioris. Values are percentage frequencies, grey shading indicates phi ≥ 0.3 and dark grey ≥ 0.5. Species are arranged according to decreasing phi value. Only species with relative frequency > 25% in at least one association are shown. The full synoptic table is available in Suppl. material 2.

Association	VU	CF	GF
Number of plots	30	19	23
Fraxinus angustifolia	100	11	.
Melissa officinalis	73	5	13
Ulmus minor	77	21	9
Juglans regia	73	21	4
Torilis arvensis	67	.	22
Lapsana communis	87	32	43
Calystegia sepium	43	.	4
Quercus faginea	50	11	4
Frangula alnus	37	.	.
Sonchus oleraceus	40	.	4
Anisantha sterilis	30	.	.
Salix alba	30	.	.
Lythrum salicaria	37	11	.
Vitis vinifera	47	.	22
Ligustrum vulgare	97	84	57
Anagallis arvensis	23	.	.
Prunus domestica	67	32	30
Rubia peregrina	57	5	39
Rosa canina aggr.	83	42	65
Solanum dulcamara	27	5	4
Fraxinus excelsior	7	84	100
Quercus robur	.	74	.
Narcissus pseudonarcissus subsp. nobilis	.	68	.
Sambucus nigra	37	100	30
Filipendula ulmaria	3	84	39
Ornithogalum pyrenaicum	.	53	.
Lamium galeobdolon	3	84	61
Glechoma hederacea	23	74	13
Cardamine pratensis	.	53	13
Crataegus laevigata	.	42	.
Aconitum lycoctonum subsp. neapolitanum	.	32	.
Aconitum napellus subsp. vulgare	.	32	.
Arum italicum	50	84	39
Ficaria verna	60	84	26
Mercurialis perennis	.	47	26
Fritillaria pyrenaica	.	26	.
Lamium maculatum	17	58	26
Scilla lilio-hyacinthus	.	42	22
Lilium martagon	.	32	9
Quercus faginea × robur	7	32	.
Melica uniflora	23	53	100
Veronica chamaedrys	17	47	96
Fagus sylvatica	7	37	87
Moehringia trinervia	3	32	83
Fragaria vesca	.	.	61
Buxus sempervirens	73	.	100
Poa nemoralis	47	42	100
Torilis japonica	17	26	78
Tilia platyphyllos	37	21	87
Bromopsis ramosa	50	53	100
Hepatica nobilis	10	42	78
Geranium phaeum	3	16	61
Silene dioica	.	63	78
Silene vulgaris	13	.	57
Knautia arvernensis	.	16	52
Malus sylvestris	3	.	43
Pulmonaria affinis	.	.	39
Potentilla sterilis	.	37	61
Carex sylvatica	43	53	91
Acer opalus	10	.	48
Ulmus glabra	10	16	57
Galium laevigatum	.	.	35
Crepis lampsanoides	7	32	61
Luzula sylvatica	3	21	52
Quercus pubescens	13	.	48
Hieracium murorum	3	5	39
Lathyrus laevigatus	.	.	30
Cerastium fontanum	3	.	35
Milium effusum	7	16	48
Trifolium repens	17	.	48
Saxifraga hirsuta	.	21	43
Clinopodium vulgare	10	.	39
Cruciata laevipes	50	37	83
Vicia sepium	17	42	65
Solidago virgaurea	.	.	26
Veronica montana	40	58	83
Trifolium pratense	37	5	61
Campanula trachelium	37	16	65
Ranunculus acris	60	16	78
Viburnum lantana	67	37	87
Anemone nemorosa	17	5	43
Carex divulsa	17	5	43
Carex flacca	13	11	43
Festuca rubra aggr.	3	.	26
Veronica ponae	3	.	26
Aquilegia vulgaris	10	11	39
Viburnum opulus	13	32	52
Euphorbia amygdaloides	77	63	96
Equisetum arvense	33	21	61
Deschampsia cespitosa	57	32	78
Saponaria officinalis	10	.	30
Primula veris	7	21	39
Myosotis martini	.	11	26
Stachys sylvatica	27	26	57
Other diagnostic species of the class Alno-Populetea:
Acer campestre	100	100	100
Hedera hibernica	100	100	100
Brachypodium sylvaticum	100	95	100
Alliaria petiolata	90	95	100
Geum urbanum	87	100	100
Elymus caninus	97	74	100
Rubus caesius	100	79	91
Clematis vitalba	93	63	100
Galium aparine	93	95	43
Dioscorea communis	70	74	48
Schedonorus giganteus	47	63	65
Carex pendula	60	47	30
Heracleum sphondylium	47	53	39
Urtica dioica	37	58	39
Alnus glutinosa	47	32	39
Circaea lutetiana	30	26	57
Symphytum tuberosum	20	53	35
Fraxinus angustifolia × excelsior	30	47	4
Angelica sylvestris	23	11	35
Carex remota	13	32	30
Conopodium pyrenaeum	3	95	78
Primula elatior	.	63	65
Isopyrum thalictroides	.	53	43
Chaerophyllum temulum	30	.	26
Iris foetidissima	30	32	.
Allium ursinum	.	42	26
Populus nigra	30	5	9
Clinopodium menthifolium	27	.	13
Humulus lupulus	17	26	.
Other species:
Crataegus monogyna	100	100	100
Cornus sanguinea	100	84	100
Viola reichenbachiana	80	100	100
Euonymus europaeus	70	100	96
Geranium robertianum	77	95	74
Corylus avellana	53	95	100
Rosa arvensis	43	95	96
Taraxacum sp.	90	37	83
Rubus ulmifolius	77	53	57
Lonicera xylosteum	57	58	74
Viola hirta	67	53	57
Anthriscus sylvestris	40	58	61
Dactylis glomerata	60	32	70
Arctium minus	67	32	48
Prunus spinosa	33	58	70
Helleborus viridis	3	84	83
Prunella vulgaris	70	11	57
Stellaria holostea	7	79	74
Agrostis stolonifera	53	32	43
Rhamnus cathartica	47	42	39
Lonicera periclymenum	37	47	30
Ruscus aculeatus	27	58	35
Acer pseudoplatanus	30	53	30
Ajuga reptans	10	58	52
Helleborus foetidus	40	11	52
Hypericum perforatum	43	16	43
Orobanche hederae	50	16	35
Ranunculus repens	30	26	39
Lathraea clandestina	40	21	9
Melilotus albus	30	.	39
Picris hieracioides	27	.	43
Arrhenatherum elatius	27	.	39
Galeopsis tetrahit	7	32	39
Eupatorium cannabinum	20	11	35
Hesperis matronalis	3	37	35
Medicago sp.	30	.	30
Plantago major	30	.	26
Poa trivialis	20	16	26
Allium oleraceum	30	.	17
Veronica hederifolia	17	32	9
Chaerophyllum aureum	17	.	30
Cardamine flexuosa	3	21	26
Hippocrepis emerus	13	.	30
Ilex aquifolium	.	21	30
Oxalis acetosella	3	26	22
Populus pyramidalis	27	.	13
Brachypodium rupestre	3	16	26

Figure 5.

Dendrogram with the three forest types produced by the agglomerative hierarchical clustering of 72 plots (above). The three groups were supported by PERMANOVA (p < 0.001, R² = 0.30). NMDS ordination diagram with the structural and climatic variables significantly correlated with the ordination plotted (below).

Figure 6.

Vernal aspect of the ground layer in a forest stand of the Carici pendulae-Fraxinetum excelsioris (Baia river). Photo: J. Belmonte.

Figure 7.

Late-spring view from outside of a stand belonging to the Viburno lantanae-Ulmetum minoris (Irati river). Photo: I. Biurrun.

Diversity patterns

Submediterranean forests of the Viburno-Ulmetum minoris had the highest Whittaker’s beta diversity (4.7), which was considerably lower in the Fraxinus excelsior-dominated forests, with lowest values in the subcantabrian forests of the Carici-Fraxinetum excelsioris (3.2) (Table 2). Carici-Fraxinetum and Viburno-Ulmetum shared the same fine-grain within plot beta diversity in the herb layer, measured as the mean z value of the nested plots for each forest type. This value was highest in the western Pyrenean forests of the Galio-Fraxinetum excelsioris, but differences were not significant after the post-hoc Dunn test with Bonferroni correction (Figure 8).

Regarding plot-level alpha diversity, species richness and Shannon diversity significantly differed among the three forest types (Table 2). Stands of the Galio-Fraxinetum contained significantly more species of vascular plants than stands of the Carici-Fraxinetum and the Viburno-Ulmetum, while Shannon diversity was significantly lower in Viburno-Ulmetum than in Galio-Fraxinetum and Carici-Fraxinetum (Figure 8).

Table 3 shows the species richness of the herb layer across grain sizes, with significant differences for sizes ≥ 0.1 m². In all sizes from 0.1 m² to 200 m² richness was highest in Galio-Fraxinetum and lowest in Viburno-Ulmetum minoris. The higher species richness of Galio-Fraxinetum was especially significant for the biggest sizes, from 10 to 200 m², while for 0.1 and 1 m² these differences were only significant when compared to the richness of Viburno-Ulmetum.

Table 2.

Download as

CSV

XLSX

Mean values of diversity indices, structural and climatic variables, and ecological and disturbance indicator values in the three forest types. GF: Galio laevigati-Fraxinetum excelsioris, CF: Carici pendulae-Fraxinetum excelsioris, VU: Viburno lantanae-Ulmetum minoris. The column “Test” indicates the analysis applied to test differences among forest types, and the column “Stat.” the value of the statistics of the ANOVA, Kruskal-Wallis or Permutation tests. Different letters in mean values indicate significant between-pair differences after significant ANOVA, Kruskal-Wallis or Permutation tests.

	GF n = 23	CF n = 19	VU n = 30	Test	Stat.	p-value
Gamma diversity	308	207	319
Whittaker’s β	3.37	3.10	4.68
Species richness	91.8 ± 17.3 (a)	66.8 ± 20.3 (b)	68.2 ± 15.9 (b)	Kruskal-Wallis	21.02	< 0.001
Shannon diversity index	2.54 ± 0.22 (a)	2.63 ± 0.15 (a)	2.33 ± 0.23 (b)	ANOVA	25.18	< 0.001
z value	0.32 ± 0.05 (a)	0.29 ± 0.05 (a)	0.29 ± 0.05 (a)	Kruskal-Wallis	6.22	0.04
Elevation (m a.s.l.)	671 ± 55 (b)	568 ± 80 (a)	460 ± 38 (c)	Kruskal-Wallis	56.56	< 0.001
Tree layer cover (%)	84 ± 13	88 ± 9	87 ± 10	Kruskal-Wallis	0.53	0.769
Shrub layer cover (%)	52 ± 22	50 ± 16	47 ± 21	ANOVA	0.48	0.619
Herb layer cover (%)	74 ± 17	83 ± 13	76 ± 16	Kruskal-Wallis	4.53	0.104
Cryptogam layer cover (%)	25 ± 23 (b)	3 ± 5 (a)	5 ± 8 (a)	Kruskal-Wallis	18.22	< 0.001
Litter cover (%)	32 ± 17 (b)	54 ± 20 (a)	43 ± 22 (ab)	Kruskal-Wallis	10.01	0.006
Dead wood cover (%)	9 ± 5 (b)	6 ± 5 (a)	10 ± 6 (b)	Kruskal-Wallis	8.35	0.01
Stones and rocks cover (%)	11 ± 20 (b)	1 ± 2 (a)	4 ± 9 (a)	Kruskal-Wallis	16.80	< 0.001
Gravel cover (%)	7 ± 10 (b)	0 ± 0 (a)	3 ± 9 (a)	Kruskal-Wallis	17.55	< 0.001
BIO1 (°C)	11.0 ± 0.5 (a)	11.2 ± 0.4 (a)	12.5 ± 0.4 (b)	Kruskal-Wallis	50.92	< 0.001
BIO12 (mm)	930 ± 61 (a)	928 ± 89 (a)	773 ± 56 (b)	ANOVA	46.37	< 0.001
BIO7 (°C)	24.3 ± 0.5 (a)	23.7 ± 0.6 (a)	26.0 ± 0.5 (b)	Kruskal-Wallis	51.17	< 0.001
BIO10 (°C)	18.1 ± 0.6 (a)	17.9 ± 0.5 (a)	20.0 ± 0.5 (b)	Kruskal-Wallis	50.57	< 0.001
BIO18 (mm)	158 ± 11 (a)	158 ± 19 (a)	128 ± 8 (b)	Kruskal-Wallis	48.14	< 0.001
Ios₃	2.92 ± 3.04 (a)	2.94 ± 3.71 (a)	2.14 ± 1.90 (b)	Kruskal-Wallis	49.24	< 0.001
Disturbance frequency	0.13 ± 0.04 (a)	0.12 ± 0.03 (a)	0.17 ± 0.04 (b)	Permutation	15.40	< 0.001
Disturbance frequency herb layer	1.35 ± 0.02 (a)	1.35 ± 0.01 (a)	1.39 ± 0.03 (b)	Permutation	31.78	< 0.001
Disturbance severity	0.67 ± 0.01 (ab)	0.68 ± 0.01 (a)	0.67 ± 0.01 (b)	Permutation	4.02	0.022
Disturbance severity herb layer	0.20 ± 0.02 (a)	0.20 ± 0.02 (a)	0.24 ± 0.03 (b)	Permutation	33.39	< 0.001
EIVE_Moisture	4.62 ± 0.19	4.74 ± 0.13	4.65 ± 0.25	Permutation	1.95	0.15
EIVE_Nitrogen	5.60 ± 0.22 (a)	5.89 ± 0.19 (b)	5.85 ± 0.26 (b)	Permutation	10.87	< 0.001
EIVE_Reaction	6.45 ± 0.16 (a)	6.43 ± 0.13 (a)	6.68 ± 0.10 (b)	Permutation	30.20	< 0.001
EIVE_Light	5.32 ± 0.28 (a)	5.17 ± 0.20 (a)	5.80 ± 0.31 (b)	Permutation	35.97	< 0.001

Table 3.

Download as

CSV

XLSX

Scale-dependent diversity patterns of the herb layer in the three associations. GF: Galio laevigati-Fraxinetum excelsioris, CF: Carici pendulae-Fraxinetum excelsioris, VU: Viburno lantanae-Ulmetum minoris. The column “Test” indicates the analysis applied to test differences among forest types, and the column “Stat.” the value of the statistics of the ANOVA or Kruskal-Wallis tests. Different letters in mean values indicate significant between-pair differences after significant ANOVA or Kruskal-Wallis analysis.

Plot size (m²)	GF n = 23	GF n = 19	VU n = 30	Test	Stat.	p-value
0.0001	1.2 ± 0.7	1.1 ± 0.6	1.3 ± 0.6	Kruskal-Wallis	0.54	0.76
0.001	2.0 ± 1.2	1.9 ± 0.8	1.7 ± 0.8	Kruskal-Wallis	1.29	0.53
0.01	3.6 ± 1.9	3.9 ± 1.4	2.9 ± 1.2	Kruskal-Wallis	4.74	0.09
0.1	8.2 ± 3.3 (a)	7.7 ± 2.6 (ab)	5.7 ± 2.4 (b)	Kruskal-Wallis	12.60	0.002
1	19.5 ± 6.8 (a)	14.8 ± 4.4 (ab)	12.1 ± 4.6 (b)	Kruskal-Wallis	19.22	< 0.001
10	37.3 ± 8.1 (a)	27.6 ± 8.7 (b)	23.6 ± 6.5 (b)	ANOVA	21.43	< 0.001
100	68.1 ± 13.2 (a)	49.3 ± 14.7 (b)	46.8 ± 12.5 (b)	ANOVA	18.48	< 0.001
200	85.4 ± 16.4 (a)	64.9 ± 20.9 (b)	60.4 ± 15.6 (a)	Kruskal-Wallis	20.68	< 0.001

Figure 8.

Box plots of diversity parameters showing the post-hoc Tukey’s post-hoc or Dunn’s multiple comparison test with Bonferroni correction, where ‘ns’ indicates non-significant differences, while significant differences are indicated by ** (p < 0.01) and *** (p < 0.001).

Structural and climatic variables

As mentioned before, cryptogam cover, litter cover, gravel cover, stone cover and BIO7 were the only structural and climatic variables significantly correlated with the results of the NMDS ordination (Figure 5). According to the analyses of variance, the most outstanding difference among the forest types regarding their structure was the more than five-fold higher mean cover of the cryptogam layer in the Galio-Fraxinetum vs. Viburno-Ulmetum, where the cryptogam layer was also considerably more developed than in Carici-Fraxinetum, but without significant differences. Litter and dead wood cover also varied across forest types, with highest values for litter and lowest for dead wood in the Carici-Fraxinetum (Table 2), although differences were only slightly significant. Regarding the fractions of soil surface, there were significant differences among forest types, especially between Galio-Fraxinetum, where cover of stones, rocks and gravel were highest, and the other two forest types, among which Viburno-Ulmetum also showed higher cover of rocks, stones and gravel than Carici-Fraxinetum, but without significant differences.

Climatic variables also significantly varied across the three forest types. The warmest mean annual temperature and the lowest annual precipitation were recorded in Viburno-Ulmetum, which was also subject to the highest annual temperature range and the warmest and driest summer (Table 2). The two temperate ash forests shared climatic features, and their summer ombrothermic index (Ios₃) was clearly over 2, while the submediterranean forests of Viburno-Ulmetum had a mean Ios₃ of 2.14 (Table 2), significantly lower than in the temperate forests, and close to the threshold that defines the Mediterranean climate (Ios₃ < 2).

Ecological and disturbance indicator values

All ecological indicator values analyzed, except moisture, showed significant differences among forest types (Table 2). Galio-Fraxinetum had significantly fewer nitrogen than Carici-Fraxinetum and Viburno-Ulmetum, while pH and light were significantly higher in Viburno-Ulmetum than in Carici-Fraxinetum and Galio-Fraxinetum (Figure 9).

Disturbance frequency of the whole forest and of the herb layer were highest in Viburno-Ulmetum, with significant differences with Carici-Fraxinetum and Galio-Fraxinetum. Regarding the disturbance severity, the highest differences among forest types were for the herb layer, once again with the highest values in Viburno-Ulmetum (Table 2, Figure 10).

Figure 9.

Box plots of unweighted community means of Ecological Indicator Values for Europe showing the results of the permutation tests, where ‘ns’ indicates non-significant differences, while significant differences are indicated by ** (p < 0.01) and *** (p < 0.001).

Figure 10.

Box plots of community weighted means of Disturbance Indicator Values for European plants showing the results of the permutation tests, where ‘ns’ indicates non-significant differences, while significant differences are indicated by ** (p < 0.01) and *** (p < 0.001).

Discussion

We have sampled 72 vegetation plots following a standardized protocol including the establishment of permanent plots and a nested sampling design. Many of these plots have been sampled in the same location as historical plots and have been included in ReSurvey database (Knollová et al. 2024). All of them will allow for future monitoring of these habitats.

Nested sampling in forests

To our knowledge, only a few studies have tried nested sampling designs for plant diversity assessment in forests, besides the classical Whittaker plots (Schmida 1984; Stohlgren et al. 1995), where different grain-sizes are independent and non-overlapping, so nestedness is not perfect. The sampling design used in MONITO Project (Chiarucci et al. 1996) does fit a perfect nesting, with non-independent plots ranging from 1 m² to 2500 m². Other nested-sampling designs in forests include the nested plot series used by Qiao et al. (2012) in Chinese forests, with plot sizes ranging from 0.25 m² to 0.5 ha. In our opinion, the size range used in the two mentioned nested plot designs is not applicable in all forest types, especially in riparian forests, where one can hardly find forest patches big enough to lay 50 m × 50 m plots and even less 70 m × 70 m plots. Staubli et al. (2021) resurveyed forest nested plot series formed by three concentric circles of 30 m², 200 m² and 500 m². This pattern is more suited to small forest patches, but the circular shape is difficult to fit within the linear form of riparian forests. Another important difference is that the mentioned protocols included all vegetation layers in the nested sampling, while ours was limited to the herb layer.

General floristic patterns

Our dataset includes 456 vascular plant taxa, some of them rare or endangered species in the region with their few populations often sheltered in these hardwood riparian forests (Bañares et al. 2004; Uribe-Echebarría and Campos 2006). The mean species richness per 200 m²-plot is 75.3, clearly higher than the richness predicted by Večeřa et al. (2019) for deciduous forests of this region at 400 m² (see figure S2.8 in Večeřa et al. 2019), but within or a bit higher than the range of the input data used for modelling (see figure S1.1 in Večeřa et al. 2019). These results confirm the northern part of the Iberian Peninsula as a secondary biodiversity hotspot for European deciduous forests, after the main hotspots in Central and south-eastern European mountain ranges, the lowland forests of the hemiboreal zone and the Carpathian foothills (Večeřa et al. 2019).

Alien plant invasions are often favored in sites with increased soil nutrient availability (Davis et al. 2000) and have repeatedly been reported as a major issue in riparian ecosystems from the global to the regional scale (Richardson et al. 2007; Schnitzler et al. 2007; Campos et al. 2013), where they are particularly favored, at the regional level, by additional human disturbance, including infrastructures, agricultural fields, and forest plantations (Liendo et al. 2016; Wagner et al. 2021). However, alien plant invasion is not for the moment a big issue in these subcantabrian, western Pyrenean and submediterranean floodplain forests of northern Spain, with the most frequent alien species being archeophytes (Juglans regia and Populus nigra var. italica).

At a European level, however, temperate hardwood riparian forests are among the top-six most invaded forest habitats (Wagner et al. 2017), with a percentage of alien species in the species pool of 8.9%, while in our dataset this percentage is 6.1%. It is interesting to highlight that the percentage of alien species in the species pool reaches 15.7% in the nearby Cantabrian streams (Liendo et al. 2021), where the riparian forest corresponds to the habitat T12 (Alnus glutinosa-Alnus incana forest on riparian and mineral soils). These Cantabrian riparian forests also show a much higher invasion level at plot-level (6.8%) than our target forests in the inner Basque-Navarran valleys (1.6%), which have a quite similar value to European hardwood forests (1.8%). This big difference might be due to several factors, 1) low temperatures constraining the spread of alien species (Svitková et al. 2024), as many alien species common in Cantabrian riparian forests have a subtropical origin (Liendo et al. 2015) and are not adapted to the frequent frosts subcantabrian and submediterranean valleys undergo; 2) dispersal limitation, as invasive species common in Central European forests and thus are adapted to colder environments did not reach northern Spain (Kalusová et al. 2024); and last, but not least 3) lower anthropogenic disturbance of inner valleys comparing to the highly populated and industrialized Cantabrian valleys (Liendo et al. 2015). It might be that there is potentially a high invasion debt in these northern Iberian hardwood riparian forests (Rouget et al. 2016). Wagner et al. (2021) hypothesized that invasion debt is actually more prominent in southern European woodlands.

EUNIS expert system

Our results after the application of the EUNIS-ESy are quite surprising, especially if we consider that 39 plots were only classified at the level 1, and nine were not classified to a unique level 3 habitat, that is, more than half of the plots were not classified at EUNIS level 3. We repeated the procedure in Juice (Tichý 2002), but results were similar, with only slight changes. Even if we consider that the application of the EUNIS expert system at a regional scale might produce some misclassifications due to the optimization of habitat definition for the whole of Europe (Chytrý et al. 2020) it becomes quite evident that the system needs further adjustment. It is not acceptable that more than half of the plots are not classified at level 3 as this is an impediment for the application of the expert system at the regional scale.

González-García et al. (2024) combined the application of the EUNIS-ESy with unsupervised and semi-supervised methods to classify 4,361 plots of all forest types from the Cantabrian Mixed Forest Ecoregion, and eventually obtained 24 habitat types at EUNIS level 4. Although the EUNIS-ESy classified some plots in habitat T13, the authors do not indicate how many they were and to which associations they belonged to, and reclassified them in the regional level 4 units T111, T121, and T1E1 within T11 (Temperate Salix and Populus riparian forest), T12 (Alnus glutinosa-Alnus incana forest on riparian and mineral soils), and T1E (Carpinus and Quercus mesic deciduous forest), respectively. These results are comparable with those from the present study, as most of the plots classified at level 3 were included in habitats T1E, T12 and T13.

Syntaxonomy

The classification of riparian forests in the study area at association level agrees with the framework provided by Biurrun et al. (2016) in their survey of the entire Iberian Peninsula. Three clearly distinct associations are identified: Galio laevigati-Fraxinetum excelsioris for the temperate forests from the western Pyrenean valleys, Carici pendulae-Fraxinetum excelsioris for the temperate forests from the subcantabrian valleys, and Viburno lantanae-Ulmetum minoris for the Fraxinus angustifolia-dominated forests in the submediterranean valleys. Galio-Fraxinetum excelsioris was described as a new association by Biurrun et al. (2016), who selected a relevé sampled in Zaraitzu river as holotype. This type lacks alder, which is absent in this river probably due to its torrential regime (Peralta et al. 2018); nevertheless, alder is present in forests of this association in the rivers Irati and Urrobi, where it is a common component of the tree layer.

The association Carici pendulae-Fraxinetum excelsioris was formally described by Biurrun and García-Mijangos (2002a) to include ash forests from the subcantabrian rivers and streams in the Basque Country and Navarre. The association was first provisionally published by Biurrun (1999) with six relevés from Basaburua and other tributaries of Arga river, all of them in subcantabrian valleys in Navarre. Originally, the association was meant to include ash forests growing on the banks of small streams where alder was absent, but Biurrun (1999) mentioned that it was also present in rivers with enough water flow to support the growth of alder. In these rivers, Carici-Fraxinetum would occupy an external position, between the alder forest and the zonal vegetation, i.e., the mesic oak forest. This alder forest growing on the frequently flooded riverbanks belongs to the association Lonicero xylostei-Alnetum glutinosae, first described by Biurrun et al. (1994) (under the name Hyperico androsaemi-Alnetum glutinosae loniceretosum xylostei), who provided 18 relevés from subcantabrian valleys, most of them with Alnus glutinosa as the dominant species. This association was recognized by Douda et al. (2016) in their pan-European survey of alder forests but was synonymized to Carici-Fraxinetum excelsioris in the Iberian survey by Biurrun et al. (2016). Whether there is a place for two riparian forests in subcantabrian rivers remains unresolved, as this work has focused on the upper riverbanks. It is worth highlighting that our clustering analysis has included in this association one plot from Urrobi river, unlike the rest of plots from this river, classified in the western Pyrenean Galio-Fraxinetum. This plot is located in the upper section of Urrobi, in the high plateau of Burguete, which is also the eastern limit of Quercus robur and mesic oak forests in the region.

The submediterranean association Viburno lantanae-Ulmetum minoris was formally described by Biurrun and García-Mijangos (2002b) to include floodplain elm and ash forests from the Castilian-Cantabrian biogeographic sector (Mediterranean region), with 14 relevés from northern Burgos to Eastern Navarre. The type of the association was sampled in Irati river (San Vicente, Lumbier), at a location where we have resurveyed the historic plot for this study. Similarly to the Carici pendulae-Fraxinetum excelsioris, this association was provisionally and thus invalidly published by Biurrun (1999), who already indicated that this forest occupies an external position to the alder and willow flooded forests of the association Humulo-Alnetum in rivers, while in small streams they grow on the riverbank. Although elm is not usually a dominant species, it can reach high cover beneath the upper canopy formed by ashes, and according to our observations, some big elms can also be found. Elm individuals do not only correspond to Ulmus minor, but also to Ulmus laevis, which was indicated as a naturalized species in Flora iberica (Navarro and Castroviejo 1993), but recent molecular findings revealed that it originates from the Iberian Peninsula (Fuentes-Utrilla et al. 2014).

The three associations, Carici-Fraxinetum excelsioris, Galio-Fraxinetum excelsioris and Viburno-Ulmetum minoris, were included by Biurrun et al. (2016) in the alliance Alnion incanae, but they argued important floristic differences towards Central European floodplain forests to support their proposal of a new suballiance Buxo sempervirentis-Alnenion glutinosae. This new suballiance included non-pioneer temperate and submediterranean floodplain forests from the southern piedmonts of the Pyrenees, Basque-Cantabrian mountains and Cantabrian range (Biurrun et al. 2016), and replaced in northern Spain the two suballiances described from Central Europe: Alnenion glutinoso-incanae Oberd. 1953 and Ulmenion Oberd. 1953, which include streamside and spring-fed forests, and alluvial hardwood forests in large rivers, respectively (Douda 2008; Petrášová and Jarolímek 2012). However, shortly after the publication of the Iberian review, the European checklist of high level syntaxa was published (Mucina et al. 2016), where the name Fraxino-Quercion roboris Passarge 1968 was recovered to include the associations traditionally assigned to Ulmenion in the vegetation syntheses of western and central Europe (Wallnöfer et al. 1993; Stortelder et al. 1999), except for some that only indicated the four main hierarchical levels (Chytrý 2013). The French vegetation prodrome also separates both floodplain forest types, but indicates the name Ulmion minoris (Oberd. 1953) Seytre et Renaux in Renaux et al. 2019 (basionyme Ulmenion minoris Oberd. 1953) for true alluvial hardwood forests, while applied the name Fraxino-Quercion roboris for hygrophilous non-alluvial forests (Renaux et al. 2019). Relationships of the northern Iberian suballiance Buxo-Alnenion glutinosae with the submediterranean alliance Alno-Quercion roboris should also be explored. The latter gathers alder-oak riparian floodplain forests on nutrient-rich alluvial soils in the Apennine and Balkan peninsulas (Mucina et al. 2016; Preislerová et al. 2022), but they are mostly restricted to marshy places (Brullo and Spampinato 1999).

Therefore, the question now arises which is the correct classification at alliance level of the Iberian riparian forests currently classified in the Buxo-Alnenion glutinosae and the Alnion incanae. Should Buxo-Alnenion be transferred to Fraxino-Quercion roboris? Or to Alno-Quercion roboris? Should it be kept in Alnion incanae? According to the ecological conditions, while it is evident that these forests are not growing on the floodplains of big river valleys, it is also true that many stands are not subject to frequent floods but are affected by the fluctuations in the ground water table and irregularly by short-term flooding, as are stands included in Ulmenion (Petrášová and Jarolímek 2012), and consequently in Fraxino-Quercion roboris. Moreover, the forest structure in Ulmenion is also quite similar to our studied forests, with a well-developed and species-rich shrub layer and high cover of the herb layer, with an important temporal turnover between spring and summer. Similar to our forests, low geophytes prevail in spring, while higher hemicryptophytes dominate later (Petrášová and Jarolímek 2012; Pielech 2021).

Species composition also approaches Iberian hardwood forests to Fraxino-Quercion, as species typical of mesic forests are common, and the abundance of Acer campestre and Ligustrum vulgare, among other woody species, separates them from typical Alnion incanae stands (Renaux et al. 2019). Using only trees and tall shrubs, Willner (2024) classified Austrian forests at the level of alliance, resulting in a clear separation of Alnion incanae and Fraxino-Quercion, which led the author to suggest keeping the hardwood alluvial forests of Fraxino-Quercion roboris within the zonal class Carpino-Fagetea. Among the diagnostic species of Fraxino-Quercion in Austria Acer campestre stands out once again, although in this case the ash species bring contrasting arguments, as Fraxinus excelsior and F. angustifolia appear as diagnostic of Alnion incanae and Fraxino-Quercion, respectively. However, we must consider that when taken altogether, both ash species are constant in our dataset, and that Austrian hardwood forests are a special Pannonian type where F. angustifolia replaces F. excelsior, which is typical in many other Central European hardwood riparian forests (Chytrý et al. 2020).

Preislerová et al. (2024) provide only minor differences among both alliances regarding their structural and ecological attributes, such as the presence of broad-leaved evergreen shrubs as dominants in Alnion incanae, or the affinity of this alliance for moist to wet substrate, while Fraxino-Quercion is limited to intermittently wet substrate. In any case, these attributes are based on the current syntaxonomy, therefore there is an intrinsic circularity on their assignation to the alliances, as is evidenced by the geographical and biogeographical attributes: while Alnion incanae is present in the Mediterranean region, Fraxino-Quercion is absent; similarly, Alnion incanae is indicated from lowlands to the montane belt, while Fraxino-Quercion is restricted to lowlands. It is clear that these differences cannot be considered, as they would disappear if Buxo-Alnenion was transferred from Alnion incanae to Fraxino-Quercion.

The correct classification of Buxo-Alnenion glutinosae is only a small-scale aspect of the more complex issue of the floristic, ecological and geographical delimitation of the alliance Alnion incanae regarding the temperate and submediterranean alluvial forests of the Fraxino-Quercion roboris and the Alno-Quercion roboris, respectively, as well as their relationship with the mesic oak forests of the order Carpinetalia betuli, already brought up by Willner (2024). The pan-European classification of floodplain forests conducted by Douda et al. (2016) did not attempt to distinguish stream and seepage forests from hardwood forests, and included them all in the alliance Alnion incanae, without any mention to suballiances. This is understandable as they dealt not only with temperate riparian forests, but also with Mediterranean ones and alder carrs. Therefore, and once the internal classification of European mesic oak forests has been clarified (Novák et al. 2023), we call here for a comprehensive analysis of European temperate riparian, alluvial and mesic oak forests with the aim to clarify the number of alliances and define them with regards to floristic composition, site conditions and biogeography, based on a clear definition of the alliance concept (Willner 2020).

Diversity patterns along environmental gradients

The three associations show high levels of vascular plant species richness, which is expected due to the relatively high productivity of floodplain habitats, but as predicted by the hump-shape model of productivity-richness relationship (Grime 1979), the indicator value of soil nitrogen is significantly lower in the most species-rich association, i.e., Galio-Fraxinetum (see Figure 9). This is consistent with the results provided by Slezák et al. (2022), who found that species richness declined with productivity in Slovakian floodplain forests. Interestingly, light conditions and soil reaction are highest in Viburno-Ulmetum, which shows relatively low species richness, although light and soil pH have emerged as positively influencing species richness in forests (Chiarucci et al. 2001; Wohlgemuth et al. 2008; Hofmeister et al. 2009; Dormann et al. 2020) with open canopy forests with moderately high soil pH being the richest ones. It was not the aim of this study to model the species richness of the studied forests; however, it seems that other factors are more important than light and soil pH for species richness in the context of riparian hardwood forests in northern Iberian Peninsula. If we pay attention to the variables included in the modelling of richness of European deciduous forests in Večeřa et al. (2019), the three top-ranked predictors were temperature seasonality, annual precipitation and forest area. At our regional scale, the richest association Galio-Fraxinetum receives the highest annual precipitation (although without significant differences with Carici-Fraxinetum), but assuming temperature annual range as a proxy of temperature seasonality, the highest values are in Viburno-Ulmetum, which has, together with Carici-Fraxinetum, the lowest species richness. In the absence of modelling, we can conclude that alpha richness of these hardwood forests is driven by a set of biogeographic and environmental factors. Among them, the mountainous character of the western Pyrenean Galio-Fraxinetum, linked to lower site productivity and higher flooding intensity, might be a positive factor for its high alpha species richness, as already observed in Central European floodplain alder forests, where richness was significantly higher in the Carpathians compared to the Polish Plain and the Pannonian lowland (Hrivnák et al. 2022). This mountainous character is also reflected on the high soil heterogeneity in Galio-Fraxinetum, where cover of stones, rocks and gravel is higher than in the other forest types, leading to higher local species richness. Schuster and Diekmann (2005) already stated the importance of micro-heterogeneity in habitat resources, especially edaphic factors, for the determination of species diversity patterns in deciduous forests.

Shannon diversity is lowest in the submediterranean forests of the Viburno-Ulmetum minoris, with significant differences regarding the two other forest types. This is especially relevant if we compare it to species richness, as Viburno-Ulmetum and Carici-Fraxinetum show similar values. This might be due to the fact that submediterranean forests receive more light in the ground layer (see Figure 9), where a few ruderal or grassland species are able to achieve high cover values: Rubus caesius, Urtica dioica, Carex flacca, Melisssa officinalis and even herbaceous climbers as Humulus lupulus.

Viburno-Ulmetum minoris has the highest Whittaker’s beta diversity, i.e., there is higher species turnover among plots belonging to this association than among plots belonging to Galio-Fraxinetum and especially to Carici-Fraxinetum. One of the reasons behind this might be that the river system is more complex in the sections where Viburno-Ulmetum develops, and we have sampled a higher variety of geomorphological conditions and landscape configurations. It is important to highlight that in our studied forests the environmental distance seems to play a more important role than geographical distance on beta diversity, as has been already seen for other regions (Vega et al. 2020). In fact, the lowest species turnover was measured in Carici-Fraxinetum, which is also the forest type where plots altogether are more geographically distant, but, according to our observations, is the most homogeneous regarding site conditions and landscape configuration.

The mean z values in the studied forests were highest in Galio-Fraxinetum (0.32), while Carici-Fraxinetum and Viburno-Ulmetum shared a lower value of 0.29. Similar values among 0.26 and 0.30 were found by Chiarucci et al. (2001) in several forest types from Tuscany and Qiao et al. (2012) in Chinese forests across large latitudinal and altitudinal gradients. It has to be taken into account that the mentioned studies included all life-forms in the analysis, while we only included herbaceous species and seedlings and saplings of woody species. For comparison, mean z values were ≤ 0.24 in grassland vegetation both at regional (Filibeck et al. 2019) and Palaearctic scale (Dembicz et al. 2021a, 2021b). Although differences were not significant, the higher z value in the Galio-Fraxinetum is probably a consequence of the higher cover of stones, rocks and gravel in this association, as already demonstrated for grassland habitats at regional (Filibeck et al. 2019) and Palaearctic scale (Dembicz et al. 2021a).

Structural and climatic variables

The most relevant structural differences among forest types refer to the cryptogam cover, i.e., the moss cover, and the cover of stones, rocks and gravel. These parameters are significantly higher in the Galio-Fraxinetum, probably related to the more torrential water flow, which allows the deposition of coarse material and the removal of accumulated litter and thus offering more space for bryophyte development.

The submediterranean forests of Viburno-Ulmetum showed significant climatic differences regarding the two Fraxinus excelsior-dominated associations. It is worth highlighting the low value of the summer ombrothermic index (Ios₃), close to the threshold that defines the Mediterranean climate. This result is in line with the results of Biurrun et al. (2014), who analyzed riparian forests of northern Iberian Peninsula and found that the threshold value of the ombrothermic index of the two warmest months (Ios₂) for a true Mediterranean climate, i.e., Ios₂ < 2 approximately coincides with the replacement of Fraxinus excelsior by F. angustifolia.

Conclusions

The present study represents the first attempt to assess the relationships between species richness and spatial scale in Iberian forest ecosystems. With this aim we have adapted to forest habitats the EDGG sampling protocol developed for the nested sampling of grasslands and other open habitats. Moreover, we have established such nested plots as permanent plots that will facilitate monitoring the effects of global change on these forests. This is particularly relevant because hardwood riparian forests are especially vulnerable to climate change in northern Iberian Peninsula, as this region represents their European southwestern distribution limit. Their fragility facing climate change and their high conservation value, both as an endangered and diverse habitat and as refuge for threatened flora and fauna commit us to in-depth investigate their diversity patterns and ecological requirements. It becomes evident that in order to establish efficient conservation measures we need to clarify their identity both in the frame of EUNIS habitat classification and the European phytosociological classification system.

Data availability

Full plot-level information, including geographic coordinates and composition data, are available under request from the Vegetation Database of the University of the Basque Country (EU-00-011), as well as from JB and IS, who also store the data of the nested-plot series.

Author contributions

IB and JAC planned the research, JB and IS conducted the field sampling with participation of IB and JAC, IB performed the vegetation classification and diagnostic species analyses and led the writing, IS and JAC calculated the diversity indices and performed the ANOVAs and permutation tests, JB prepared the map, extracted the climatic variables and performed the EUNIS-Esy and the ordination analysis, and all authors critically revised the manuscript.

Acknowledgments

We would like to express our gratitude to Itziar García-Mijangos, who helped with the expert system, and to Eneko Barriola and Denys Vynokurov, who helped with the sampling. This work has been funded by the Basque Government (IT1487-22), Jokin Belmonte has been supported by Next Generation funds from the European Union (Investigo Program n°58) and Irati Sanz-Zubizarreta by a predoctoral grant of the University of the Basque Country UPV/EHU (PIF21/255).

References

Bañares A, Blanca G, Güemes J, Moreno Saiz JC, Ortiz S [Eds] (2004) Atlas y libro rojo de la flora vascular amenazada de España: táxones prioritarios. Ministerio de Medio Ambiente, Madrid, ES, 1069 pp. https://bit.ly/3P9vUZV

Biurrun I (1999) Flora y vegetación de los ríos y humedales de Navarra. Guineana 5: 1–338.

Biurrun I, Font X (2020) SIVIM Floodplain Forests – Database of riverine forests and scrubs from the Iberian Peninsula. Vegetation Classification and Survey 1: 171–172. https://doi.org/10.3897/VCS/2020/61660

Biurrun I, García-Mijangos I (2002a) Carici pendulae-Fraxinetum excelsioris ass. nova hoc loco. Itinera Geobotanica 15: 77–78.

Biurrun I, García-Mijangos I (2002b) Viburno lantanae-Ulmetum minoris ass. nova hoc loco. Itinera Geobotanica 15: 222–224.

Biurrun I, García-Mijangos I, Loidi J (1994) Study of alder forests in the Basque Country and bordering territories by means of multivariate analysis. Botanica Helvetica 104: 31–54.

Biurrun I, García-Mijangos I, Campos JA, Herrera M, Loidi J (2012) Vegetation-plot database of the University of the Basque Country (BIOVEG). Biodiversity and Ecology 4: 328. https://doi.org/10.7809/b-e.00121

Biurrun I, Campos JA, Herrera M, Loidi J, García-Mijangos I (2014) A survey of the riverine forests of the northern Iberian Peninsula: numerical classification versus traditional syntaxonomy and relationships with climatic gradients. Documents Phytosociologiques, Série 3: 90–107.

Biurrun I, Campos JA, García-Mijangos I, Herrera M, Loidi J (2016) Floodplain forests of the Iberian Peninsula: Vegetation classification and climatic features. Applied Vegetation Science 19: 336–354. https://doi.org/10.1111/avsc.12219

Boletín Oficial de Navarra (2023) DECRETO FORAL 10/2023, de 15 de febrero, por el que se crea el Listado Navarro de Especies de Flora Silvestre en Régimen de Protección Especial y se establece el Catálogo de Especies de Flora Amenazada de Navarra. Government of Navarre, Pamplona-Iruñea, ES.

Boletín Oficial del Estado (2011) Real Decreto 139/2011, de 4 de febrero, para el desarrollo del Listado de Especies Silvestres en Régimen de Protección Especial y del Catálogo Español de Especies Amenazadas. Spanish Government, Madrid, ES.

Boletín Oficial del País Vasco (2011) ORDEN de 10 de enero de 2011, de la Consejera de Medio Ambiente, Planificación Territorial, Agricultura y Pesca, por la que se modifica el Catálogo Vasco de Especies Amenazadas de la Fauna y Flora Silvestre y Marina, y se aprueba el texto único. Basque Government, Vitoria-Gasteiz, ES.

Bray JR, Curtis JT (1957) An ordination of the upland forest communities of Southern Wisconsin. Ecological Monographs 27: 325–349. https://doi.org/10.2307/1942268

Bruelheide H, Dengler J, Jiménez-Alfaro B, Purschke O, Hennekens SM, Chytrý M, Pillar VD, Jansen F, Kattge J, … Zverev A (2019) sPlot – A new tool for global vegetation analyses. Journal of Vegetation Science 30: 161–186. https://doi.org/10.1111/jvs.12710

Bruelheide H, Tichý L, Chytrý M, Jansen F (2021) Implementing the formal language of the vegetation classification expert systems (ESy) in the statistical computing environment R. Applied Vegetation Science 24: e12562. https://doi.org/10.1111/avsc.12562

Brullo S, Spampinato G (1999) Syntaxonomy of hygrophilous woods of the Alno-Quercion roboris. Annali di Botanica 57: 133–146.

Campos JA, Biurrun I, García-Mijangos I, Loidi J, Herrera M (2013) Assessing the level of plant invasion: A multi-scale approach based on vegetation plots. Plant Biosystems 147: 1148–1162. https://doi.org/10.1080/11263504.2013.861538

Chiarucci A, Bonini I, De Dominicis V (1996) Vascular plant diversity. In: Bartolozzi L, Bussotti F, De Dominicis V, Ferretti M (Eds) Program MONITO: Concepts, Structure and Results. Regione Toscana, Giunta Regionale Publisher, Firenze, IT, 57–61.

Chiarucci A, De Dominicis V, Wilson JB (2001) Structure and floristic diversity in permanent monitoring plots in forest ecosystems of Tuscany. Forest Ecology and Management 141: 201–210. https://doi.org/10.1016/S0378-1127(00)00329-7

Chytrý M [Ed.] (2013) Vegetation of the Czech Republic. 4. Forest and scrub vegetation. Academia, Prague, CZ.

Chytrý M, Tichý L, Holt J, Botta-Dukát Z (2002) Determination of diagnostic species with statistical fidelity measures. Journal of Vegetation Science 13: 79–90. https://doi.org/10.1111/j.1654-1103.2002.tb02025.x

Chytrý M, Hennekens SM, Jiménez-Alfaro B, Knollová I, Dengler J, Jansen F, Landucci F, Schaminée JHJ, Aćić S, … Yamalov S (2016) European Vegetation Archive (EVA): an integrated database of European vegetation plots. Applied Vegetation Science 19: 173–180. https://doi.org/10.1111/avsc.12191

Chytrý M, Tichý L, Hennekens SM, Knollová I, Janssen JAM, Rodwell JS, Peterka T, Marcenò C, Landucci F, … Schaminée JHJ (2020) EUNIS Habitat Classification: expert system, characteristic species combinations and distribution maps of European habitats. Applied Vegetation Science 23: 648–675. https://doi.org/10.1111/avsc.12519

Chytrý M, Tichý L, Hennekens SM, Knollová I, Janssen JAM, Rodwell JS, Peterka T, Marcenò C, Landucci F, … Schaminée JHJ (2021) EUNIS-ESy: Expert system for automatic classification of European vegetation plots to EUNIS habitats. Version v2021-06-01. https://doi.org/10.5281/zenodo.4812736 [accessed 1 Sep 2024]

Chytrý M, Řezníčková M, Novotný P, Holubová D, Preislerová Z, Attorre F, Biurrun I, Blažek P, Bonari G, … Axmanová I (2024) FloraVeg.EU — An online database of European vegetation, habitats and flora. Applied Vegetation Science 27: e12798. https://doi.org/10.1111/avsc.12798

Clarke KR, Gorley RN, Somerfield PJ, Warwick RM (2014) Change in marine communities: an approach to statistical analysis and interpretation, 3^rd ed. PRIMER-E, Plymouth, UK.

Davis MA, Grime P, Thompson K (2000) Fluctuating resources in plant communities: A general theory of invasibility. Journal of Ecology 88: 528–534. https://doi.org/10.1046/j.1365-2745.2000.00473.x

de Bello F, Lepš J, Sebastià MT (2007) Grazing effects on the species–area relationship: variation along aclimatic gradient in NE Spain. Journal of Vegetation Science 18: 25–34. https://doi.org/10.1111/j.1654-1103.2007.tb02512.x

De Cáceres M, Font X, Oliva F, Vives S (2007) GINKGO, a program for non-standard multivariate fuzzy analysis. Advances in Fuzzy Sets and Systems 2: 41–56.

Dembicz I, Dengler J, Steinbauer MJ, Matthews TJ, Bartha S, Burrascano S, Chiarucci A, Filibeck G, Gillet F, … Biurrun I (2021a) Fine-grain beta diversity of Palaearctic grassland vegetation. Journal of Vegetation Science 32: e13045. https://doi.org/10.1111/jvs.13045

Dembicz I, Dengler J, Gillet F, Matthews TJ, Steinbauer MJ, Bartha S, Campos JA, De Frenne P, Dolezal J, … Biurrun I (2021b) Fine-grain beta diversity in Palaearctic open vegetation: variability within and between biomes and vegetation types. Vegetation Classification and Survey 2: 293–304. https://doi.org/10.3897/VCS/2021/77193

Dengler J, Boch S, Filibeck G, Chiarucci A, Dembciz I, Guarino R, Henneberg B, Janišová M, Marcenò C, … Biurrun I (2016) Assessing plant diversity and composition in grasslands across spatial scales: the standardised EDGG sampling methodology. Bulletin of the Eurasian Dry Grassland Group 32: 13–30.

Dengler J, Matthews TJ, Steinbauer MJ, Wolfrum S, Boch S, Chiarucci A, Conradi T, Dembicz I, Marcenò C, … Biurrun I (2020) Species–area relationships in continuous vegetation: Evidence from Palaearctic grasslands. Journal of Biogeography 60: 72–86. https://doi.org/10.1111/jbi.13697

Dengler J, Biurrun I, Dembicz I (2021) Standardised EDGG methodology for sampling grassland diversity: second amendment. Palaearctic Grasslands 49: 22–26. https://doi.org/10.21570/EDGG.PG.49.22-26

Dengler J, Jansen F, Chusova O, Hüllbusch E, Nobis MP, Van Meerbeek K, Axmanová I, Bruun HH, Chytrý M, … Gillet F (2023) Ecological Indicator Values for Europe (EIVE) 1.0. Vegetation Classification and Survey 4: 7–29. https://doi.org/10.3897/VCS.98324

Dormann CF, Bagnara M, Boch S, Hinderling J, Janeiro-Otero A, Schäfer D, Schall P, Hartig F (2020) Plant species richness increases with light availability, but not variability, in temperate forests understorey. BMC Ecology 20: e43. https://doi.org/10.1186/s12898-020-00311-9

Douda J (2008) Formalized classification of the vegetation of alder carr and floodplain forests in the Czech Republic. Preslia 80: 199–224.

Douda J, Boublík K, Slezák M, Biurrun I, Nociar J, Havrdová A, Doudová J, Aćić S, Brisse H, … Zimmermann NE (2016) Vegetation classification and biogeography of European floodplain forests and alder cars. Applied Vegetation Science 19: 147–163. https://doi.org/10.1111/avsc.12201

Drakare S, Lennon JJ, Hillebrand H (2006) The imprint of the geographical, evolutionary and ecological context on species–area relationships. Ecology Letters 9: 215–227. https://doi.org/10.1111/j.1461-0248.2005.00848.x

EEA (2016) Biogeographical Regions in Europe. European Environment Agency, Copenhagen, DK. http://bit.ly/3BKqjq0 [accessed 2 Sep 2024]

European Commission (2011) European red list of vascular plants. Publications Office of the European Union, Luxembourg, LU.

European Union (1992) Council Directive 92/43/EEC of 21 May 1992 on the conservation of natural habitats and of wild fauna and flora. Official Journal of the European Union, Luxembourg, LU.

Euro+Med (2024) The Euro+Med PlantBase – the information resource for Euro-Mediterranean plant diversity. http://ww2.bgbm.org/EuroPlusMed/ [accessed 15 Jul 2024]

Faith DP, Minchin PR, Belbin L (1987) Compositional dissimilarity as a robust measure of ecological distance. Vegetatio 69: 57–68. https://doi.org/10.1007/BF00038687

Fick SE, Hijmans RJ (2017) WorldClim 2: new 1-km spatial resolution climate surfaces for global land areas. International Journal of Climatology 37: 4302–4315. https://doi.org/10.1002/joc.5086

Filibeck G, Sperandii MG, Bazzichetto M, Mancini LD, Rossini F, Cancellieri L (2019) Exploring the drivers of vascular plant richness at very fine spatial scale in sub-Mediterranean limestone grasslands (Central Apennines, Italy). Biodiversity and Conservation 28: 2701–2725. https://doi.org/10.1007/s10531-019-01788-7

Fischer HS (2015) On the combination of species cover values from different vegetation layers. Applied Vegetation Science 18: 169–170. https://doi.org/10.1111/avsc.12130

Fuentes-Utrilla P, Venturas M, Hollingsworth PM, Squirrell J, Collada C, Stone GN, Gil L (2014) Extending glacial refugia for a European tree: genetic markers show that Iberian populations of white elm are native relicts and not introductions. Heredity 112: 105–113. https://doi.org/10.1038/hdy.2013.81

García-Mijangos I, Berastegi A, Biurrun I, Dembicz I, Janišová M, Kuzemko A, Vynokurov D, Ambarlı D, Etayo J, … Dengler J (2021) Grasslands of Navarre (Spain), focusing on the Festuco-Brometea: classification, hierarchical expert system and characterisation. Vegetation Classification and Survey 2: 195–231. https://doi.org/10.3897/VCS/2021/69614

González-García V, Fernández-Pascual E, Font X, Jiménez-Alfaro B (2024) Forest habitat diversity in the Cantabrian Mixed Forests ecoregion (NW Iberian Peninsula), a climatic refugium in western Europe. Applied Vegetation Science 27: e12793. https://doi.org/10.1111/avsc.12793

Grime JP (1979) Plant strategies and vegetation processes. Wiley, Chichester, UK.

Hofmeister J, Hošek J, Modrý M, Roleček J (2009) The influence of light and nutrient availability on herb layer species richness in oak-dominated forests in central Bohemia. Plant Ecology 205: 57–75. https://doi.org/10.1007/s11258-009-9598-z

Hrivnák R, Jarčuška B, Jarolímek I, Kochjarová J, Májeková J, Hegedüšová Vantarová K, Slezák M (2022) Comparative diversity of vascular plants in black alder floodplain and swamp forests of Central European biogeographical regions. Biodiversity Data Journal 10: e90281. https://doi.org/10.3897/BDJ.10.e90281

Janssen JAM, Rodwell JS, García Criado M, Gubbay S, Haynes T, Nieto A, Sanders N, Landucci F, Loidi J, … Valachovič M (2016) European red list of habitats. Part 2. Terrestrial and Freshwater Habitats. Publications Office of the European Union, Luxembourg, LU. https://doi.org/10.2779/091372

Jurasinski G, Retzer V, Beierkuhnlein C (2009) Inventory, differentiation, and proportional diversity: A consistent terminology for quantifying species diversity. Oecologia 159: 15–26. https://doi.org/10.1007/s00442-008-1190-z

Kalusová V, Čeplová N, Danihelka J, Večeřa M, Pyšek P, Albert A, Anastasiu P, Biurrun I, Boch S, … Axmanová I (2024) Alien plants of Europe: an overview of national and regional inventories. Preslia 96: 149–182. https://doi.org/10.23855/preslia.2024.149

Knollová I, Chytrý M, Bruelheide H, Dullinger S, Jandt U, Bernhardt-Römermann M, Biurrun I, de Bello F, Glaser M, … Essl F (2024) ReSurveyEurope: A database of resurveyed vegetation plots in Europe. Journal of Vegetation Science 35: e13235. https://doi.org/10.1111/jvs.13235

Liendo D, Biurrun I, Campos JA, Herrera M, Loidi J, García-Mijangos I (2015) Invasion patterns in riparian habitats: The role of anthropogenic pressure in temperate streams. Plant Biosystems 149: 289–297. https://doi.org/10.1080/11263504.2013.822434

Liendo D, García-Mijangos I, Campos JA, López-Muniain U, Biurrun I (2016) Drivers of Plant Invasion at Broad and Fine Scale in Short Temperate Streams. River Research and Application 32: 1730–1739. https://doi.org/10.1002/rra.3024

Liendo D, Biurrun I, Campos JA, García-Mijangos I, Pearman PB (2021) Effects of disturbance and alien plants on the phylogenetic structure of riverine communities. Journal of Vegetation Science 32: e12933. https://doi.org/10.1111/jvs.12933

Loidi J, Biurrun I, Campos JA, García-Mijangos I, Herrera M (2011) La vegetación de la Comunidad Autónoma del País Vasco. Leyenda del mapa de series de vegetación a escala 1:50.000. University of the Basque Country, Bilbao, ES.

Matthews TJ, Guilhaumon F, Triantis KA, Borregaard MK, Whittaker RJ, Santos A (2016) On the form of species-area relationships in habitat islands and true islands. Global Ecology and Biogeography 25: 847–858. https://doi.org/10.1111/geb.12269

McCune B, Grace JB (2002) Analysis of ecological communities. MjM Software Design, Gleneden Beach, OR, US.

Mercadal G, Vilar L (2013) Caracterització de les freixenedes al·luvials inundables del nord-est de Catalunya (Carici remotae-Fraxinetum oxycarpae Pedrotti 1970 corr. Pedrotti 1992) [Characterization of alluvial and floodplain ash forests from NE Catalonia (Carici remotae-Fraxinetum oxycarpae Pedrotti 1970 corr. Pedrotti 1992)]. Orsis 27: 53–94. [In Catalan]

Midolo G, Herben T, Axmanová I, Marcenò C, Pätsch R, Bruelheide H, Karger DN, Acic S, Bergamini A, … Chytrý M (2023) Disturbance indicator values for European plants. Global Ecology and Biogeography 32: 24–34. https://doi.org/10.1111/geb.13603

MITECO (2024) Lista patrón de los hábitats terrestres presentes en España. https://bit.ly/3BEPQkn [accessed 10 Nov 2024]

Mucina L, Bültmann H, Dierßen K, Theurillat JP, Raus T, Čarni A, Šumberová K, Willner W, Dengler J, … Tichý L (2016) Vegetation of Europe: hierarchical floristic classification system of vascular plant, bryophyte, lichen, and algal communities. Applied Vegetation Science 19 (Suppl. 1): 3–264. https://doi.org/10.1111/avsc.12257

Navarro C, Castroviejo S (1993) Ulmus. In: Castroviejo et al. (Eds) Flora iberica. Vol. III. Plumbaginaceae (partim)-Capparaceae. Real Jardín Botánico, CSIC, Madrid, ES, 244–248.

Novák P, Willner W, Biurrun I, Gholizadeh H, Heinken T, Jandt U, Kollár J, Kozhevnikova M, Naqinezhad A, … Chytrý M (2023) Classification of European oak–hornbeam forests and related vegetation types. Applied Vegetation Science 26: e12712. https://doi.org/10.1111/avsc.12712

Oksanen J, Simpson G, Blanchet F, Kindt R, Legendre P, Minchin P, O’Hara R, Solymos P, Stevens M, … Weedon J (2024) vegan: community ecology package. R package version 2.6-6.1. https://CRAN.R-project.org/package=vegan [accessed 20 Sep 2024]

Peralta J, Biurrun I, García-Mijangos I, Remón JL, Olano JM, Lorda M, Loidi J, Campos JA (2018) Manual de Hábitats de Navarra. 2^nd ed. Government of Navarre, Pamplona, ES.

Petrášová M, Jarolímek I (2012) Hardwood floodplain forests in Slovakia: syntaxonomical revision. Biologia 67: 889–908. https://doi.org/10.2478/s11756-012-0078-x

Pielech R (2021) Fraxino-Quercion roboris Passarge 1968. Elm-ash and oak riparian floodplain forests on nutrient-rich brown soils in the nemoral zone of Europe. In: Mandžukovski D, Čarni A, Sotirovski K (Eds) Interpretative manual of European riparian forests and shrublands. Ss Cyril and Methodius University in Skopje, Skopje, MK, 57–62.

Preislerová Z, Jiménez-Alfaro B, Mucina L, Berg C, Bonari G, Kuzemko A, Landucci F, Marcenò C, Monteiro-Henriques T, … Chytrý M (2022) Distribution maps of vegetation alliances in Europe. Applied Vegetation Science 25: e12642. https://doi.org/10.1111/avsc.12642

Preislerová Z, Marcenò C, Loidi J, Bonari G, Borovyk D, Gavilán RG, Golub V, Terzi M, Theurillat J-P, … Chytrý M (2024) Structural, ecological and biogeographical attributes of European vegetation alliances. Applied Vegetation Science 27: e12766. https://doi.org/10.1111/avsc.12766

Qiao X, Tang Z, Shen Z, Fang J (2012) What causes geographical variation in the species–area relationships? A test from forests in China. Ecography 35: 1110–1116. https://doi.org/10.1111/j.1600-0587.2011.06869.x

R Core Team (2024) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, AT. https://www.R-project.org/ [accessed 20 Sep 2024]

Renaux B, Timbal J, Gauberville C, Thébaud G, Bardat J, Lalanne A, Royer J-M, Seytre L (2019) Contribution au Prodrome des végétations de France: les Carpino betuli-Fagetea sylvaticae Jakucs 1967. Documents phytosociologiques, série 3, 11: 1–423.

Richardson DM, Holmes PM, Esler KJ, Galatowitsch SM, Stromberg JC, Kirkman SP, Hobbs RJ (2007) Riparian vegetation: Degradation, alien plant invasions, and restoration prospects. Diversity and Distributions 13: 126–139. https://doi.org/10.1111/j.1366-9516.2006.00314.x

Rivas-Martínez S (2011) Mapa de series, geoseries y geopermaseries de vegetación de España. Parte II. Itinera Geobotanica 18: 5–424.

Rivas-Martínez S, Rivas Sáenz S, Penas A (2011) Worldwide bioclimatic classification system. Global Geobotany 1: 1–643.

Rodwell J, Ssymank A (2016) G1.2b Temperate and boreal hardwood riparian woodland. https://bit.ly/41HTvse [accessed 15 Oct 2024]

Rouget M, Robertson MP, Wilson JRU, Hui C, Essl F, Renteria JL, Richardson DM (2016) Invasion debt – quantifying future biological invasions. Diversity and Distributions 22: 445–456. https://doi.org/10.1111/ddi.12408

Scherrer D, Guisan A (2019) Ecological indicator values reveal missing predictors of species distributions. Scientific Reports 9: e3061. https://doi.org/10.1038/s41598-019-39133-1

Schmida A (1984) Whittaker’s plant diversity sampling method. Israel Journal of Botany 33: 41–46.

Schnitzler A, Hale BW, Alsum EM (2007) Examining native and exotic species diversity in European riparian forests. Biological Conservation 138: 146–156. https://doi.org/10.1016/j.biocon.2007.04.010

Schuster B, Diekmann M (2005) Species richness and environmental correlates in deciduous forests of Northwest Germany. Forest Ecology and Management 206: 197–205. https://doi.org/10.1016/j.foreco.2004.10.063

Slezák M, Douda J, Šibíková M, Jarolímek I, Senko D, Hrivnák R (2022) Topographic indices predict the diversity of Red List and non-native plant species in human-altered riparian ecosystems. Ecological Indicators 139: e108949. https://doi.org/10.1016/j.ecolind.2022.108949

Staubli E, Dengler J, Billeter R, Wohlgemuth T (2021) Changes in biodiversity and species composition of temperate beech forests in Switzerland over 26 years. Tuexenia 41: 87–108.

Stohlgren TJ, Falkner MB, Schell LD (1995) A Modified-Whittaker nested vegetation sampling method. Vegetatio 117: 113–121. https://doi.org/10.1007/BF00045503

Stortelder AHF, Schaminée JHJ, Hommel PWFM (1999) De Vegetatie van Nederland. Deel 5. Plantengemeenshappen van ruigten, struwelen en bossen [The vegetation of the Netherlands. Vol. 5. Plant communities of shrubs, thickets and forests]. Opulus Press, Uppsala, SE.

Svitková I, Svitok M, Čejka T, Širka P, Galvánek D, Gömöry E, Gömöryová E, Jochjarová J, Senko D, … Hrivnák R (2024) Contrasting diversity patterns of native and alien species across multiple taxa in Central European river corridors. Ecological Indicators 169: e112859. https://doi.org/10.1016/j.ecolind.2024.112859

Tichý L (2002) JUICE, software for vegetation classification. Journal of Vegetation Science 13: 451–453. https://doi.org/10.1111/j.1654-1103.2002.tb02069.x

Tichý L, Chytrý M (2006) Statistical determination of diagnostic species for site groups of unequal size. Journal of Vegetation Science 17: 809–818. https://doi.org/10.1111/j.1654-1103.2006.tb02504.x

Uribe-Echebarría PM, Campos JA (2006) Flora vascular amenazada en la Comunidad Autónoma del País Vasco. Servicio Central de Publicaciones del Gobierno Vasco, Vitoria-Gasteiz, ES, 389 pp.

Večeřa M, Divíšek J, Lenoir J, Jiménez-Alfaro B, Biurrun I, Knollová I, Agrillo E, Campos JA, Čarni A, … Chytrý M (2019) Alpha diversity of vascular plants in European forests. Journal of Biogeography 46: 1919–1935. https://doi.org/10.1111/jbi.13624

Vega E, Martínez-Ramos M, García-Oliva F, Oyama K (2020) Influence of environmental heterogeneity and geographic distance on beta-diversity of woody communities. Plant Ecology 221: 595–614. https://doi.org/10.1007/s11258-020-01036-x

Wagner V, Chytrý M, Jiménez-Alfaro B, Pergl J, Hennekens S, Biurrun I, Knollová I, Berg C, Vassilev K, … Pyšek P (2017) Alien plant invasions in European woodlands. Diversity and Distributions 23: 969–981. https://doi.org/10.1111/ddi.12592

Wagner V, Večeřa M, Jiménez-Alfaro B, Pergl J, Lenoir J, Svenning J-C, Pyšek P, Agrillo E, Biurrun I, … Chytrý M (2021) Alien plant invasion hotspots and invasion debt in European woodlands. Journal of Vegetation Science 32: e13014. https://doi.org/10.1111/jvs.13014

Wallnöfer S, Mucina L, Grass V (1993) Querco-Fagetea. In: Mucina L, Grabherr G, Wallnöfer S (Eds) Die Pflanzengesellschaften Österreichs. Teil III. Gustav Fischer Verlag, Jena, DE, 85–236.

Whittaker RH (1960) Vegetation of the Siskiyou mountains, Oregon and California. Ecological Monographs 30: 279–338. https://doi.org/10.2307/1943563

Willner W (2020) What is an alliance? Vegetation Classification and Survey 1: 139–144. https://doi.org/10.3897/VCS/2020/56372

Willner W (2024) How to classify forests? A case study from Central Europe. Vegetation Classification and Survey 5: 17–26. https://doi.org/10.3897/VCS.117703

Wohlgemuth T, Moser B, Brändli U-B, Kull P, Schütz M (2008) Diversity of forest plant species at the community and landscape scales in Switzerland. Plant Biosystems 142: 604–613. https://doi.org/10.1080/11263500802410975

Zelený D (2018) Which results of the standard test for community weighted mean approach are too optimistic? Journal of Vegetation Science 29: 953–966. https://doi.org/10.1111/jvs.12688

Zelený D (2020) weimea: Weighted Mean Analysis. R package version 0.1.18. https://www.davidzeleny.net/anadat-r/doku.php/en:cwm_r [accessed 28 Nov 2024]

Zelený D, Schaffers AP (2012) Too good to be true: Pitfalls of using mean Ellenberg indicator values in vegetation analyses. Journal of Vegetation Science 23: 419–431. https://doi.org/10.1111/j.1654-1103.2011.01366.x

Zhang J, Gillet F, Bartha S, Alatalo JM, Biurrun I, Dembicz I, Grytnes JA, Jaunatre R, Pielech R, … Dengler J (2021) Scale dependence of species-area relationships is widespread but generally weak in Palaearctic grasslands. Journal of Vegetation Science 32: e13044. https://doi.org/10.1111/jvs.13044

E-mail and ORCID

Idoia Biurrun (Corresponding author, idoia.biurrun@ehu.eus), ORCID: https://orcid.org/0000-0002-1454-0433
Jokin Belmonte (jokin.belmonte@ehu.eus), ORCID: https://orcid.org/0009-0002-8862-5584
Irati Sanz-Zubizarreta (irati.sanz@ehu.eus), ORCID: https://orcid.org/0009-0000-9816-2574
Juan Antonio Campos (juanan.campos@ehu.eus), ORCID: https://orcid.org/0000-0001-5992-2753

Supplementary materials

Supplementary material 1

Header data and metadata of the 72 plots (*.xlsx)

Download file (42.87 kb)

Supplementary material 2

Full synoptic table with relative frequency values of all species across the three forest types (*.xlsx

Download file (27.75 kb)

﻿Abstract

Keywords

﻿Introduction

﻿Study area

﻿Methods

﻿Vegetation data

﻿Diversity variables

﻿Ecological and disturbance indicator values

﻿Climatic data

﻿Statistical analyses

﻿Expert system for automatic classification to EUNIS habitats

﻿Unsupervised classification, ordination and diagnostic species

﻿Analyses of differences between forest types

﻿Results

﻿General floristic results

﻿Expert system

﻿Floristic differentiation in riparian forests

﻿Diversity patterns

﻿Structural and climatic variables

﻿Ecological and disturbance indicator values

﻿Discussion

﻿Nested sampling in forests

﻿General floristic patterns

﻿EUNIS expert system

﻿Syntaxonomy

﻿Diversity patterns along environmental gradients

﻿Structural and climatic variables

﻿Conclusions

﻿Data availability

﻿Author contributions

﻿Acknowledgments

﻿References

﻿E-mail and ORCID

Supplementary materials

Abstract

Introduction

Study area

Methods

Vegetation data

Diversity variables

Ecological and disturbance indicator values

Climatic data

Statistical analyses

Expert system for automatic classification to EUNIS habitats

Unsupervised classification, ordination and diagnostic species

Analyses of differences between forest types

Results

General floristic results

Expert system

Floristic differentiation in riparian forests

Diversity patterns

Structural and climatic variables

Ecological and disturbance indicator values

Discussion

Nested sampling in forests

General floristic patterns

EUNIS expert system

Syntaxonomy

Diversity patterns along environmental gradients

Structural and climatic variables

Conclusions

Data availability

Author contributions

Acknowledgments

References

E-mail and ORCID