Marcescent Forests of the Iberian Peninsula: Floristic and Climatic Characterization

Abstract

Forests dominated by marcescent oak species represent the transition between deciduous forests adapted to rainy summers and cold winters, on the one hand, and evergreen sclerophyllous Mediterranean forests. In the Iberian Peninsula marcescence is shown by some oak species, including Quercus pubescens, Q. pyrenaica, Q. faginea and Q. canariensis; it suggests an old evergreen habit forced to become deciduous by the cold winters. In this paper we analyse the floristic diversity of marcescent forests in the Iberian Peninsula and their proportion of evergreen broad-leaved and sub-Mediterranean species, and relate them to climatic conditions. This analysis uses 494 phytosociological relevés from the Information System of Iberian and Macaronesian Vegetation (SIVIM) and the BIOVEG data-bases. The data-set was submitted to an agglomerative clustering, which produced four clusters. An NMDS gradient analysis was also applied, in order to assess the relationship between the clusters and bioclimatic variables. The hierarchical and syntaxonomical classifications show a high correspondence, as reflected in the dominance of different Quercus species in each cluster. Quercus broteroi and Q. canariensis forests show a higher proportion of evergreen broad-leaved species, while Q. pubescens and Q. faginea forests are characterized by sub-Mediterranean species. As for climatic relationships, Q. broteroi and Q. canariensis forests present the highest values of thermicity (It) and the lowest values of the ombrothermic (Io₂) index. These values indicate their Mediterranean, thermophilous character, which in turn is related to their high proportions of evergreen broad-leaved species. As a conclusion, it can be stated that marcescence is not related to evergreeness, the latter being linked to mediterraneity. Therefore, the idea of considering marcescence as a residual feature of ancient evergreen laurophyll forests is not supported by our results. Those marcescent forests, particularly the basophilous ones, are related to the sub-Mediterranean floristic element.

1 Introduction

The late Tertiary and Pleistocene were times of dramatic climatic changes, which led to massive extinctions and migrations in the European flora. The changes started with progressive aridification and the emergence of the Mediterranean climate in the late Miocene, with episodes such as the Messinian salinity crisis (Hsu et al. 1977). This resulted in a reversal of the rainfall seasonality from a summer to a winter maximum, with the appearance of summer drought along the southern fringe of the European continent. Subsequently, successive ice ages occurred during the Quaternary. All these events caused deep changes and extinctions in this part of the world. Those episodes, combined with the uplift of several mountain ranges throughout the Cenozoic, fostered the differentiation of mountain floras and set up barriers hampering southward migrations in each of the extension phases of the glacial period. The thermophilous laurophyll flora and vegetation of the mid-Tertiary practically disappeared in Europe, while remaining much better preserved on the Atlantic islands (Macaronesia), i.e., the Azores, Madeira and the Canaries (Rodríguez-Sánchez and Arroyo 2011).

Nevertheless, continental Europe has preserved some remains of the laurophyll vegetation (mesophytic palaeoclimatic relicts, in the sense of Honrado et al. 2001), which is generally recognized based on living populations of Rhododendron ponticum ssp. baeticum or Prunus lusitanica (Calleja et al. 2009). Even the more widespread bay laurel (Laurus nobilis) can be related to this group (Rodríguez-Sánchez and Arroyo 2008, 2011; Rodríguez-Sánchez et al. 2010). Some ferns, including Culcita macrocarpa, Davallia canariensis, Stegnogramma pozoi, Woodwardia radicans, Hymenophyllum tunbrigense and Vandenboschia speciosa, are also related to this relict element from the mid-Tertiary. The three southern European peninsulas, i.e., the Iberian, Italian and Balkan, are known to have been refuges for the thermophilous flora, including Quercus species, during the cold periods of the Pleistocene (Dumolin-Lapègue et al. 1997; Brewer et al. 2002; Ferris et al. 1998; Olalde et al. 2002). Nowadays these peninsulas, particularly Iberia, are mainly covered by sclerophyllous, often scrubby vegetation well adapted to the Mediterranean climate. Temperate Europe, on the other hand, is dominated by deciduous summergreen broad-leaved forests that reached their maximum extension by approximately 6,000 BP (Brewer et al. 2002). These forests now extend over most of Europe and are adapted to a seasonal climate with rainy summers and more or less cold winters. The transition between these forests and the evergreen Mediterranean vegetation is formed by forests in which the dominant species are marcescent oaks, whose leaves wither in autumn but remain dry on the branches until spring, when they are shed. This practice is shown by many European oak species: Quercus pubescens, Q. pyrenaica, Q. faginea, Q. canariensis, Q. cerris, Q. frainetto, Q. ithaburensis, Q. trojana, etc., and resembles the old evergreen practice, which has been forced into deciduousness by the cold winters. In fact, one of these species, Q. canariensis, is even “almost evergreen”: it keeps some of its large leaves functioning during winter until spring, when they are shed immediately before the sprouting of the new leaves.

This woody vegetation dominated by oaks could be interpreted historically as a remnant of the mid-Tertiary vegetation, and these marcescent species as direct descendants of some of the evergreen oaks that covered the mid-latitudes of Europe, as they still do on other continents, particularly in East Asia (Japan, China, and southernmost Korea).

We have carried out a survey of these forests in the Iberian Peninsula, analyzing their flora and the climatic conditions they are linked to. We highlighted the evergreen broad-leaved species living in the canopy or understory of these forests, as they can be used as indicators of evergreens in the communities of Tertiary origin. Many of these species are now considered to be Mediterranean species, as they show a certain sclerophylly (hardness in the leaves); in the context of Mediterranean vegetation, though, they always occur in wet or rainy areas. Another relevant group of plants is the sub-Mediterranean forest species, considered to be the genuine or characteristic species of these forests.

The aims of our study are: (1) to analyse the floristic diversity of the marcescent forests of the Iberian Peninsula; (2) to analyse the proportion of evergreen broad-leaved and sub-Mediterranean species across the different marcescent forest types; and (3) to determine the climatic features linked to different marcescent forests.

We consider the Iberian Peninsula as particularly well suited for performing this survey because of the climatic evolution of the post-glacial Holocene period (Benito Garzón et al. 2007; Rodríguez-Sánchez et al. 2010) and the current climatic variability, from the wettest extreme near the Atlantic to the most extreme Mediterranean climate (summer drought) in some areas of southeastern Spain.

2 Materials and Methods

2.1 Study Area

The study area (Fig. 1) covers the Iberian Peninsula (ca 582,000 km²), located in the transition zone between the Euro-Siberian and Mediterranean biogeographic regions. The climate changes eastward, from temperate hyper-oceanic to semi-continental mediterranean (Rivas-Martínez 2007). In the transition between the mediterranean and temperate climates, we consider the submediterranean variant. This can be defined as the type of climate in which there is only one summer month with p < 2t, i.e., only one dry month in the warm season. Formally, Rivas-Martínez (2007) included this variant in the temperate macroclimate, since at least two dry months are needed to be included in the mediterranean macroclimate. This submediterranean variant is widespread in the Iberian Peninsula, as well as in other parts of southern Europe, providing a favourable climatic context for this survey. It also seems similar to the concept of warm-temperate deciduous areas in parts of East Asia and southeastern North America.

Fig. 1

Location of the study area (Iberian Peninsula) in southwestern Europe and distribution of forest cluster types resulting from floristic classification. Dark areas correspond to uplands and light areas to lowlands

2.2 Data Collection and Preparation

To construct the primary data-set, phytosociological relevés were obtained from the Information System of the Iberian and Macaronesian Vegetation (SIVIM, Font et al. 2010) and the BIOVEG data-base of the University of the Basque Country. Relevés dominated (tree cover >3) by the marcescent tree species Quercus pyrenaica, Q. faginea ssp. faginea (from now on Q. faginea), Q. faginea ssp. broteroi (from now on Q. broteroi), Q. pubescens and Q. canariensis were selected. All relevés were compiled using the module QUERCUS of the VEGANA Package (De Cáceres et al. 2003). The primary data-set grouped 1,284 relevés representing 39 associations and five communities (see syntaxonomic scheme in Appendix 1). Floristic nomenclature follows Flora Iberica (Castroviejo et al. 1986–2011), and Flora Europaea (Tutin et al. 1976, 1980) for taxa not included in the former. The references for these sources are listed in Appendix 2.

The effects of possible oversampling of some areas or particular vegetation types were reduced by a stratified re-sampling (Knollová et al. 2005). The data-set was stratified geographically in such a way that only one relevé for each type of dominant tree was selected for each 10 × 10 km UTM grid.

After stratified re-sampling, a floristic homogeneization was performed by removing bryophytes (many relevés have no records of this group) and by assuming that Hedera helix s.l. includes Hedera helix L., Hedera maderensis K. Koch ex A. Rutherf. ssp. iberica McAllister and Hedera hibernica (G. Kirchn.) Bean; that Rubus sp. includes Rubus sect. Corylifolii Lindl. and Rubus sect. Rubus; that Lonicera periclymenum s.l., Juniperus oxycedrus s.l. and Sanguisorba minor s.l., include all subspecies within each genus in the territory; and that Viola gr. sylvestris includes Viola riviniana Rchb. and V. reichenbachiana Jord. ex Boreau. Quercus faginea Lam. includes Q. alpestris Boiss, and Q. pubescens Willd. includes Q. subpyrenaica Villar. Finally, all Quercus hybrids were included in the parent species dominant in the relevé or in the surroundings. So the data-set was reduced to 494 relevés and 1,003 species.

Climatic variables were calculated from the Digital Climatic Atlas of the Iberian Peninsula (Ninyerola et al. 2005), using a grid of 200 m pixel size generated from the existing network of meteorological stations and the official digital elevation model (DEM) of 100 m pixel size. The mean values of annual temperature (°C), annual rainfall (mm) and summer rainfall (sum of rainfall in June, July and August) per 10 × 10 km UTM cell were calculated, as well as five bioclimatic indices proposed by Rivas-Martínez (2007): continentality index (Ic), thermicity index (It) and the ombrothermic index of the warmest 1, 2 and 3-month periods (Ios₁, Ios₂, Ios₃, respectively). In order to reduce colinearity in potential explanatory climatic variables, a Pearson correlation matrix was calculated and variables with pairwise |r| > 0.75 were eliminated. In this way, mean annual temperature, summer rainfall, and Ios₁ and Ios₃ were seen as superfluous and excluded.

Ordinal species abundance estimates were replaced by metric average values prior to data analysis (r and +: 0.5%. 1: 5%. 2: 17.5%. 3: 37.5%. 4: 62.5%. 5: 87.5%). Percentages of evergreen broad-leaved species and sub-Mediterranean species were calculated in each relevé and normalized by an arcsine-square-root transformation, as recommended by Sokal and Rohlf (1987) for proportional data. Species present only in three relevés or fewer were deleted (0.6% of the total number). In this way, a definitive matrix with 494 relevés and 521 species was obtained.

2.3 Statistical Analysis

The GINKGO program of the VEGANA package (De Cáceres et al. 2003) was used to perform the floristic-numerical analysis. After transformation of cover values, the relevé data were √-transformed to reduce the impact of large cover values in the subsequent floristic analysis (McCune and Grace 2002). A dissimilarity matrix was constructed using Bray-Curtis distance. The definitive data-set was submitted to agglomerative clustering, according to species composition, using the hierarchical β-flexible cluster algorithm (Lance and Williams 1967; Wesche and von Wehrden 2011), with β = −0.25.

The Ochiai fidelity index (De Cáceres et al. 2008) of each species, for the four clusters from the numerical classification, was calculated in order to decide which species are the most diagnostic for each cluster.

Gradient analysis by NMDS (Kruskal 1964a, b; Mather 1976) was applied to assess the relationship between clusters and bioclimatic variables. NMDS has the advantage of not relying on a species response-curve model and can be used with various dissimilarity measures (McCune and Grace 2002; Podani 2006). Correlations between ordination axis and variables were evaluated by Pearson’s correlation coefficient (r).

Relationships between the clusters and the proportion of both evergreen broad-leaved species (EB) and sub-Mediterranean species (SM) were assessed by the non-parametric Kruskal-Wallis test. The same test was used to evaluate the affinities between clusters and bioclimatic indices. When differences among the clusters were significant (p < 0.05), Mann-Whitney’s U-test was applied to determine the categories among which those differences were given. All univariate statistical analyses were performed using SPSS 19.0 statistics package.

3 Results

3.1 Forest Classification

Numerical analysis leads to a classification into four main clusters at dissimilarity level 3.3 (Fig. 2), which reflects the syntaxonomical classification. Cluster 1 includes all relevés dominated by Quercus pyrenaica belonging to the Quercion pyrenaicae alliance (98.5%), except those of the Arbuto unedonis-Quercetum pyrenaicae dominated by Quercus broteroi. Cluster 2 comprises relevés corresponding to the Quercion pubescenti-petraeae (75.6%) and Quercion roboris (17.1%) alliances and some relevés of the Aceri-Quercion fagineae dominated by Quercus pubescens. Relevés of the Aceri-Quercion fagineae dominated by Quercus faginea are mainly grouped in cluster 3, and those corresponding to the Quercion broteroi and Querco-Oleion sylvestris alliances inside the Quercetea ilicis class are grouped in cluster 4. In this latter the dominant trees are Quercus broteroi and Q. canariensis, but some relevés from the eastern coastal mountains, dominated by Quercus faginea or Quercus pubescens and assigned to Violo willkommii-Quercetum fagineae, are also included. Table 1 shows the correspondence between clusters, dominant Quercus species and alliance.

Fig. 2

Dendrogram resulting from the agglomerative clustering analysis explained in the main text. Principal discriminant species, based on their values of the Ochiai fidelity index, are shown. Pearson’s r cophenetic correlation: 0.758; Spearman’s r cophenetic correlation: 0.706; Gower Distance (Stress 1): 1472825.37261

Table 1 Correspondence between clusters obtained from hierarchical numerical analysis, tree dominance and alliances for marcescent and related forests in the Iberian Peninsula. The correspondence (%) between clusters, dominant Quercus species and alliances is shown in parentheses. N = number of relevés; % EB = percentage of evergreen broad-leaved woody species (median values and interquartile range; X² = 198.89; df = 3; p < 0.001). % SM = percentage of sub-Mediterranean species (median values and interquartile range; X² = 127.1; df = 3; p < 0.001). The values followed by the same letter (a) within a column do not differ at p < 0.05, based on a non-parametrical Kruskal-Wallis test and a post hoc Mann-Whitney U-test

The geographical distribution of relevés is represented in Fig. 1, with different symbols for each cluster. Quercus pyrenaica, Q. pubescens and Q. faginea forests assigned to the Querco-Fagetea class have a sub-Mediterranean distribution, whereas Q. broteroi and Q. canariensis forests have a Mediterranean distribution and are included in the Quercetea ilicis class. Quercus pyrenaica forests (Cluster 1) occur mainly throughout the western half of the Iberian Peninsula, associated with acidic soils. Cluster 2 includes Quercus pubescens forests from the Pyrenees. Basophilous forests of Quercus faginea (Cluster 3) are found in the mountains of the eastern half of the Peninsula. Ecotones are frequent mainly in the north, where the presence and even dominance of hybrids is common.

The classification of the four clusters, with statistically determined diagnostic species for each cluster, is presented in Table 2. Cluster 1 is well defined by Quercus pyrenaica (0.97), Clinopodium vulgare (0.65), Pteridium aquilinum (0.64), Holcus mollis (0.63), Teucrium scorodonia (0.63), Arenaria montana (0.59), Luzula forsteri (0.59), Melampyrum pratense (0.57), Lonicera periclymenum s.l. (0.55), Physospermum cornubiense (0.52) and Poa nemoralis (0.50). This cluster is characterized by acidophilous species, which supports their inclusion in the Quercetalia roboris order.

Table 2 Floristic interpretation of the clusters obtained (see main text). The value of the Ochiai fidelity index is shown for each cluster of the most frequent species

Clusters 2 and 3 share some diagnostic species with sub-Mediterranean distributions, such as Viburnum lantana, Primula veris, Acer monspessulanum, Buxus sempervirens (classified as EB), Helleborus foetidus and Cytisophyllum sessilifolium, together with other species with a wider distribution, such as Acer campestre, Cornus sanguinea, Juniperus communis, Lonicera xylosteum, Corylus avellana and Amelanchier ovalis. The characteristic species that separate the two groups of relevés are:

• For cluster 2, Quercus pubescens (0.92), Emerus major (0.50) and Acer opalus (0.30) among sub-Mediterranean species and Galium pumilum (0.38) and Campanula trachelium (0.35) among widespread species; and

• For cluster 3, Quercus faginea (0.91), Teucrium chamaedrys (0.50), Genista scorpius (0.47), Bupleurum rigidum (0.44) and the evergreen broad-leaved species Rubia peregrina (0.58) and Quercus rotundifolia (0.51) (see Table 2).

Finally, cluster 4 is characterized by Quercus broteroi (0.66), Quercus canariensis (0.54), Asparagus acutifolius (0.43), Arisarum simorrhinum (0.38) and a large group of evergreen broad-leaved species such as Pistacia lentiscus (0.63), Viburnum tinus (0.62), Phillyrea latifolia (0.61), Smilax aspera (0.60), Phillyrea angustifolia (0.56), Arbutus unedo (0.55), Chamaerops humilis (0.54), Lonicera implexa (0.50), Ruscus aculeatus (0.47), Rosa sempervirens (0.43) and Quercus suber (0.39). This cluster is quite well defined by thermophilous evergreen broad-leaved species, many of which are sclerophyllous, together with the absence of nemoral plants of the Querco-Fagetea class. In fact, these forests belonging to the Quercion broteroi and Querco-Oleion sylvestris alliances have traditionally been assigned to the Quercetea ilicis class.

3.2 Floristic Analysis

Table 1 summarizes, for each cluster, the proportion of evergreen broad-leaved species (EB) and sub-Mediterranean species (SM) present in the relevés. The highest average percentage of EB species (40%) is in cluster 4, which comprises Quercetea ilicis forests (65.5%) and those of the Aceri-Quercion fagineae alliance (20%) located in the warmest areas. The average percentages of Q. pubescens (cluster 2: 10.4%) and Q. faginea (cluster 3: 12.8%) forests are quite similar. Cluster 1 shows the lowest mean value (3%). For SM species, clusters 2 and 3 have the highest values (mean of 12% and 11% respectively), followed by Q pyrenaica forests (7.7%), and cluster 4 (2.7%). The Kruskal-Wallis test was significant for the percentage of EB and SM species (p < 0.001). Comparison of clusters (Fig. 3) shows that there is no difference between clusters 2 and 3 for both types of species (EB and SM), but clusters 1 and 3 differ significantly from them and also between them for both variables (Table 1).

Fig. 3

Comparison of clusters obtained from the clustering analysis (see main text), using the proportions of evergreen broad-leaved species (EB) and sub-Mediterranean species (SM). Boxes and whiskers include 25–75% and 5–95% of the observed values, respectively, and lines inside the boxes are medians. Clusters marked with the same letter (a) do not differ at p < 0.05, based on a non-parametrical Kruskal-Wallis test and a post hoc Mann-Whitney U-test

3.3 Climatic Analysis

Climatic characterization of the four clusters is summarized in Table 3, and clusters for climatic parameters and indices are compared in the box diagram of Fig. 4. The Mediterranean character of a climate is expressed by low values of Ios₂ (the 2-month summer ombrothermic index: Rivas-Martínez 2007). Quercus pubescens forests (cluster 2) show the highest median values (3.0), corresponding to a temperate climate, with significant differences (p < 0.001) with respect to the other three clusters, which are linked to a more mediterranean climate. Cluster 4 shows the lowest value, which indicates its mediterranean character; clusters 1 and 3 have intermediate values, in agreement with a submediterranean climate. The annual ombrothermic index (Io) also shows significant differences among clusters, but in this case Quercus pyrenaica forests (cluster 1) show values similar to those of Q. pubescens forests (cluster 2), indicating their affinity for the humid ombrotype; clusters 3 and 4 are related to the subhumid ombrotype (upper and lower, respectively). There is no difference, however, between clusters 3 and 4 in terms of annual precipitation (P), which suggests that temperature controls the dominance by Q. faginea, Q. broteroi or Q. canariensis.

Table 3 Median and interquartile ranges (50% of cases, in parentheses) of climatic parameters for the clusters from hierarchical numerical classification (see main text). Values followed by the same letter (a,b) within a row do not differ at p < 0.05, based on a non-parametrical Kruskal-Wallis test (X²) and a post hoc Mann-Whitney U-test. N = 494. df = 3. *** = p < 0.001. ** = p < 0.01

Fig. 4

Comparison of the clusters obtained (see main text) using climatic parameters. Boxes and whiskers include 25–75% and 5–95% of the observed values, respectively, and lines inside the boxes are medians. Clusters marked with the same letter (a,b) do not differ at p < 0.05, based on a non-parametrical Kruskal-Wallis test and a post hoc Mann-Whitney U-test

Regarding temperature, cluster 4 bears the highest thermicity index (353) and is significantly different from the other three clusters, which have similar values. This reflects the location of relevés belonging to cluster 4 in the thermo-Mediterranean belt. Continentality does not show any clear differences between clusters (X² = 16.86).

The diagrams made by NMDS analysis show a clear segregation of the clusters obtained in the hierarchical classification, along axes 1, 2 and 3 (Figs. 5 and 6). These three axes are correlated with climatic indices (Table 4), explaining 33.49% of the accumulated variance. Axis 1 explains the highest accumulated variance (18.25%) and is negatively correlated with annual precipitation (P; r = −0.319; p < 0.001) and ombrothermic index (Io; r = −0.313; p < 0.001). Axis 2 shows a large negative correlation with Ios₂ (r = −0.619; p < 0.001). The diagram representing axes 1 and 2 separates clusters 1, 2 and 3 (Fig. 5). The relevés of cluster 1 (Q. pyrenaica forests) are grouped in the first quadrant of the diagram, linked to higher precipitation. The position of relevés of cluster 2 (Q. pubescens forests) indicates their temperate character, due to the positive relation with Ios2. Axis 3, which is negatively correlated with the thermicity index (It; r = −0.702; p < 0.001), involves relevés of cluster 4 (Fig. 6).

Fig. 5

NMDS ordination of the data-set (see main text), showing axes 1 and 2. G1 – Cluster 1, G2 – Cluster 2, G3 – Cluster 3, G4 – Cluster 4

Fig. 6

NMDS ordination of the data-set (see main text), showing axes 1 and 3. G1 – Cluster 1, G2 – Cluster 2, G3 – Cluster 3, G4 – Cluster 4

Table 4 Pearson’s correlation coefficients for climatic variables with axes from NMDS ordination analysis (see main text). Accumulated variance explained (Var) by the first three axis = 33.49%. *= p < 0.05; ** = p < 0.01; *** = p < 0.001; ns = not significant

4 Discussion

The results of the numerical classification mostly match current syntaxonomy (Rivas-Martínez 2011). Four groups are clearly differentiated, each related to the dominance of one or more Quercus species: Quercus pyrenaica (cluster 1), Q. pubescens (cluster 2), Q. faginea (cluster 3) and Q. broteroi plus Q. canariensis (cluster 4). Nevertheless, some differences can be seen: the associations Pteridio aquilini-Quercetum pubescentis and Carici depauperatae-Q pubescentis, currently in Quercion roboris (Quercetalia roboris), are included in cluster 2, and we propose to change them from Quercion roboris to Quercion pubescenti-petraeae (Quercetalia pubescenti-petraeae).

Relevés of the Quercion broteroi and Querco-Oleion sylvestris alliances are grouped together into cluster 4, due to the high number of Quercetea ilicis species that they share. Thus we accept their current assignment to this class, which groups both sclerophyllous and marcescent forests from Mediterranean Ibero-Atlantic territories with dry to hyperhumid ombrotypes (Rivas-Martínez 2011). Some relevés dominated by Q. faginea and Q. pubescens from the eastern coast of the Iberian Peninsula are also included in this cluster 4. These relevés belong to the Fraxino orni-Quercetum fagineae and Violo willkommi-Quercetum fagineae associations in the Aceri-Quercion fagineae alliance.

Fraxino orni-Quercetum fagineae was described by Rivas Goday et al. (1960), who included it in the class Quercetea ilicis due to the presence of the Mediterranean species Lonicera etrusca, Arbutus unedo, Viburnum tinus, and Juniperus oxycedrus. They admitted, however, that it was quite deviant, as shown by the presence of mesophytic species such as Acer granatense, A. monspessulanum, Quercus faginea, Rhamnus infectoria, and Colutea arborescens. Later on, Rivas-Martínez (1972) proposed that the new syntaxon Aceri-Quercion fagineae within Quercetalia pubescenti-petraeae should go together with Q. faginea forests and other mixed sub-Mediterranean Iberian forests developed on base-rich soils. This new syntaxon would occupy, according to Rivas-Martínez (1972), the southern and westernmost areas within the range of Quercetalia pubescentis. He also proposed the Fraxino orni-Quercetum fagineae association as the type of the new syntaxon. On the other hand, the subassociation Violo willkommii-Quercetum fagineae asparagetosum acutifolii includes marcescent forests from Catalonian coastal ranges (Royo 2006). It is characterized by the presence of many evergreen character species of Quercetea ilicis (Viburnum tinus, Smilax aspera, Phillyrea angustifolia, Arbutus unedo, Lonicera implexa, and Asparagus acutifolius).

The syntaxonomic location of Fraxino orni-Quercetum fagineae and Violo willkommii-Quercetum fagineae asparagetosum acutifolii might not be very appropriate, and their future reassignment to the class Quercetea ilicis should be considered.

The almost complete correspondence of the numerical classification with dominance by a particular tree species indicates the existence of a companion flora for these forests. This fits with the syntaxonomic classification, in which the alliances are characterized by tree species together with other diagnostic species (Rivas-Martínez et al 2002).

Quercus pubescens and Q. faginea forests have many diagnostic species in common, many of them occurring in sub-Mediterranean areas. Most of the diagnostic species common to clusters 2 and 3 are deciduous, whereas the proportion of evergreen broad-leaved species is only 10% and 12% respectively, not enough to support the idea that these forests originated from the evergreen Tertiary forests.

The evergreen broad-leaved species are linked to cluster 4, which includes the most thermophilous, mediterranean forests, mainly Q. canariensis and Q. broteroi forests belonging to Quercetea ilicis, as stated in the syntaxonomic discussion. In fact, these forests hold many of the laurophilous Tertiary species that became adapted to submediterranean conditions (Viburnum tinus, Arbutus unedo, Myrtus communis and others) (Costa et al. 2015). Quercus broteroi and Q. canariensis have been defined as semi-deciduous species with Mediterranean distributions (Sánchez de Dios et al. 2009), in contrast to Q. faginea and Q. pyrenaica, which are marcescent and have sub-Mediterranean distributions. Abadia et al. (1996) also considered that it is the ability to cope with cold winters that differentiates submediterranean marcescent forests from thermophilous Mediterranean forests. In fact, in thermo-mediterranean areas, Q. broteroi and Q. canariensis can only live in especially humid places, such as ravines or north-exposed cloud forests (Cantó 2004; Rivas-Martínez et al. 1990; Pinto and Paiva 2005), where they replace evergreen sclerophyllous forests.

The small number of sub-Mediterranean and especially of evergreen broad-leaved species in the Quercus pyrenaica forests, grouped in cluster 1, may be caused by the acidic soils combined with cooler conditions and high annual rainfall. The Q. pyrenaica forests in the south, where it is warmer, have more evergreen broad-leaved species (Cano and Valle 1990; Pereira 2009). This confirms the relation of such plants with warm climates. A comparative study of these forests together with temperate and true Mediterranean acidophilous forests would be of interest, in order to find out how the submediterranean character is reflected in their floristic composition.

The optimum for marcescent forests seems to be linked to sub-Mediterranean areas characterized by an attenuated summer drought with only one arid month in the warm season (Rivas-Martínez 2007). Studies on forests in Castile and Leon indicated that the ombrothermic indices (Io and Ios) are the best bioclimatic indicators for discriminating between marcescent forests (Del Río and Penas 2006). They concluded that the annual ombrothermic index (Io) discriminates between Q. pyrenaica and Q. faginea forests, the former requiring more humidity than the latter. This agrees with our results, which show significant differences between these forests and also between Q. faginea and Q. pubescens forests. On the contrary, regarding Ios₂, there is no difference between Q. faginea and Q. pyrenaica forests but there are significant differences between these forests and Q. pubescens forests, which suffer only one month of summer drought. Values of the continentality and thermicity indices do not show differences among these three forests types. Q. canariensis and Q. broteroi forests, though, differ clearly from the rest in temperature levels, and this fact could explain the relict character of these latter forests in the Mediterranean area.

Marcescence is thus an adaptation to submediterranean areas with winter frost. Many authors have studied the advantages of marcescent vegetation when competing with deciduous or evergreen forests. Montserrat Martí et al. (2004) attributed it to an early budding in spring and delayed leaf senescence in autumn. Marcescent leaves are capable of some photosynthetic activity in the last 1–2 months of the growing season, unlike senescent ones (Abadía et al. 1996). The companion flora of these marcescent forests is also adapted to a submediterranean and somewhat continental climate, as seen by its distribution; it may have been associated with deciduous forests that could have migrated to the south in the Quaternary, since the Pyrenees may not have formed a strong barrier to colonization after the last ice age (Petit et al. 2003).

Quercus broteroi and Q. canariensis forests have high proportions of evergreen species and are related to thermophilous areas with clear summer drought. Quercus canariensis is considered to be a semi-deciduous tree (Bingre et al. 2007), as it keeps some of its large leaves functional during winter and then sheds them in spring immediately before the sprouting of the new leaves. This adaptation to mild winters may suggest an evergreen laurophyllous origin partially adapted to summer drought and thus sheltered in especially humid biotopes in thermo-mediterranean territories.

As a conclusion of this survey, it can be stated that the marcescent condition is not related to evergreeness. The abundance of evergreen plants is linked to mediterraneity, and thus the idea of considering marcescence as a residual feature of ancient evergreen laurophyll forests is not supported by our results. Those marcescent forests, particularly basophilous forests, are related to the sub-Mediterranean floristic element.

We propose to move the associations Pteridio aquilini-Quercetum pubescentis and Carici depauperatae-Quercetum pubescentis into Quercion pubescenti-petraeae from the Quercion roboris alliance. We also suggest the inclusion of the syntaxa Fraxino orni-Quercetum fagineae and Violo willkommii-Quercetum fagineae asparagetosum acutifolii in Quercetea ilicis. Nevertheless, further studies considering all the associations of Quercetalia ilicis and Quercetalia pubescentis are necessary.

Appendices

Appendix 1. Syntaxonomical Scheme

Salici purpureae-Populetea nigrae (Rivas-Martínez & Cantó ex Rivas-Martínez, Báscones, T.E. Díaz, Fernández-González & Loidi 1991) Rivas-Martínez & Cantó 2002

Populetalia albae Br.-Bl. ex Tchou 1948

Populion albae Tchou 1948

Fraxino angustifoliae-Ulmenion minoris Rivas-Martínez 1975

1. Querco pyrenaicae-Fraxinetum angustifoliae Rivas Goday 1964 corr. Rivas-Martínez, Fernández-González & A- Molina in Fernández-González & A. Molina 1988

Querco-Fagetea sylvaticae Br.-Bl. & Vlieger in Vlieger 1937

Quercetalia roboris Tüxen 1931

Quercion roboris Malcuit 1929

Quercenion robori-petraeae Rivas-Martínez 1978

2. Carici depauperatae-Quercetum pubescentis Zeller 1959 nom. inv. propos.

3. Carici depressae-Quercetum canariensis O. Bolòs 1959 nom. inv. propos.

4. Lathyro linifolii-Quercetum petraeae (Lapraz 1966) Rivas-Martínez 1983 nom mut. propos.

5. Pteridio aquilini-Quercetum pubescentis O. Bolòs 1983

Quercion pyrenaicae Rivas Goday ex Rivas-Martínez 1965

Quercenion pyrenaicae Rivas-Martínez 1975

6. Adenocarpo decorticantis-Quercetum pyrenaicae Martínez-Parras & Molero 1983

7. Arbuto unedonis-Quercetum pyrenaicae (Rivas Goday in Rivas Goday, Esteve, Galiano, Rigual & Rivas-Martínez 1960) Rivas-Martínez 1987

8. Arisaro vulgare-Quercetum pyrenaicae C. Pinto-Gomes, R. Paiva-Ferreira, C. Aguiar, M. Lousã, C. Costa, M. Ladero & S. Rivas-Martínez 2007

9. Berberido hispanicae-Quercetum pyrenaicae F. Valle, Gómez-Mercado & Mota 1988 nom. mut. propos.

10. Cephalanthero rubrae-Quercetum pyrenaicae O. Bolòs & Vigo in O. Bolòs 1967

11. Festuco merinoi-Quercetum pyrenaicae (Rivas-Martínez & Sánchez-Mata in Sánchez-Mata 1989) Sánchez-Mata 1999 corr. Entrocassi, Gavilán & Sánchez-Mata 2004

12. Genisto falcatae-Quercetum pyrenaicae Penas & T.E. Díaz ex Rivas-Martínez 2002

13. Holco mollis-Quercetum pyrenaicae Br.-Bl., P. Silva & Rozeira 1956

14. Luzulo baeticae-Quercetum pyrenaicae Rivas-Martínez 2002

15. Luzulo forsteri-Quercetum pyrenaicae Rivas-Martínez 1963

16. Pulmonario longifoliae-Quercetum pyrenaicae Oberdorfer & Tüxen in Tüxen & Oberdorfer 1958

17. Sorbo torminalis-Quercetum pyrenaicae Rivas Goday ex Rivas-Martínez 1987

Quercenion robori-pyrenaicae (Br.-Bl., P. Silva & Rozeira 1956) Rivas-Martínez 1975

18. Linario triornithophorae-Quercetum pyrenaicae Rivas-Martínez, T.E. Díaz, F. Prieto, Loidi & Penas 1984

19. Lonicero periclymeni-Quercetum pyrenaicae Rivas-Martínez 2002

20. Melampyro pratensis-Quercetum pyrenaicae Rivas-Martínez ex Rivas-Martínez, T.E. Díaz, F. Prieto, Loidi & Penas 1984

21. Rusco aculeati-Quercetum roboris Br.-Bl. P. Silva & Rozeira 1956

Quercetalia pubescentis Klika 1933

Quercion pubescenti-petraeae Br.-Bl. 1932 nom. mut. propos.

22. Buxo sempervirentis-Quercetum pubescentis Br.-Bl. ex Bannes-Puygiron 1933

23. Buxo sempervirentis-Quercetum subpyrenaicae (O. Bolòs & P. Montserrat 1984) Rivas-Martínez in Rivas-Martínez 2011

24. Roso arvensis-Quercetum pubescentis Rivas-Martínez, Báscones, T.E. Díaz, Fernández-González & Loidi 1991 nom. mut. propos.

Aceri granatensis-Quercion fagineae (Rivas Goday, Rigual & Rivas-Martínez in Rivas Goday, Borja, Esteve, Galiano, Rigual & Rivas-Martínez 1960) Rivas-Martínez 1987

25. Berberido hispanicae-Quercetum alpestris Rivas-Martínez 2011

26. Cephalanthero rubrae-Quercetum fagineae Rivas-Martínez in Rivas Goday, Borja, Esteve, Galiano, Rigual & Rivas-Martínez 1960 corr. Rivas-Martínez 1972

27. Cytiso reverchonii-Quercetum fagineae Inocencio, Alcaraz & Ríos 1998

28. Daphno latifoliae-Aceretum granatensis Rivas-Martínez 1965

29. Fraxino orni-Quercetum fagineae Rivas Goday & Rigual in Rivas Goday, Borja, Esteve, Galiano, Rigual & Rivas-Martínez 1960 corr. Rivas-Martínez 1972

30. Pulmonario longifoliae-Quercetum fagineae Loidi & Herrera 1990

31. Sileno melliferae-Quercetum faginae Rivas Goday & Borja in Rivas Goday, Borja, Esteve, Galiano, Rigual & Rivas-Martínez 1960 corr. Rivas-Martínez, T.E. Díaz, Fernández-González, Izco, Loidi, Lousã & Penas 2002

32. Spiraeo obovatae-Quercetum fagineae O. Bolòs & P. Montserrat 1984

33. Telino patentis-Quercetum fagineae Rivas Goday & Borja (1960) 1961 corr. Rivas-Martínez 2011

34. Vinco difformis-Quercetum fagineae Pérez Latorre & Cabezudo 2009

35. Violo willkommii-Quercetum fagineae Br.-Bl. & O. Bolòs 1950 corr. Rivas-Martínez 1972

Quercetea ilicis Br.-Bl. ex A. & O. Bolòs 1950

Quercetalia ilicis Br.-Bl. ex Molinier 1934 em. Rivas-Martínez 1975

Quercion broteroi Br.-Bl., P. Silva & Rozeira 1956 em. Rivas-Martínez 1975 corr. Ladero 1974

Quercenion broteroi Rivas-Martínez, Costa & Izco 1986 corr. Rivas-Martínez 1987

36. Arisaro-Quercetum broteroi Br.-B., P. Silva & Rozeira 1956 corr. Rivas-Martínez 1975

37. Euphorbio monchiquensis-Quercetum canariensis Malato-Beliz in Rivas-Martínez, Lousã, T.E. Díaz, Fernández-González & J.C. Costa 1990

38. Pistacio terebinthi-Quercetum broteroi Rivas Goday in Rivas Goday, Borja, Esteve, Galiano, Rigual & Rivas-Martínez 1960

39. Quercetum alpestris-broteroi Carlos J. Pinto & Rodrigo J.P. Paiva 2005

40. Sanguisorbo hybridae-Quercetum broteroi M. Pereira 2009

41. Viburno tini-Quercetum alpestris Torres & Cano in Cano & al. 2002 corr. Rivas-Martínez 2011

Querco rotundifoliae-Oleion sylvestris Barbéro, Quézel & Rivas-Martínez in Rivas-Martínez, Costa & Izco 1986

42. Oleo sylvestris-Quercetum alpestris Galán, A.V. Pérez & Cabezudo in A.V. Pérez, Galán, P. Navas, D. Navas, Y. Gil & Cabezudo 1999 corr. Rivas-Martínez 2011

43. Rusco hypophylli-Quercetum canariensis Rivas-Martínez 1975

Appendix 2. Bibliographic Sources of the Relevés

Aguilella, A. (1985). Flora y vegetación de la Sierra del Toro y las Navas de Torrijas; Estribaciones sudorientales del macizo de Javalambre. Tesi doctoral. Universitat de València.

Alonso, R. (2002). Valoración del estado de conservación de la vegetación y propuestas de ordenación y uso del territorio de la margen izquierda de la cuenca alta del río Esla (León). Servicio de publicaciones de la Universidad de León.

Amigo, J. & Romero, M.I. (1994). Vegetación atlántica bajo clima mediterráneo: un caso en el noroeste ibérico. Phytocoenologia, 22(4):583–603.

Amigo, J., Izco, J., Guitián, J. & Romero, M.I. (1998). Reinterpretación del robledal termófilo galaico-portugués: Rusco aculeati-Querc. Lazaroa, 19:85–98.

Amor, A., Ladero, M. & Valle, C.J. (1993). Flora y vegetación vascular de la comarca de La Vera y laderas meridionales de la Sierra de Tormantos (Cáceres, España). Studia Botanica 11: 11–207.

Ascaso Martorell, J. (1990). Estudio fitocenológico y valoración de los recursos pastorales de las zonas forestales y arbustivas del Prepirineo aragonés. Institución Fernando el Católico, 152 pp. Zaragoza.

Asensi, A., Díez-Garretas, B. & Nieto, J. M. (2005). Torcal de Antequera - Desfiladero de los Gaitanes. Guia geobotánica. XX Jornadas internacionales de fitosociología. Academia malagueña de Ciencias. 87 pp. Málaga.

Barrera Martínez, I. (1985). Contribución al estudio de la flora y de la vegetación de la Sierra de Albarracín. Tesis Doctoral. Universidad Complutense de Madrid.

Báscones, J.C. (1978). Relaciones suelo-vegetación en la Navarra húmeda del noroeste. Estudio florístico-ecológico. Tesis Doctoral. Universidad de Navarra

Baudière, A. (1974). Contribution à l'étude structurale des forêts des Pyrénées-Orientales: hêtraies. Coll. Phytosociol. 3:17–44.

Belmonte López, D. (1986). Flora y vegetación del Parque Natural de Monfragüe; Tesis Doctoral. Universidad Complutense de Madrid.

Benito Alonso, J. L. (2006). Vegetación del parque nacional de Ordesa y Monte Perdido (Sobarbe, Pirineo central aragonés). Publicaciones del Consejo de Protección de la Naturaleza de Aragon. Zaragoza.

Bolòs, O. de & Montserrat, P. (1983). Datos sobre algunas comunidades vegetales, principalmente de los pirineos de Aragón y de Navarra. Lazaroa, 5:89–96.

Bolòs, O. de (1954). De Vegetatione Notulae, I. Collect. Bot., 4(2): 253–286.

Bolòs, O. de (1959). El paisatge vegetal de dues comarques naturals: la Selva i la Plana de Vic. Arx. Secc. Ciènc. 26:1–175.

Bolòs, O. de (1967). Comunidades vegetales de las comarca próximas al litoral situadas entre los rios Llobregat y Segura. Mem. R. Acad. Cienc. Art. Barcelona, 38(1):3–281.

Bolòs, O. de (1983). La vegetació del Montseny. Servei de Parcs Naturals. Diputació de Barcelona. Barcelona.

Bolòs, O. de (1988). La roureda acidòfila (Quercion robori-petraeae) a Catalunya. Monogr. Inst. Pir. Ecol. (Hom. a P. Montserrat), 4:447–453.

Bolòs, O. de (1996). Contribució al coneixement de la vegetació del territori Auso-Segàrric. Mem. R. Acad. Cienc. Art. Barcelona, LV 4:147–272.

Braun-Blanquet, J. (1967). Vegetationsskizzen aus dem Baskenland mit Ausblicken auf das weitere Ibero-Atlantikum. II Teil. Vegetatio, 14(1–4):1–126.

Cano, E., Pinto-Gomes, C., Valle, F., Torres, J.A., García Fuentes, A., Melendo, M. & Mendes, S. (2002). Primera aproximación al conocimiento de los quejigares del sur de la Península Ibérica (Portugal y España). Quercetea, 3:175–182

Cano, E. & Valle, F. (1990). Formaciones boscosas en Sierra Morena Oriental (Andalucía, España). Acta Bot. Malacitana, 15:231–237.

Cantó, P. (2004). Estudio fitosociológico y biogeográfico de la sierra de San Vicente y tramo i nferior del valle del Alberche (Toledo, España). Lazaroa, 25:187–249

Carreras, J. Carrillo, E. Ninot, J.M. & Vigo, J. (1997). Contribution to the phytocenological knowledge of Pyrenean forest. Fragm. Flor. Geobot., 42(1):95–129.

Carreras, J., Carrillo, E., Font, X., Ninot, J.M., Soriano, I. & Vigo, J. (1995). La vegetación de las sierras prepirenaicas situadas entre los ríos Segre y Llobregat. 1- Comunidades forestales (bosques, mantos marginales y orlas herbáceas). Ecol. Medit., 21(3/4):21–73.

Carreras, J., Carrillo, E., Masalles, R.M., Ninot, J.M. & Vigo, J. (1993). El poblament vegetal de les valls de Barravés i de Castanesa. I- Flora i Vegetació. Acta Bot. Barcinon., 42:1–392.

Carrillo, E. & Ninot, J. M. (1992). La Flora i la vegetació de les valls d'Espot i de Boí (II). I.E.C. Arx. Secc. Ciènc. 99(2):1–351.

Casas, I., Díaz, R., Echevarría, J. E. & Gavilán, R. (1989). Datos sobre la vegetación de Morata de Tajuña (Madrid, España). Lazaroa, 11:61–76.

Catalán, P. (1987). Geobotánica de las cuencas Bidasoa-Urumea (NO de Navarra-NE de Guipúzcoa). Estudio ecológico de los suelos y de la vegetación de la cuenca de Artikutza (Navarra). Tesis doctoral. Universidad del País Vasco.

Conesa, J.A. (1991). Flora i vegetació de les Serres Marginals Pre-pirinenques compreses entre els rius Segre i Noguera Ribagorçana. Tesi Doctoral. Universitat de Barcelona.

De la Cruz Rot, M. (1994). El paisaje vegetal de la Cuenca del río Henares (Guadalajara). Tesis Doctoral. Universidad de Alcalá de Henares.

Devis Ortega, J. (2006). Flora i vegetació del territori comprès entre el riu Segre i el Port del Comte (Prepirineus catalans, Lleida). Tesi Doctoral. Universitat de Barcelona.

Díez Garretas, B., Cuenca, J. & Asensi, A. (1986). Datos sobre la vegetación del subsector aljíbico (provincia Gaditano-Onubo-Algarviense). Lazaroa, 9:315–332.

El Aallali, A., López Nieto, J. M., Pérez Raya, F. & Molero Mesa, J. (1998). Estudio de la vegetación forestal en la vertiente sur de Sierra Nevada (Alpujarra Alta granadina). Itinera Geobotanica, 11:387–402.

Fernández González, F. (1991). La vegetación del valle del Paular (Sierra de Guadarrama, Madrid), I. Lazaroa, 12:153–272.

Fernández López, C., Carballo, A. & Guixá, R. (1984). Vegetación natural de Jabalcuz-La Pandera (Jaén). Blancoana, 2:17–38.

Fernández Prieto, J.A. & Vázquez, V. M. (1987). Datos sobre los bosques asturianos orocantábricos occidentales. Lazaroa, 7:363–382.

Ferrer Plou, J. J. (1990). Marojales y quejigales del noroeste de la provincia de Teruel. Teruel, 80–81(1):181–194.

Ferrer, J. (1993). Flora y vegetación de las Sierras de Herrera, Cucalón y Fonfría. Naturaleza en Aragón nº4. Gobierno de Aragón. 333 pp. Zaragoza.

Fuente, V. de la (1985). Vegetación orófila del occidente de la provincia de Guadalajara (España). Lazaroa, 8:123–219.

Galán de Mera, A. (1993). Flora y Vegetación de los términos municipales de Alcalá de los Gazules y Medina Sidonia (Cádiz, España). Universidad Complutense de Madrid, Facultad de Farmacia, Departamento de Biología Vegetal II. 534 pp.

García González, M.E. (1990). Flora y vegetación de la sierra del Brezo y de la comarca de la Peña (Palencia). Tesis Doctoral. Universidad de León.

García-Abad Alonso, J.J., Gómez Delgado, M. & Rodríguez Espinosa, V.M. (2009). Cartografía detallada de plantas vasculares en un sector de la Alta Alcarria, Guadalajara. Utilidad en la detección de enclaves naturales de interés. Lazaroa, 30:161–175

García-Baquero, Gonzalo (2005). Flora y vegetación del Alto Oja (Sierra de la Demanda, La Rioja, España). Guineana, 11:1–250.

García-Mijangos, I., Biurrun, I., Darquistade, A., Herrera, M. & Loidi, J. (2004). Nueva cartografía de los hábitats en los lugares de interés comunitario (L.I.C.) fluviales de Navarra. Manual de interpretación de los hábitats. Informe para Viveros y Repoblaciones de Navarra

García-Mijangos, I. (1997). Flora y vegetación de los montes Obarenes (Burgos). Guineana 3:1–457.

Gómez Mercado, F. & Valle, F. (1990). Notas fitosociológicas sobre las comunidades arbóreas de las sierras de Cazorla y Segura. Acta Bot. Malacitana, 15:239–246.

Gómez Navarro, J. (2008). Aportaciones al estudio de la flora y vegetación del extremo NE de la provincia de Albacete y zonas adyacentes de la provincia de Valencia (España). Facultad de Ciències Biològiques, Universitat de València, 2:457–926.

Gonçalves Aguiar, C.F. (2000). Flora e vegetação da Serra de Nogueira e do Parque Natural de Montesinho; Universidade Técnica de Lisboa. Inst. Superior de Agronomia. 688 pp. Lisboa

Grüber, M. (1989). Les chenaies acidiphiles à Quercus petraea et Q. pyrenaica de la partie collinéenne des Hautes Pyrénées. Bull. Soc. Hist. Nat. Toulouse, 125:73–78.

Hernández i Cardona, À. M. (1997). Les plantes i el paisatge vegetal d'Olesa de Montserrat. Publicacions de l'Abadia de Montserrat. Col.lecció Vila d'Olesa 6:151–190.

Herrera, M. (1995). Estudio de la vegetación y flora vascular de la cuenca del río Asón (Cantabria). Guineana, 1:1–453.

Herrero Cembranos, L. (1989). Flora y vegetación de la margen izquierda de la cuenca alta del río Pisuerga (Palencia). Tesis Doctoral. Universidad de León

Inocencio Pretel, C., Alcaraz Ariza, F.J. & Ríos Ruiz, S. (1998). El paisaje vegetal de la cuenca albacetense del Guadalmena. Inst. Est. Albacetenses. 327 pp. Albacete.

Ladero Álvarez, M., Luengo Ugidos, M., Santos Bobillo, M.T., González Iglesias, J., Alonso Beato, M.T. & Sánchez Rodríguez, M.E. (2006). Vegetación del entorno del Balneario de Cervantes, Santa Cruz de Mudela (Ciudad Real). An. R. Acad. Nac. Farm., 72:321–368.

Ladero Álvarez, M.T., Luengo Ubigos, M.A., Santos Bobillo, M.T., Alonso Beato, M.T., Sánchez Rodríguez, M.E., González Iglesias, F.J. & Ladero Santos, I. (2008). Vegetación del entorno del Balneario de Valdelateja, Valle de Sedano (Burgos). An. R. Acad. Farm. 74:541–581.

Llamas Garcia, F. (1984). Flora y vegetación de la Maragateria (León). Institución Fray Bernardino de Sahagún, 273 pp. León.

Llansana, R. (1976). Estudio florístico y geobotánico de la zona comprendida entre Balaguer y els Aspres de la Noguera. Tesis de Llicenciatura Universitat de Barcelona.

Loidi Arregui, J., Biurrun Galarraga, I. & Herrera Gallastegui, M. (1997). La vegetación del centro-septentrional de España. Itinera Geobot., 9:161–618.

Loidi, J. & Herrera, M. (1990). The Quercus pubescens and Quercus faginea forests in the Basque Country (Spain): distribution and typology in relation to climatic factors. Vegetatio, 90:81–92

Loidi, J. & Fernández-Prieto, J. A. (1986). Datos sobre la biogeografía y la vegetación del sector Castellano-Cantábrico (España). Doc. Phytosoc. 10:323–362

López Pacheco, M. J. (1988). Flora y vegetación de las cuencas alta y media del río Curueño (León). Institución Fray Bernardino de Sahagun. Diputación provincial de León. 384 pp. León.

López, G. (1976). Contribución al conocimiento fitosociológico de la Serranía de Cuenca I. Comunidades fruticosas: bosques, matorrales, tomillares y tomillar-praderas. Anales Inst. Bot. Cavanilles, 33:5–87.

Martínez Parras, J. M. & Molero Mesa, J. (1982). Ecología y fitosociología de Quercus pyrenaica Willd. en la provincia Bética. Los melojares béticos y sus etapas de sustitución. Lazaroa, 4:91–104.

Medrano, M. (1994). Flora y Vegetación de las Sierras de la Demanda y Cameros (La Rioja). Tesis Doctoral. Universidad de Navarra.

Melendo Luque, M. (1998). Cartografía y ordenación vegetal de Sierra Morena: Parque Natural de las Sierras de Cardeña y Montoro (Córdoba). Tesis Doctoral. Universidad de Jaén.

Mercadé López, A. (2008). Aportació al coneixement dels boscos mesòfils de la Catalunya central, I; Fagedes i Rouredes. Acta Bot. Barcinon., 51:93–125.

Merle Farinós, H., Ferriol Molina, M. (2008). Aportación al conocimiento de los melojares relictos de Quercus pyrenaica de la Sierra de Espadán (Castellón, España). Lazaroa, 29:125–128.

Molero, J. & Vigo, J. (1981). Aportació al coneixement florístic i geobotànic de la Serra d'Aubenç. Treb. Inst. Bot. Barcelona, 6:1–82.

Molero, J., Sáez, Ll. & Villar, L. (1998). Interés florístico y geobotánico de la sierra de Alcubierre (Monegros, Aragón). Acta Bot. Barcinon., 45:363–390.

Molina Abril, J. A. & Pertíñez, C. (2005). Adiciones al informe final sobre el estudio integrado de la vegetación de ribera del tramo medio del río Cinca. Centro de Estudios de Monzón y Cinca Medio, 32:51–100.

Molina Cantos, R., Valdés Franzi, A. & Alcaraz Ariza, F.J. (2008). Flora y vegetación del tramo medio del valle del río Júcar (Albacete). Instituto de estudios albacetenses 'Don Juan Manuel' de la excma. Diputación de Albacete. 663 pp.

Navarro Andrés, F. & Valle Gutiérrez, C. J. (1983). Fitocenosis fruticosas de las comarcas zamoranas de Tabara, Alba y Aliste. Studia Botanica, 2:69–121.

Navarro Andres, F. (1974). La vegetación de la Sierra del Aramo y sus estribaciones (Asturias). Rev. Fac. Ci. Oviedo, 15:111–243.

Navarro, G. (1986). Vegetación y flora de las Sierra de Urbión, Neila y Cabrejas. Tesis Doctoral. Universidad Complutense Madrid

Navarro, G. (1989). Contribución al conocimiento de la vegetación del Moncayo. Opusc. Bot. Pharm. Complutensis, 5:5–64.

Nieto, J. M. & Cabezudo, B. (1988). Series de vegetación climatófilas de las sierras Tejeda y Almijara (Málaga-Granada, España). Acta Bot. Malacitana, 13:229–260.

Ninot, J.M. Quadrada, R.V. & Carrillo, E. (2009). Vegetació del Massís de la Fembra Morta (Anoia, Catalunya Central). Miscellanea Aqualatensia, 9:11–136.

Nuet, J. (1983). La vegetació de la muntanya dels Mollons, a la comarca d'Anoia. Miscellanea Aqualatensia, 3:15–52.

Onaindia Olalde, M. (1986). Ecología vegetal de las Encartaciones y Macizo del Gorbea, Vizcaya. Universidad del País Vasco, 271 pp. Bilbao.

Penas, A. & Díaz-González, T.E. (1985). Datos sobre la alianza Corynephoro-Plantaginion radicatae Rivas Goday & Rivas-Martínez 1963 nom. invers. Rivas-Martínez 1975 en el sector Orensano-sanabriense. Acta Bot. Malacitana, 10:155–166.

Penas, A. Rivas-Martínez, S. Díaz González, T.E. (2001). Un itinerario botánico por los alrededores de León. Servicio de Publicaciones de la Universidad de León, pp 47–55.

Peralta, J. Báscones, J.C. & Íñiguez, J. (1990). Bosques de la Sierra de Leyre (Navarra-Zaragoza, NE de España. Botánica pirenaico-cantabrica, 5:559–564.

Pereira, M. (2009). A Flora e Vegetação da Serra de Monfurado (Alto Alentejo-Portugal). Guineana, 15:1–316.

Pérez Badia, M.R. (1997). Flora vascular y vegetación de la comarca de la Marina Alta (Alicante). Colección Técnica 16. Instituto de Cultura Juan Gil Albert. Alicante.

Pérez Latorre, A. V., Caballero, G., Casimiro-Soriguer Solanas, F., Gavira, O. & Cabezudo, B. (2009). Vegetación de la Cordillera Antequerana Oriental (Subsector Torcalense). Málaga-Granada (España). Acta Bot. Malacitana, 34:1–29.

Pérez Latorre, A. V., Galán de Mera, A., Deil, U. & Cabezudo, B. (1996). Fitogeografía y vegetación del sector aljíbico (Cádiz-Málaga, España). Acta Bot. Malacitana, 21:241–267.

Pérez Latorre, A. V., Galán de Mera, A., Navas, P., Gil, Y. & Cabezudo, B. (1999). Datos sobre la flora y vegetación del Parque Natural de los Alcornocales (Cádiz-Málaga, España). Acta Bot. Malacitana, 24:133–184.

Pérez Latorre, A. V., Navas Fernández, D., Gavira, O., Caballero, G. & Cabezudo, B. (2004). Vegetación del Parque Natural de Las Sierras Tejeda, Almijara y Alhama (Málaga-Granada, España). Acta Bot. Malacitana, 29:117–190.

Pérez Morales, C. (1988). Flora y vegetación de la cuenca alta del río Bernesga. Institución Fray Bernardino de Sahagún. Diput. Prov. de León. 437 pp. León.

Pinillos López, J.A. (2000). Estudio de la Vegetación y Flora del Campo de Garcimuñoz: Baja y Media Serranía (Cuenca). Tesis Doctoral. Universidad de Valencia.

Pinto Gomes, C. & Pavia Ferreira, R. (2005). Flora e Vegetação do Barrocal Algarvio (Tavira-Portimão). Comissão de Coordenação e Desenvolvimento Regional do Algarve, pp 9–354.

Pinto-Gomes, Carlos., Paiva-Ferreira, R. & Meireles, C. (2007). New Proposals on Portuguese Vegetation. Lazaroa, 28:67–77.

Pitarch García, R. (2002). Estudio de la flora y vegetación de las sierras orientales del Sistema Ibérico: La Palomita, Las Dehesas, El Rayo y Mayabona (Teruel). Publicaciones del Consejo de Protección de la Naturaleza de Aragón. Serie Investigación. 537 pp. Zaragoza.

Puente García, E. (1988). Flora y vegetación de la cuenca alta del río Sil. Instituación Fray Bernardino de Sahagún. Diput. Prov. de León. 536 pp. León

Ramos Teles, R. J. (2005). Caracterização da flora e vegetação do Vale da Ribeira de Almoster; Universidad de Èvora

Rivas Goday, S. & Borja, J. (1961). Estudio de la vegetación y flórula del macizo de Gúdar y Jabalambre. Anales Inst. Bot. A. J. Cavanilles, 19:1–550.

Rivas Goday, S. (1964). Vegetación y flórula de la cuenca extremeña del Guadiana. Publ. Diput. Prov. Badajoz. 777 pp. Badajoz.

Rivas Goday, S. Borja Carbonell, J. Esteve Chueca, F. Fernández-Galiano, E. Rigual Magallon, A. & Rivas Martínez, S. (1959). Contribución al estudio de la Quercetea ilicis hispánica. Anales Inst. Bot. Cavanilles, 17(2):285–403.

Rivas Goday, S. & Madueño Box, M. (1946). Consideraciones acerca de los grados de vegetación del Moncayo y sobre la habitación de las Digitalis purpurea L. y parviflora Jacq. Farmacognosia, 5(9):97–122.

Rivas Martínez, S. (1962). Contribución al estudio fitosociológico de los hayedos españoles. Anales Inst. Bot. A. J. Cavanilles, 20:97–128.

Rivas Martínez, S. (1975). La vegetación de la clase Quercetea ilicis en España y Portugal. Anales Inst. Bot. Cavanilles, 31(2):205–259.

Rivas-Martínez, S., Báscones, J.C., Díaz, T.E. & Fernández-González, F. (1991). Vegetación del Pirineo-Occidental y Navarra. Itinera Geobot., 5:5–455

Rivas-Martínez, S., Díaz, T.E., Fernández-González, F., Izco, J., Loidi, J., Lousa, M. & Penas, A. (2002). Vascular plant communities of Spain and Portugal, addenda to the syntaxonomical checklist of 2001. Itinera Geobot., 15(1):5–432

Rivas-Martínez, S., Diaz, T.E., FernÃndez Prieto, J.A., Loidi, J. & Penas, A. (1984). Los Picos de Europa. Ediciones Leonesas.

Rivas-Martínez, Salvador., Lousa, M., Díaz, T. E., Fernández-González, F. & Costa, J. C. (1990). La vegetación del sur de Portugal (Sado, Alentejo y Algarve). Itinera Geobot., 3:5–126

Romero Abello, A. (1991). Contribución al estudio de la flora y vegetación vascular de las cuencas inferiores de los ríos Arlanza, Arlanzón y Carrión (provincias de Palencia y Burgos, España). Tesis Doctoral. Universidad complutense de Madrid

Romo, A.M. (1989). Flora i vegetació del Montsec (Prepirineus catalans). I.E.C. Arx. Secc. CiÒnc., 90:1–534.

Roselló, R. (1994). Catálogo florístico y vegetación de la comarca natural del Alto Mijares. Diputació de Castelló. Castelló.

Royo Pla, F. (2006). Flora i vegetació de les planes i serres litorals compreses entre el riu Ebro i la serra d'Irta. Tesi Doctoral. Universitat de Barcelona.

Sánchez Mata, D. (1989). Flora y vegetación del macizo oriental de la Sierra de Gredos (Avila). Institución Gran Duque de Alba. Diputación Provincial de Avila. Avila.

Sardinero, S. (2004). Flora y vegetación del macizo occidental de la Sierra de Gredos (Sistema Central, España). Guineana, 10:1–474.

Sobrino Vesperinas, E. & Sanz Elorza, M. (1998). Datos sobre la flora y vegetación de la Sierra de Alcubierre (Huesca). Lagascalia, 20(2):231–237.

Tüxen, R. & Oberdorfer, E. (1958). Die Pflanzenwelts Spaniens. Teil II. Veröff. Geob. Inst. Rübel Zurich, 32. 328 pp.

Valle, Francisco., Gómez-Mercado, F. & Mota, J. F. (1988). Los robledales de la sierra de Segura y otras comunidades relacionadas con ellos. Anales Jard. Bot. Madrid, 45(1):247–257.

Vicente Orellana, J.A. & Galán De Mera, A. (2008). Nuevas aportaciones al conocimiento de la vegetación Luso-Extremadurense. Estudio de las Sierras de Las Villuercas (Extremadura, España) y San Mamede (Alto Alentejo, Portugal). Acta Bot. Malacitana, 33:1–49.

Vigo, J. (1968). La vegetació del massís de Penyagolosa. I.E.C. Arx. Secc. Ciènc. 37:1–247.

Vigo, J. (1996). El poblament vegetal de la vall de Ribes. Institut Cartogràfic de Catalunya, 468 pp. Barcelona.

Vilar, L. & Viñas, X. (1990). Sobre los robledales del Llano de la Selva. Acta Bot. Malacitana, 15:277–281.

Villaescusa Reig, C. (1998). Flora vascular de la comarca de El Baix Maestrat (Castellón). Tesi Doctoral. Universitat de València.

Villegas i Alba, N. (1993). Flora i vegetació de les muntanyes del Puigsacalm-serra de Milany. Tesi doctoral. Universitat de Barcelona

Viñas, X. & Polo, L. (1992). La vegetació dels enclavaments silicis de l'alta Garrotxa. Actes del Simposi Internacional de Botànica Pius Font i Quer, 2:317–329.

Viñas, X. (1993). Flora i vegetació de l'Alta Garrotxa. Tesi Doctoral. Universitat de Girona.

Vives, J. (1964). Vegetación de la alta cuenca del Cardener. Estudio florístico y fitocenológico comarcal. Acta Geobot. Barcinon., 1:1–218.

Wattez, J.-R. (1979). Affinités phytosociologiques de l'alisier torminal (Sorbus torminalis (L.) Crantz) en Picardie occidentale; Doc. Phytosoc. N.S., 4:951–965