Comparative study of Central and Eastern European alliances of thermophilous oak forests (Quercion petraea, Betonico-Quercion, and Lathyro-Quercion) within the temperate zone of Europe

According to the results of the agglomerative cluster analysis, 28 analyzed syntaxa (see Table S1) were divided into three groups, conventionally designated as A, B, C (Figure 1). The cluster A united Central European syntaxa (1–10). The cluster B consists of Eastern European syntaxa with the main distribution in the Middle Russian Upland (11–20). The cluster C unites the South Uralian group of syntaxa (21–28).

Figure 1

The clusters of syntaxa coincide with the division into the alliances of Quercion petraeae (A), Betonico-Quercion (B), Lathyro-Quercion (C). This proves the essential differences in species composition between the Eastern European and Central European thermophilous oak forests (Semenishchenkov, 2012; Semenishchenkov & Poluyanov, 2014; Goncharenko et al., 2020). In addition, the Eastern European group was divided into two parts, namely the cluster B, which occupies an intermediate position between A and C, and the cluster С from the Southern Urals. If we take into account the agglomeration at higher levels, cluster B is still closer to cluster A from Central Europe than to cluster C, which is geographically and floristically distant.

According to the cluster analysis results, the geographic boundaries of cluster B proved wider than previously assumed (Goncharenko et al., 2020). Cluster B also included associations from Central Ukraine, in particular Digitali grandiflorae-Quercetum roboris (11, Figure 1). This extends the notion about the western border of the Betonico-Quercion alliance further west including regions of Central Ukraine in the Dnieper basin (Goncharenko & Kovalenko, 2019). As for syntaxa from Kazakhstan (19 and 20, Figure 1), they also joined the Betonico-Quercion alliance (cluster B). This differs from the authors’ original treatment of these associations because the association Astragalo ciceri-Quercetum roboris (19) was assigned to the Betonico-Quercion alliance, but the association Sanguisorbo officinalis-Quercetum roboris (20) was classified into another alliance Lathyro-Quercion (Kozhevnikova & Prokhorov, 2021). Such a discrepancy may be explained by the inclusion in cluster C (21–28, Figure 1) of specific associations from the Southern Urals with a high content of mountain species. This increased the distance between clusters B and C, and hence lead to the accession of associations from Kazakhstan to the closer cluster B. However, the associations from Kazakhstan (19 and 20) occupy a transitional position in floristic respect between the Betonico-Quercion and the Lathyro-Quercion, as will be shown later.

The floristic differentiation of Central and Eastern European forests is multilevel (Table S2). It is based on 5 groups (blocks) of differential species. In addition to groups A, B, and C, whose species are concentrated in a single cluster, there are groups AB and BC the species of which overlap pairs of clusters, specifically A + B and B + C, respectively. The group ABC consists of common species for all three groups of syntaxa. Each of the mentioned groups is supported by a significant number of differentiating species.

The multilevel floristic differentiation is explained by the presence of not only local ecological but geographical reasons. Group A (Quercion petraeae) concentrates the species associated with Central Europe. The most prominent representatives are such species as Quercus petraea (Mattuschka) Liebl., Ligustrum vulgare L., Lembotropis nigricans (L.) Griseb., Melittis melissophyllum L., Symphytum tuberosum L., Festuca heterophylla Lam., Melica uniflora Retz., Sorbus torminalis (L.) Crantz. and others. Most of them in Ukraine have their eastern distribution border in the basins of the Dniester and South Bug rivers (Meusel et al., 1965; Meusel & Jäger, 1992). But on the other hand, cluster B (Betonico-Quercion) already includes syntaxa from Central Ukraine in the Dnieper basin. Taking both facts into account, we may assume that the boundary between the alliances Quercion petraeae and Betonico-Quercion runs along the eastern boundary of the above-listed Central European species and in coincidence with the boundary of the Central European floristic province in western Ukraine (Meusel & Jäger, 1992).

Cluster B (Betonico-Quercion) occupies an intermediate position between the Central European Quercion petraeae and the South Uralian Lathyro-Quercion groups, as indicated by the presence of the species groups AB and BC (Table S2). Similarly to group A, there are many species restricted in their eastern distribution in groups B and AB. The main reason for this is the increasing continentality of climate. These include species like Allium oleraceum L., Anthericum ramosum L., Carex montana L., Potentilla alba L., Lathyrus niger (L.) Bernh., Serratula tinctoria L. and others. Their eastern distribution boundaries are mostly confined to the Volga basin (Meusel et al., 1965; Meusel & Jäger, 1992). Following the species of Group A, these seem to represent the «second wave», but they do not reach the Southern Urals and therefore are absent from the syntaxa of Lathyro-Quercion (Table S2).

Although associations from Kazakhstan joined the Betonico-Quercion alliance (Figure 1), they have a transitional character between the alliances Betonico-Quercion and Lathyro-Quercion (columns 19 and 20, Table S2). Taking into account the aforesaid, we can assume that the boundary between the alliances Betonico-Quercion and Lathyro-Quercion approximately runs through the Volga valley. Further accumulation of material from continental regions of Russia and Kazakhstan will bring more clarity in this question.

Restrictions on the distribution of eastern species in the reverse western direction are much less expressed. Most of the differentiators of groups C and BC are also found in Ukraine and other European countries. However, the differential species in these groups are associated with other habitats, mainly the meadow-steppe and steppe, due to which they have much less constancy in the composition of thermophilous oak forests.

Thus, there are two trends revealed in Table S2. In the eastward spreading, there are climatic constraints due to harsh continental conditions and especially low winter temperatures, to which western thermophilous species are not adapted. Conversely, in the opposite western direction, restrictions on the distribution of species of groups B and BC are less pronounced, which is explained by a milder climate of Central Europe and therefore, coenotic and edaphic constraints prevail.

In the ordination space, three groups of thermophilous oak forests syntaxa extend mainly along the first axis. The Central European (A) and the South Uralian (C) clusters are located at opposite ends (Figure 2).

Figure 2

The first axis is most correlated with the continentality variable Kn, followed by the temperature variable Tm (Table S3). As mentioned, the increasing continentality of climate may explain the distribution limitations of many thermophilous species towards the east (groups A and AB, Table S2). Edaphic factors, namely soil nitrogen, light regime and moisture, are related more to the second axis (Table S3). A different tendency for the soil acidity vector Rc, which differs from the directions of other vectors of edaphic factors, could be explained by the confinement of thermophilous oak forests to the carbonate substrates, as noted by many authors (Table S1).

Based on the comparison of mean values and results of Tukey’s HSD test, we observed significant differences in the continentality variable Kn in all three groups (Figure 3f).

Figure 3

The mean values of Kn indicator form a monotonic increasing series (39.4, 41.7, 44.0 for clusters A, B, and C, respectively; see Table S4). For other variables, such as soil acidity and nitrogen, significant differences in group means are observed only for the outermost clusters, namely Central European A and South Uralian C, while the intermediate cluster B has no significant differences (shared letters ‘ab’ above Figure 3b, 3c). On the contrary, the two Eastern European clusters B and C are similar concerning the temperature variable, while the cluster A differs from both (letter ‘b’ above Figure 3e). This is explained by the fact that the syntaxa of the cluster A concentrate much more thermophilous species. There are no significant differences in the moisture and light indicators (Figure 3a, 3d), as thermophilous oak forests are similar in this respect in all studied regions.

Although thermophilous oak forests are formally categorised within the class Quercetea pubescentis, they are not typical in northern regions where they grow outside the climatic optimum. So their species composition is a phytosociological admixture of species of different vegetation classes. A comparison of the phytosociological structure of the studied syntaxa showed that they affiliate to four forest vegetation classes (Quercetea pubescentis, Quercetea robori-petraeae, Carpino-Fagetea, Brachypodio-Betuletea pendulae), three classes of herbaceous vegetation (Festuco-Brometea, Molinio-Arrhenatheretea, and Trifolio-Geranietea), and a class of xerophytic shrub vegetation (Crataego-Prunetea) (Table 1).

The shares of species of some classes vary within a narrow range, and this may be both in low numbers, e.g., as for the class Crataego-Prunetea, RHA (0.01–0.06), and in relatively high numbers, e.g., as for Trifolio-Geranietea, GER (0.18–0.26, Table 1). The high richness of Trifolio-Geranietea species can be explained by the growth of the communities in light and warm habitats optimal for this class.

While similar in the sense that all syntaxa are a phytosociological mixture of species in which no single vegetation class plays a dominant role, they nevertheless differ in the proportions in which species of the different classes are combined in the different regions. Major differences between Central European and Eastern European forests are mainly observed not concerning a single class, but if we take into account the proportions of species of the classes Quercetea pubescentis / Brachypodio-Betuletea, PUB/BRA. The first class is more represented in the western syntaxa of cluster A, and the number of species of the second class increases significantly in the eastern syntaxa of cluster C from the Southern Urals. The ratio PUB/BRA of these two classes forms the following series: 0.14/0.07 = 2.14 in cluster A, 0.14/0.11 = 1.27 in cluster B, and 0.07/0.19 = 0.38 in cluster C. Thus, the phytosociological structure of thermophilous oak forests display patterns associated with the proximity of the region to the climatic optimum of one or another class. Thus, the class Quercetea pubescentis has a Mediterranean centre of the climatic optimum (Jakucs, 1960) and the class Brachypodio-Betuletea has the south Siberian centre of diversity (Ermakov et al., 1991). Most likely in the northern regions, thermophilous oak forests represent an aggregate type syntaxonomically, so the question of their placement into only one of the forest vegetation classes cannot be solved. This problem will be explained by the complexity of the classification of ecotone vegetation types and should be considered concerning individual syntaxa.

In addition to floristic, ecological and phytosociological differences between the three groups of thermophilous forests discussed in the previous sections, there is a differentiation in the lists of the main layer-forming species. In Table 2 species that are ranked by the constancy value in each of the clusters A, B and C form a diagonal-like pattern in three main layers.

In the western syntaxa of cluster A, the most constant species are Quercus petraea and Carpinus betulus L. in the tree layer. Moving eastward, in cluster B, they are replaced by Quercus robur L. with Betula pendula Roth. In cluster B the role of Pinus sylvestris L., Pyrus communis L., Acer platanoides L. is also increasing (Table 2). In cluster C, the species Quercus robur comes first, followed by the species Tilia cordata Mill., which has the second rank. Such species as Acer platanoides and Ulmus glabra Huds. have also significant participation, occupying the third and fourth ranks, respectively, in the communities from the Southern Urals.

Differences in the main layer-forming species extend to all main layers. The species Ligustrum vulgare is the most constant species in the shrub layer of the Central European cluster A. The composition of the shrub layer of cluster B is more acidophilous, including such species as Frangula alnus Mill., Sorbus aucuparia L., Cytisus ruthenicus Wol. The constancy of species such as Caragana frutex (L.) C. Koch, Prunus fruticosa Pall. etc. in the shrub layer of cluster C is explained by the formation on the border with petrophytic steppe communities in the Southern Urals (Table S1).

One of the most characteristic features of thermophilous oak forests is a floristically rich herbaceous layer. Although it usually has no clearly expressed dominants, there is also a differentiation in the herbaceous layer, similar to that in the tree and shrub layers. The most constant species in cluster A are Veronica chamaedrys L., Poa nemoralis L., Clinopodium vulgare L. (Table 2). In the Eastern European cluster B they are replaced by Betonica officinalis L., Veronica chamaedrys, Convallaria majalis L., and in cluster C the most constant are species like Origanum vulgare L., Stellaria holostea L., Brachypodium pinnatum (L.) Beauv. (Table 2).

Two factors seem to play a major role in the succession of major layer-forming species. The first one is the degree of tolerance of species to climate continentality, and the second reason is in contact with what types of vegetation the thermophilous oak forests form in each region. Although the species listed in Table 2 are not exclusive to just one of the clusters, their constancies vary greatly depending on whether the climatic conditions are optimal or not for one or another species. This causes a change in the constancy-ranked series of the layer-forming species, as shown in Table 2. Regarding the role of surrounding communities, this environment is different in different regions of Europe, as the ecological conditions and predominant substrates on which communities are formed change, too.