This study aimed at an investigation of the structure, ecology and mapping of mixed communities with the participation of spruce, pine and broad-leave trees in one of the regions of broad-leave–coniferous zone. Despite the long history of the nature use of the study area, including forestry practices (Kurnayev, 1968; Rysin, Saveliyeva, 2007; Arkhipova, 2014; Belyaeva, Popov, 2016), the communities kept the main features of the indigenous forests of the broad-leave–coniferous zone — the tree species polydominance of the stands, the multilayer structure of communities and the high species diversity.
In the course of field works in the southwestern part of the Moscow Region (2000–2016) 120 relevés were made. Spatial structure, species composition as well as cover values (%) of all vascular plants and bryophytes were recorded in each stand.
The relevés were analysed following the ecology-phytocenotic classification approach and methods of multivariate statistical analysis that allowed correctly to differentiate communities according the broad-leave species participation. The accuracy of the classification based on the results of discriminant analysis was 95.8 %. Evaluation of the similarity of the selected units was carried out with the help of cluster analysis (Fig. 12). Clustering into groups is performed according to the activity index of species (A) (Malyshev, 1973) within the allocated syntaxon using Euclidean distance and Ward’s method. The classification results are corrected by DCA ordination in PC-ORD 5.0 (McCune, Mefford, 2006) (Fig. 1). Spatial mapping of forest cover was carried out on the basis of ground data, Landsat satellite images (Landsat 5 TM, 7 ETM +, 8 OLI_TIRS), digital elevation (DEM) and statistical methods (Puzachenko et al., 2014; Chernenkova et al., 2015) (Fig. 13 а, б).
The obtained data and the developed classification refine the existing understanding of the phytocenotic structure of the forest cover of the broad-leave–coniferous zone. Three forest formation groups with different shares of broad-leave species in the canopy with seven groups of associations were described: a) coniferous forests with broad-leave species (small- and broad-herb spruce forests with oak and lime (1)); broad-herb spruce forests with oak and lime (2); small- and broad-herb pine forests with spruce, lime, oak and hazel (3); broad-herb pine forests with lime, oak and hazel (4)), b) broad-leave–coniferous forests (broad-herb spruce–broad-leave forests (5)), and c) broad-leave forests (broad-herb oak forests (6), broad-herb lime forests (7)). In the row of discussed syntaxa from 1 to 7 group, the change in the ratio of coniferous and broad-leave species of the tree layer (A) reflects regular decrease in the participation of spruce in the plant cover (from 66 to 6 %; Fig. 3 A1, A2) and an increase in oak and lime more than threefold (from 15 to 65 %; Fig. 4 a). Nemoral species predominate in the composition of ground layers, the coverage of which increases (from 40 to 80 %) in the range from 1 to 7 group, the coverage of the boreal group varies from 55 to 8 % (Fig. 11) while maintaining the presence of these species, even in nemoral lime and oak forests. In forests with equal share of broad-leave and coniferous trees (group 5) the nemoral species predominate in herb layer. In oak forests (group 6) the species of the nitro group are maximally represented, which is natural for oak forests occurring on rich soils, and also having abundant undergrowth of hazel.
Practically in all studied groups the presence of both coniferous (in particular, spruce) and broad-leave trees in undergrowth (B) and ground layer (C) were present in equal proportions (Fig. 3). This does not confirm the unambiguity of the enrichment with nemoral species and increase in their cover in complex spruce and pine forests in connection with the climate warming in this region, but rather indicates on natural change of the main tree species in the cenopopulations. Further development of the stand and the formation of coniferous or broad-leave communities is conditioned by landscape.
It is proved that the distribution of different types of communities is statistically significant due to the relief. According to the results of the analysis of remote information, the distribution areas of coniferous forests with broad-leave species, mixed and broad-leave forest areas for the study region are represented equally. The largest massifs of broad-leave–coniferous forests are located in the central and western parts of the study area, while in the eastern one the broad-leave forests predominate, that is a confirmation of the zonal ecotone (along the Pakhra River: Petrov, Kuzenkova, 1968) from broad-leave–coniferous forests to broad-leave forests.
Variable tree rooting strategies are key for modelling the distribution, productivity and evapotranspiration of tropical evergreen forests
