Browsing by Subject "Supervised k-means clustering"

Now showing 1 - 1 of 1

Analysis of phytosociological composition and spatial structure of the central zone of Lake Baikal Eastern coast vegetation
(2018) Brianskaia, Elena; Schmieder, Klaus
The object of this study is the terrestrial ecosystem of Lake Baikal enlisted by UNESCO as the World Heritage Site. The analysis of spatial and phytosociological structures of the vegetation can reveal important stages of its formation and future dynamics. Today, the present flora and vegetation of the complex Baikal Siberian ecosystem is reflected in studies of many Russian and international phytosociologists. However, despite the huge amount of data, the phytosociological vegetation structure and its spatial distribution of the central zone of Lake Baikal eastern coast has not been studied. By this thesis, we provide the first results about the flora, phytosociological composition of the vegetation and the soil diversity of the central zone of Lake Baikal eastern coast. Selecting the area to study, we hypothesize that this complex territory can be considered as a model biome that adjoins Lake Baikal central zone in the east. The major landscape of the studied area is composed of forests complicated by the bogged valleys of the rivers Cheremshanka, Talovka and Bezymyanka. The Katkovskaya and Chernaya Griva mountains range stretches from the northto the east. 167 relevés were performed by standard methods of the Braun-Blanquet approach. To reveal the phytosociological composition of the vegetation supervised k-means classification was performed in JUICE program. By comparing the vegetation data from the studied area (167 relevés) with data from the adjacent territories of Lake Baikal, Svyatoi Nos Peninsula and the Barguzin mountain range (589 relevés) was obtained the final prodromus of the vegetation. The soil identification was performed according to Russian soil classification. The vegetation mapping was performed in ArcGIS 10.3.1 by the supervised image classification of multispectral panchromatic imagery SPOT 6. The vegetation of the territory under study is represented by four classes. The dominant type of the vegetation is represented by forests which are classified into Vaccinio-Piceetea Br.-Bl. in Br.-Bl. et al. 1939 class. The wetland vegetation includes two classes Scheuchzerio-Caricetea nigrae (Nordh. 1936) Тх. 1937 and Oxycocco-Sphagnetea Br.-Bl et R. Tx. 1943. The vegetation of shifting sands of the coastal line is classified into Brometea korotkyi Hilbig et Korolyuk 2000 class. For all phytosociological vegetation units are identified seven soil types, such as, Lithozems, Brown soil, Soddy Brown Forest soil and Rzhavozems, Fen Peat, and Peat Gleyzem. The soil distribution demonstrates its contingence with an altitudinal gradient; however, transitioning from mountain to plain areas, a hydrological regime becomes crucial. Despite a relatively small territory under study (approx. 500 km2), the vegetation is relatively diverse. Location of the studied area within the zonal forest belt contributes to the leading position of the forest communities. The close ground water occurrence creates suitable conditions for wetland vegetation formation. Lake Baikal coastal line is considered as a refugium of the unique ancient Miocene-Pliocene xerophytic vegetation and flora. Thus, Lake Baikal water body, mountain landform and close ground water occurrence contribute to the formation of diverse vegetation communities. We suggest that the vegetation of this relatively small territory can be considered as a model within the central zone of Lake Baikal eastern coast.