Syntaxonomy of peatland vegetation: case study of the central zone of Lake Baikal eastern coast

Abstract

Peatland vegetation is dominant among azonal vegetation types within the central zone of Lake Baikal eastern coast. In order to classify the phytosociological diversity of peatland vegetation, the supervised k-means clustering method was applied. According to the main concept of the method, the unidentified dataset from this study was subjected to the classification among identified units (a priori groups) from the north-eastern area of Lake Baikal. The unidentified dataset is composed of 55 relevés, which were collected in the central zone of the Lake Baikal eastern coast. The identified a priori groups comprise 12 associations distinguished according to the Braun-Blanquet approach. Those associations are based on classification of 138 relevés, which were taken in the course of studies of the north-eastern part of Lake Baikal, such as Svyatoi Nos Peninsula and Barguzinskiy mountain range. As a result, we obtained the assignment of unidentified relevés among those a priori groups. The supervised k-means clustering performed well in assignment of unidentified relevés to a priori groups and proved to be a useful tool for the classification and syntaxonomical positioning of unidentified peatland vegetation. The peatland vegetation was classified into classes: Oxycocco-Sphagnetea and Scheuchzerio-Caricetea nigrae.