DOI: 10.15393/j2.art.2024.7983
Budnik Pavel Vladimirovich | Petrozavodsk State University, budnikpavel@yandex.ru |
Baklagin Vyacheslav Nikolaevich | Northern Water Problems Institute Karelian Research Centre Russian Academy of Sciences, slava.baklagin@mail.ru |
Galaktionov Oleg Nikolayevich | Petrozavodsk State University, ong66@mail.ru |
Key words: zoning factor analysis cluster analysis k-means method discriminant analysis logging forest management |
Summary: Increasing the efficiency of forestry and forest management requires solving the issues of forest areas zoning. The purpose of this study was to substantiate statistically the typification scheme for forest exploitation conditions of central forest districts in the Republic of Karelia, Murmansk and Arkhangelsk regions. Forest exploitation zoning was carried out taking into account 19 variables characterizing wood resources, natural production conditions and road infrastructure. The data sources included forest plans and forestry regulations of central forest districts. The research methodology included the consecutive use of factor, cluster and discriminant analyses. Factor analysis was used to eliminate multicollinearity and reduce the number of variables. Factors were extracted using the maximum likelihood method, and their number was determined using the Kaiser criterion. The factor structure was rotated using the Varimax method. As a result of factor analysis, 4 factors were identified, explaining 83.16 % of the total variance of 19 variables. The first factor determined the overall productivity of forests and the calculated felling rate for coniferous stands. The second factor characterized the level of deciduous production and natural production conditions characteristic of forests with a high content of deciduous species. The third factor characterized the volume of the calculated felling rate for clear felling. The fourth factor determined the degree to which natural production conditions corresponded to favorable conditions (drier soils and productive forests). Cluster analysis included two stages. At the first stage, hierarchical cluster analysis was applied to determine the number of clusters. At the second stage, the k-means method was used to divide the central forest districts into a given number of groups. Based on the results of cluster analysis, the region under study was divided into 9 forest exploitation areas and a schematic map was developed. The reliability of the cluster analysis results was confirmed by statistical criteria of discriminant analysis: canonical correlation values, Pearson χ^2 test, Wilks' lambda test. |