Possibility of Using Zoning of Fallow Vegetation according to Vegetation Indices to Assess the Patterns of Accumulation of Organic Matter in Post-Agrogenic Soils

Abstract

An array of Eutric Retisols (Loamic, Cutanic, Ochric)) was studied under a fallow aged 20–25 years, which is in the stage of overgrowth of meadow vegetation, pine and birch. The site is confined to one element of the relief, has no morphological signs of the development of erosive processes and is characterized by a homogeneous granulometric composition. To assess the influence of fallow vegetation type on the formation of soil organic matter (SOM) reserves, vegetation cover was zoned according to vegetation indices calculated on the basis of remote sensing (RS) data. The “k-means” algorithms and the “random forest” method were used for zoning. It was shown that there were statistically significant differences between the types of land cover in terms of reserves of SOM in the upper layer of the old-arable horizon with the allocation of 3 and 4 clusters. It is shown that the most expedient is the allocation of 3 classes of fallow vegetation using the “k-means” algorithm: coniferous woody vegetation, deciduous woody vegetation and herbaceous vegetation. The correctness of the allocation of these classes was confirmed by a field geobotanical survey of the territory. The results of a pairwise comparison of sites occupied by various types of fallow vegetation show the presence of significant differences in the reserves of the SOM only in the uppermost layer (0–5 cm) of the old arable horizon and only when compared with the array occupied by woody coniferous vegetation and herbaceous vegetation. Differences in accumulated humus reserves in the upper layer of 0–10 cm are statistically significant in soils under deciduous and coniferous woody vegetation, as well as between herbaceous and coniferous vegetation. There was no significant difference in this indicator between the areas occupied by woody deciduous vegetation and herbaceous vegetation.