Home – Home – Jornals – Siberian Journal of Forest Science 2026 number 1

In recent decades, there has been a rapid increase in the use of deep machine learning tools based on artificial neural networks in various fields of science. Deep neural networks vary in their architecture, for example, in a convolutional neural network, different layers can use convolutional kernels to extract key features from an image and pool the layers to generalize these features. Recurrent neural networks process sequential data series and retain memory of past data by returning the output of a layer back to the same layer. Training a neural network involves optimizing the weights of connections in the network to minimize the prediction error. Deep learning has the potential to leverage information hidden in large datasets to provide innovative solutions to complex environmental challenges. Big data consists of images, audio, videos, or unstructured text, which can be challenging to analyze using traditional statistical methods. With an exponential increase in publications on the methods and results of deep learning on neural networks in various fields of knowledge, this review attempts to analyze some of its applications in the field of forest ecology. In particular, it presents the results of using artificial neural networks to solve certain problems in Russian forestry, such as combining heterogeneous data to estimate forest phytomass, mapping and predicting forest cover dynamics, and identifying plant roots in minirizotron images. The final section describes some of the achievements, challenges, and uncertainties of deep machine learning in ecosystem ecology.

Using the example of the Isetsky granite quarry (Middle Ural taiga forest region), species diversity, aboveground phytomass, and projective cover of the living ground cover (LGC) were studied during the first years after reclamation under a forestry approach. The relevance of the study is determined by the need to assess the effectiveness of reclamation and the state of ecosystems on disturbed lands. It has been established that 1-2 years after the establishment of forest plantations, the number of LGC species on the surveyed sites ranges from 16 to 29. One year after reclamation, the aboveground phytomass varied from 101.8 to 130.3 kg/ha (oven-dry weight), and after two years, from 552.9 to 1056.5 kg/ha. Species diversity and aboveground phytomass tended to increase already within the first two years after reclamation. In the aboveground phytomass of LGC, synanthropic coenotype species absolutely dominate. Despite the significant number of LGC species, in all seven experimental variants with varying heaping soil-ground thicknesses, one or two species dominated in the aboveground phytomass, accounting for more than half of the total aboveground phytomass in a completely dry state. In the first year after completion of reclamation work, this is white goosefoot ( Chenopodium album L.), and in the second year, fireweed ( Chamaenerion angustifolium (L.) Scop.), common sorrel ( Rumex acetosella L.), and white clover ( Trifolium repens L.).

The genus willow ( Salix L.), characterized by a widespread occurrence of polyploidy (ranging from 2n = 38 to 12n = 228) and chromosome number instability, is of significant interest to evolutionary biology and breeding. Arcto-Montane willows possess valuable adaptations to extreme environmental conditions and ornamental potential for landscape design; however, they remain cytogenetically understudied, particularly in Russia, where taxonomic and ecological research predominates. This work addresses this gap by applying a flow cytometry method to determine the ploidy level of 13 Arcto-Montane willows from the collection of the Botanical Garden, Russian Academy of Sciences, Ural Branch in the city of Yekaterinburg. The study employed an external standardization method, based on calculating the ratio of samples nuclear fluorescence intensity, followed by determining DNA content using internal standardization (with Solanum pseudocapsicum L., garden parsley ( Petroselinum crispum (Mill.) Fuss), and Pisum sativum L. as standards). The protocol was optimized by adding 2 % PVP to neutralize secondary metabolites. The results confirmed diploid level (2x) for creeping willow ( Salix repens L.), Pyrenean willow ( Salix pyrenaica Gouan), mountain willow ( Salix arbuscula L.), Salix saxatilis Turcz., and fishy willow ( Salix foetida Schleich. ex DC.), and tetraploid level (4x) for Salix caesia Vill. Specimens of retuse-leaved willow ( Salix retusa L.), Salix glauca var. callicarpaea ‘Haltia’, and Salix foetida × retusa were found to be octoploid (8x), while while gray willow ( Salix glauca L.) was decaploid (10x). The obtained data on ploidy and DNA content are important for taxonomy, hybrid identification, and the selection of perspective breeding material. Further molecular genetic studies are required to confirm the parental forms of the hybrid taxa.

Siberian dwarf pine ( Pinus pumila (Pall.) Regel) , is a valuable source of dietary lipids containing biologically active fatty acids (FA), including pinolenic, linoleic and oleic acids, which have anti-inflammatory, hypolipidemic and cardioprotective properties, have a number of beneficial effects on body weight, as well as on fat deposition by increasing energy consumption (oxidation of fatty acids) and reducing energy consumption with food (decreased appetite). Siberian dwarf pine cones were collected in the Irkutsk Oblast, Bodaibinsky District, near the village of Artemovsky and in the Bilibinsky District of the Chukotka Autonomous Okrug at the Bayimka (Peschanka) deposit. Gas chromatography with mass spectrometric detection (GC-MS) has revealed a high lipid content in Siberian dwarf pine seeds: in the Irkutsk Oblast, the total lipid content is 66.25 %, and in the Chukotka Autonomous District - 64.67 %, which confirms their nutritional and functional value. Linoleic acid (42.29-42.43 %), oleic acid (24.58-25.51 %) and pinolenic acid (17.27-18.29 %) were the dominant fatty acids, while the proportion of saturated LC (palmitic and stearic) did not exceed 7.5 %, which further increases the nutritional attractiveness of raw materials. It has been established that the total lipid content and profile of most LC are maintained at a high and stable level. However, statistically significant differences in the content of pinolenic acid were found between seeds from different regions ( p < 0.05), which may be related to the climatic conditions of growth. The results obtained indicate the high prospects of using siberian dwarf pine lipids for the development of innovative functional food products, biologically active additives and pharmaceuticals aimed at correcting metabolic disorders and improving cardiovascular health.

Forests are complex biological systems that constantly interact with the atmosphere and are influenced by man-made pollutants. This article examines anthropogenic emissions of pollutants into the atmosphere from stationary sources, such as large factories, thermal power plants, and others. The morphological and anatomical characteristics of Scots pine ( Pinus sylvestris L.) undergrowth needles were studied in stands both growing under minimal anthropogenic impact and under the influence of man-made emissions. To study the impact of atmospheric pollution, two sites were selected in Krasnoyarsk (Akademgorodok and Tatyshev Island), one site in the Berezovsky District (Devil’s Finger Rock), and two sites in the Western Sayan Mountains (Medovy Klyuch Stream and Titenkino tract). To maximize the information content, undergrowth up to 2.5 m in height was selected. Data on atmospheric pollution at the selected sites were obtained from the State Report «On the State and Protection of the Environment in Krasnoyarsk Krai in 2023». The morphological and anatomical features of Scots pine needles were studied: needle length, width, and thickness, the width and thickness of the central cylinder, the size of the vascular bundles, and the number and size of resin ducts. Variable and stable features of the internal structure of the needles were identified. The most variable features were needle length, needle thickness, and the number of resin ducts. The most stable features were needle width, the size of the central cylinder, the width of the vascular bundles, and the average diameter of the resin duct. It was shown that under anthropogenic load, needles have a smaller length and width, as well as a smaller number of resin ducts. It has been established that much data regarding the morphology of needles exposed to various influences remains controversial and requires further research. The obtained results demonstrate the high sensitivity of needle morphological traits to environmental changes and once again highlight the potential for using these parameters as indicators of the ecological state of forest ecosystems. The research findings may be useful for monitoring forest health and developing conservation measures under anthropogenic, including recreational, influences.