Vol 104, No 3 (2025)

Nine rare and poorly-known species of weevils (Curculionoidea) were records in 2023 and 2024 in nature reserves in the south of the Donetsk People’s Republic, significantly supplementing the data on their distribution. Seven species were discovered in the fauna of Donbass for the first time while Bruchela musculа (K. Daniel & J. Daniel, 1903), Aizobius sedi (Germar, 1818), Tychius minor Karasyov & Caldara, 1992, and Ceutorhynchus nigritulus Schultze, 1897 were found new to the fauna of the northern Azov region. Both Ceutorhynchus alliariae H. Brisout, 1860, not known in other parts of Russia, and Pseudorchestes kostali (Dieckmann, 1985), reported from Donbass earlier from isolated records north of the Azov region, were missing from the latest major compendia for the region. New data on host plants were obtained for some species.

For the first time, a comprehensive assessment of energy resources (lipids and glycogen) in tissues and organs, and some related population characteristics was carried out using the Common frog (Rana temporaria) from the eastern periphery of its distribution range (suburbs of Perm City, Pre-Ural region).The traditional concept of the seasonal dynamics of lipid and glycogen reserves, their partitioning between the needs of growth, reproduction and survival during overwintering, all obtained using western populations, was shown to be not fully applicable to individuals from other parts of the range, even with the same duration of the activity season. The values of the relative masses of storage organs (liver and fat bodies), the content of glycogen and lipids therein, the body sizes of individuals and their reproductive characteristics revealed in the study population can serve as the basis for studying the geographic variability of these interrelated features in the eastern part of the range.

The first description of the song repertoire and song sharing in the Red-flanked bluetail is provided. The study was performed in the Khakassky State Nature Reserve, Siberia. Singing of 19 males were analysed. Songs lasted about 1 s. The repertoire size per male were 1–3 (usually 2) song types. The vast majority of songs consisted of two parts: the motif part and the trill part. In a male repertoire, different song types can have a similar motif part, but different trill parts, or vice versa. However, in each male, the song types are stereotyped, and the motif and trill parts from a male repertoire cannot be combined freely. In contrast, in the population as a whole, the diversity of songs increasing due to different males can combine a limited number of motif and trill types in different ways. Therefore, the motif types and the trill types, but not the song types, are parts of a population repertoire. To reveal the general patterns of repertoire structure and organization in Saxicolinae, the data obtained were compared with several other species from the same subfamily. Many Saxicolinae are found to be capable of increasing the song diversity at either individual level or both at individual and population levels by more or less freely combining the stereotyped parts of the song.

The objective was to determine whether the behavior of seals on ice and on shore rookeries differs, based on new data on molting and relaxation haulouts of Baikal seals (Pusa sibirica Gm.) occurring on floating ice in spring. Surveys conducted in the northern part of Lake Baikal in May–June 2022–2023 using small UAVs served as material. As ice conditions are heterogeneous habitats, in the context of the research topic we distinguished seven ice types, where the size-age composition and density of animals in rookeries, the orientation of animals on ice relative to water, and their resting and sleeping position vary. On all ice types, adult females and males (55–62%) predominate in haulouts; the preferred orientation of seals is at the edge of the ice floe (44–82%), with their head towards the water (37–71%). The preferred resting and sleeping position is the safest for quickly leaving the ice if necessary: lying on the stomach, with flippers pressed against the body (54–86%). The choice of ice floes for haulouts is likely to be determined by the physiological state of the seals, particularly the molting stage, based on the nature of the substrate (area and thickness of ice floes, and whether the surface is hard or “soft”). In late May and early June on different ice types, the relative number of seals with completed molting ranged from 22 to 76%. A comparison of the behavior patterns of seals in ice haulouts with their behavior in coastal rookeries showed that, when dispersed on ice, a larger minimum individual distance is maintained between the seals (commensurate with body size). This minimizes mutual contacts between animals, this being reflected by their behavior: most patterns of peaceful behavior described for seals lying on land are preserved, but neither aggressive nor even pacifying patterns are recorded. Various reactions of seals lying on ice to UAV flight were noted depending on the flight altitude, the number of seals in a group of haulouts, the type of ice (substrate), the physiological state of individuals (molting/molted), and individual behavioral characteristics.

A system for recognizing and counting two classic Arctic species, the Wild reindeer (Rangifer tarandus tarandus L., 1758) and the Brent goose (Branta bernicla L., 1758), in photographs from aircraft is presented. The AutoGenNet recognition system is based on a convolutional neural network (CNN) of the Mask R-CNN architecture using the Auto ML (Automated Machine Learning) concept. The system uses transfer learning, the essence of which is that at the first stage the system is trained for recognizing a variety of objects applying a standard array of images (about 328 thousand images), to be further trained on images of target objects. This approach allows for a number of images of target objects from several hundreds of thousands at one-stage training to be reduced down to several hundred at two-stage training. A synthesis of the CNN model on the basis of marked images in the AutoGenNet system is performed automatically. A special Markup program was developed for marking animals in photographs and preparing a training sample.

The first stage of system training is performed once by SNA and deep learning specialists. The second stage of training can be managed repeatedly in order to retrain the system that made errors in recognizing the objects. The work at this stage can be performed by system users who have no special education in the field of SNS training.

Two variants of work with the system are possible: a stand-alone mode in the presence of the necessary computing resources or work via the Internet with the AutoGetNet located on the servers of the SPC RAS. The CNN model presented in this contribution was trained based on 100 images of wild reindeer herds. The error of reindeer recognition using an independent data set was about 18%. 260 images of the flocks of Brent goose in different environments, be this on land, on water or in the air, were utilized to recognize Brent geese. The recognition error was about 35%. The AutoGenNet system is unified in terms of recognition objects and can be trained to recognize other animal species without any change in the program, provided they are distinguishable in the images.