On the results of the fifth IPEE RAS research expedition to study the Baikal seal in 2024

Within the framework of the Baikal seal research programme for 2020-2025 and with the financial support of the Lake Baikal Foundation, a regular research expedition of scientists from the A.N. Severtsov Institute for Problems of Ecology and Evolution (A.N. Severtsov IPEE RAS) took place from 24 July to 22 August 2024. A.N. Severtsov Institute for Problems of Ecology and Evolution (IPEE RAS) to study Baikal seals at the largest summer haulouts located in the middle part of Lake Baikal, in the archipelago of Ushkanyi Islands.

The tasks of the scientific expedition included:

  • conducting visual counts of the number and appearance of seals on Tonky Island;
  • conducting population counts using unmanned aerial vehicle (drone, UAV) at haulouts on Tonky, Dolgiy and Kruglyy Islands;
  • collecting material to assess the impact of tourists on seal behaviour;
  • collection of biological material from fallen individuals found on the coast of Tonky Island for further research.

Project results in figures and facts:

  • During the expedition, from 24 July to 22 August 2024, 39 visual counts were conducted on Tonky Island – the total number of animals, the number of injured, sick and unmolted seals were recorded. During the surveys, more than 5000 photographs of stranded seals were taken.
  • Made 16 UAV overflights of each island, and obtained 48 videos totalling 4 hours 45 minutes 23 seconds.
  • Based on the data obtained during the surveys, the distribution of seals on the islands of the Malye Ushkanyi Islands archipelago under different weather conditions was noted, the dynamics of seal abundance was determined for the observation period (July-August 2024), and inter-annual differences in the number of seals on the islands were revealed in comparison with the results of previous surveys.
  • Nineteen days of behavioural observations were conducted, during which 626 time slices were obtained (a time slice is a series of 3 photographs of a seal den, taken every 10 minutes between 6:00 and 12:00). For each time slice, weather conditions and the number of tourists at the viewing platform and their behaviour were recorded to further assess the influence of these factors on seal behaviour.
  • To analyse the behaviour of Baikal seals, 575 time slices were processed. On each slice the number of wet, semi-dry and dry seals was determined, as well as the type of behaviour exhibited by each seal in the den. Behaviour types were grouped into 2 categories – calm and anxious.
  • Count data from visual and UAV surveys, as well as behavioural data on Baikal seals, were processed using modelling. Optimal models were selected and analysed to determine the environmental factors affecting the number of animals in haulouts/seal behaviour. The environmental factors used in the analysis of count data were cloud cover, wind strength and direction, lake level, relative humidity, pressure and air temperature. Anthropogenic factors such as the number of tourists at the observation platform and their behaviour were also taken into account in the analysis of seal behaviour.
  • Samples were collected from 4 dead animals. Repeat hair samples were collected from one individual to investigate changes in hormone levels after animal death. Biological material collected from dead animals was partially processed. Cortisol concentrations in hair and vibrissae were measured, which will allow us to assess the presence of chronic stress and the dynamics of cortisol levels during the spring moult period.

The main scientific findings of the project are:

Visual records

Foot visual surveys revealed that the number of seals on the northern shore of Tonky Island in 2024 ranged from 2 to 384 individuals. The number of animals in 2024, as in previous survey years, decreased by mid-August, possibly due to redistribution of animals between islands or their departure to other areas of the lake. The level of lake swell and wind strength were the main factors affecting the number of animals at the deposit. The number of seals on land did not differ from the number of seals in the water.

The proportion of sick seals, as well as seals that have not moulted, is relatively small and can be assumed to be negligible. Differences were found from the results of similar field surveys conducted in 2020 and 2021. – The number of animals counted in 2024 is significantly lower than the number of animals counted in the previous years – 2020 and 2022, but does not differ from the number of animals counted in 2021. This may be due to the increased anthropogenic load observed during the field work in 2024.

Surveys using UAVs

A total of 13,384 seals were observed during the UAV surveys, including 5,390 in the water and 7,994 on land. The highest number of seals was recorded on Dolgiy Island and the lowest on Tonky Island. The maximum number of animals recorded at one time was 319 on Tonky Island, 1060 on Dolgiy Island and 813 on Krugly Island.

Comparison of the 2024 UAV survey data with the 2020 and 2021 surveys shows that the lowest number of seals was recorded in 2021 and the highest in 2020. For all survey years, the highest numbers of seals were observed on land for all islands. There were more seals per survey day in 2020 than in the other years, with the average number of seals per survey day for 2020 almost double that for 2024, and the average number of seals per survey day for 2021 noticeably lower than for 2020 and 2024, with 2021 covering a longer period of time than the other years.

For all years of observation, the highest numbers of seals were recorded on Dolgiy Island and the lowest on Tonky Island, which contradicts the existing ideas that Tonky Island is the most important for seals. Thin Island is the most important for seals. Comparison of UAV data with meteorological data shows that wind speed has the most significant effect on seal numbers – the higher the wind speed, the lower the number of seals on the islands (and on land in particular). Wind direction in many cases also affects the number of seals and their distribution within the archipelago – seals are more likely to be recorded on leeward (sheltered from the wind) parts of the islands.

Comparison of the UAV data with data from the foot surveys on Tonky Island shows that the results of the different counting methods are broadly consistent. However, it is noted that shore-based surveys are likely to undercount animals in the water and may also undercount animals on land during UAV counts. A combination of the two counting methods may provide the most accurate results on seal numbers. Correction factors for the number of animals on land counted by UAV and the number of animals in the water counted during shore-based surveys were calculated using available data. As a result, the correction factor k = 1.33 for animals on land with UAVs and k = 1.37 for animals in water during foot surveys. These coefficients were eventually applied to both the observations on Tonky Island and UAV surveys on other islands.

Analysis of animal behaviour

An adapted method of time slices was used to assess the influence of environmental factors on the behaviour of Baikal seals. A total of 653 time slices were processed, of which 575 were included in the analysis of Baikal seal behaviour.

Several environmental factors influence the proportion of disturbed animals on the rock: the total number of seals on the rock, the condition of the rock, cloud cover, the presence of a northerly wind, and the level of lake turbulence. The proportion of seals on the rock increases when the total number of seals on the rock decreases and when wet seals predominate, which most likely correlates with the period of recovery of the rock after the drift – this period can be considered more sensitive for seals. Weather conditions also influence seal behaviour – cloudiness, lake swell of 1-3, and the presence of northerly winds were associated with a higher proportion of disturbed seals in the den. Factors such as the presence and number of tourists and tourist behaviour have no significant effect on the proportion of disturbed seals, but violations of the rules for visiting the observation platform may lead to a mass descent of seals into the water.

We analysed the cases of mass haul-out and identified 4 scenarios of recovery: “rapid return” (5 cases), “prolonged return” (1 case), “return with a decrease in the total number of seals” (9 cases), “no return” (2 cases). The most optimal scenario is the “rapid return” scenario, in which the number of seals on the rock is restored on average after 1 hour 22 minutes. We assume that an important condition for a rapid return is the presence of a small number of seals on the rock after the descent. We observed both natural and anthropogenic reasons for the shifts. The anthropogenic causes of the wanderings included active actions of tourists at the seal den (sticking out from behind the wall), as well as passing of swimming vehicles in front of the den. Complete abandonment of the deposition was due to both natural and anthropogenic causes, and resulted in a longer recovery time. In the most negative scenario, the passage of watercraft in front of the seal haul-out resulted in all seals leaving the rock and in most cases leaving the haul-out for a long time and not returning for a long period of time. These cases are unacceptable and require immediate action to prevent them in the future.

Hormonal studies

Hormonal studies using keratinised body structures are the only way to assess the hormonal status of dead animals. Unfortunately, the small sample of dead animals found in 2024 does not allow for detailed analyses and comparisons of sample individuals. Comparison of hormone concentrations in the hair of animals from samples of different years revealed inter-annual variability – for example, among all samples, the highest values of cortisol concentration were observed in the dead animals of 2020, which was characterised by a large number of dead animals found on the shore of Tonky Island, as well as early ice melt in winter, which could lead to increased cortisol concentration in animals during the moulting period. The influence of the degree of freshness of the dead individual on cortisol concentration in hair was analysed. Despite an extremely small sample of 1 individual, an increase in hair cortisol levels was found to be associated predominantly with fat contamination of the samples. Distribution of cortisol concentrations along the vibrissae of one individual was revealed – a decrease in cortisol concentration closer to the base of the vibrissae was detected – probably, the individual did not experience a prolonged influence of the stress factor before death, which also confirms the good fatness of the dead individual.

The project was realised with the support of the World Around You Foundation of Siberian Wellness.