FrozenEcosystems - Understanding Siberia's past with a combination of state-of-the-art and next-generation ice core methods

Research Project
|
01.10.2022
- 30.09.2026

Glaciers are unique natural archives that contain geophysical and geochemical indicators of past environmental change, and ice cores from high-alpine glaciers have recently been analyzed also for microfossils, particularly pollen from terrestrial plants. Microfossils provide complementary information to other paleo indicators in ice cores and in combination with them allow reconstruction of large-scale vegetation changes, and past vegetation - fire - climate linkages. Background / Rationale: The vast boreal forests in Siberia are suffering severe degradation from climate change-related droughts, increasing fire risks, and human activities. However, paleoecological records for Siberia are scarce, usually reflect local source areas and often lack chronological precision to correlate them in detail with other paleoclimate and environmental data. Siberian high-alpine glaciers are underexplored natural archives for understanding vegetation and fire dynamics that preserve many climate and environment tracers, including microfossils that provide detailed ecosystem information at exceptional chronological precision. We will develop high-resolution records of ecosystem and fire dynamics from a new ice core from Belukha glacier, southern Siberia. Conventional, manual microscope-based identification of microfossils in ice cores and other archives is time-consuming, severely limiting the achievable temporal resolution. Hence, we will apply and further develop emerging automated detection and classification approaches for microfossil identification to overcome such limitations. Objectives: The project will provide 1) a new holistic view on millennial-scale boreal forest dynamics and drivers of past vegetation change from Siberia and 2) groundbreaking methodological advancements for automated image analyses for microfossils in ice core samples using automated detection and classification approaches. Specifically, we aim to generate a multiproxy vegetation and fire reconstruction from Belukha glacier spanning the Holocene and potentially extending to the Last Glacial period to assess supraregional ecosystems dynamics and the role of fire regimes during periods with extreme climate amplitudes. We aim to introduce a next-generation automated detection and classification approach of pollen and other microfossils to ice core research. After successful application to the Siberian ice core, characterized by low pollen diversity, we will test the approach on ice samples with various degrees of floristic diversity from around the world, and finally more complex samples from sediments which contain various non-pollen structures that complicate the analysis. Methods: We will apply state-of-the-art manual optical pollen and spore analyses to infer past vegetation and land use dynamics, as well as charcoal and black carbon analyses to achieve a multiproxy fire reconstruction. We will combine these established methods with a promising emerging technology for automated microfossil detection and classification that we will apply for the first time to ice cores. Due to the exceptional preservation of pollen and low contamination with other organic debris, ice cores are particularly well suited for further developing such methods before potential application to other natural archives with more complex interferences and matrices. Analyses will be conducted on an ice core for which many other proxy records (e.g., for temperature/ precipitation/ drought/ air pollution) have been obtained already. This will allow us to directly assess the relationship of reconstructed ecosystem patterns with other environmental records and enable testing of hypotheses regarding drivers of ecosystem change and feedback mechanisms between vegetation, fire, climate, and other environmental conditions. Expected results / Impact: We will generate an exceptionally long paleoecological ice core record for mid-latitudes from Siberia, providing an integrated view on how ecosystems have responded to centennial to millennial scale environmental change. The microfossil record will deliver a regional picture of vegetation history and allow us to assess how climate change, specifically projected future reduction in moisture, will affect boreal forest ecosystems and fire regimes in the region by providing an analogue reconstruction of the past. We will also advance image analysis for pollen and other microfossils by further developing and applying this methodology to ice core records, which will provide a tool for future investigations in glacier ice samples but also in more complex pollen samples from other environmental archives such as lake or marine sediment cores.

Collaborations & Cooperations

2026 - Participation or Organization of Collaborations within own University

Heiri, Oliver, Prof. Dr., Universität Basel, Research cooperation

2026 - Participation or Organization of Collaborations on a national level

Schwikowski, Margit, Prof. Dr., Paul Scherrer Institute, Research cooperation

Funding

FrozenEcosystems – Understanding Siberia’s past with a combination of state-of-the-art and next-generation ice core methods

SNF Ambizione / PRIMA (GrantsTool), 10.2022-09.2026 (48)

PI : Camara, Sandra.

Publications

von Allmen, Robin et al. (2024) ‘Method development and application of object detection and classification to Quaternary fossil pollen sequences’, Quaternary Science Reviews, 327, p. 108521. Available at: https://doi.org/10.1016/j.quascirev.2024.108521.

URLs

Members (3)

Sandra Camara

Principal Investigator

Julia Hasler

Robin von Allmen

Research Groups

Department of Environmental Sciences

Geoecology