Tracing functional genes in sedimentary ancient DNA to reveal ecosystem change across the Late Quaternary

The Arctic region is experiencing rapid warming. This accelerated temperature increases results from Arctic amplification, characterized by temperatures rising two to four times faster than the global average. Such dramatic warming profoundly impacts ecological processes, biodiversity patterns, and biogeochemical cycles, particularly the carbon cycle in Arctic ecosystems. Understanding ecosystem-level carbon cycling under changing climate conditions requires understanding community changes as well as their functional capacities. sedimentary ancient DNA (sedaDNA), preserved in lake and marine sediments, represents a paleoecological proxy that enables reconstruction of past communities, their functional genes, and associated metabolic pathways. In this study, I aim to reveal the functional genes involved in carbon fixation and degradation and their long-term changes with the associated communities that play a role in the carbon cycles in Arctic-subarctic marine and terrestrial ecosystems through glacial-interglacial cycles. To approach this aim, I used sedaDNA combined with a metagenomic sequencing approach and custom developed bioinformatic pipelines. Because sedaDNA is highly fragmented and degraded, specialized analytical approaches are required to retrieve these functional and taxonomic signatures. To address this methodological challenge, this thesis optimized and customized functional gene annotation pipelines to handle fragmented ancient DNA sequences. These gene annotation pipelines were validated using quality metrics, comparison with short read-based taxonomic classification (HOLI and Kraken2) and authenticated ancient DNA signals by identifying characteristic damage patterns such as cytosine deamination. This methodological improvement increased the reliability of interpreting ancient communities’ functional capabilities. In my results, I demonstrated the applicability of functional gene annotation pipelines to sedaDNA data. In particular, the contig-based pipelines were successful in authentication of ancient DNA patterns using PyDamage. I also showed that eukaryotic functional genes can be recovered from sedaDNA data although prokaryotic genes highly dominate the data. The comparative analysis between short read taxonomy from HOLI and Kraken and functional proteins reveals similar temporal patterns across many taxonomic groups, highlighting the reliability of the gene annotation pipeline. Different databases such as the Kyoto Encyclopedia of Genes and Genomes (KEGG) and Carbohydrate-Active Enzymes (CAZy) applied to sedaDNA have enabled the functional characterization of past communities to be resolved and provided comprehensive information. The functional results in the thesis reveal clear differences between marine and terrestrial communities. Overall, marine communities exhibited a significantly higher abundance of carbon fixation genes, whereas terrestrial communities were characterized by a greater abundance of genes involved in organic carbon degradation. Taxonomic profiling of carbon fixation and degradation groups showed that Pseudomonadota played a central role in carbon cycling, whereas terrestrial carbon transformation, both organic and inorganic, involved a broader range of microbial phyla such as Acidobacteriota, Planctomycetota. Furthermore, in temporal changes, there is a shift in functional gene abundance in marine communities from degrading starch-based organic matter to degrading peptidoglycan/chitin and cellulose/hemicellulose substrates. In several terrestrial communities, the communities indicate alternating trends between the breakdown of plant-derived compounds and microbial necromass constituents such as peptidoglycan/chitin based on functional gene abundances. These findings indicate that long-term differences in substrate availability have driven distinct functional responses in marine and terrestrial carbon cycling. In conclusion, this thesis advances our understanding of the functional dynamics of Arctic communities, emphasizing the role of community-level processes in mediating biogeochemical processes in the long-term carbon cycle. This work underscores the significance of sedaDNA-based paleoecological approaches, enhanced by optimized functional gene annotation pipelines, in elucidating past community functions and predicting future responses to ongoing climate changes. These findings provide valuable perspectives for future community adaptations as indicators of ecosystem resilience and vulnerability under changing climatic regimes.