Uneven preservation of ancient DNA along lake sediment cores: A case study of high-latitude and high-elevation lakes

Although sedimentary ancient DNA (sedaDNA) is increasingly used to reconstruct past ecosystem changes, we do not yet know much about its preservation conditions across geological time, resulting in potential biases and uncertainties in data interpretation. In this study, we obtained sedaDNA records from around 15 lakes from the Arctic and sub-Arctic regions and the Tibetan Plateau covering the last 2 to 80 ka BP. In addition to the four preservation proxies recently introduced by Jia et al. (2021) (https://doi.org/10.1002/edn3.259), some new potential proxies of plant DNA metabarcoding (e.g., dissimilarity between PCR replicates) and metagenomics (e.g., average DNA fragment length, duplication rate, guanine-cytosine content, and deamination rate) have also been applied to quantify the extent of ancient DNA preservation and compared with other environmental proxy records from the cores. So far, our preliminary results from Lake Ilirney (67°21’N, 168°19’E) show that DNA content generally decreases along the core over the last 18 ka BP and then maintains at a relatively stable level up to the bottom of the core (ca. 53.4 ka BP), which is consistent with the variations in lake organic productivity reflected by TC, TOC, TOC/TN, pollen and diatom abundance, and Br. In addition, sedaDNA preservation conditions revealed by our preservation proxies are variable within the core. Good sedaDNA preservation is associated with strong physical weathering and glacial abrasion in the catchment, as indicated by high K/Ti and low Zr/Rb values, resulting in increased clastic input of clay minerals and fine sediments, which favors the adsorption of DNA molecules to sediment particles. This process might also help to deepen the lake and increase its water conductivity, which is beneficial for DNA adsorption and preservation. No clear correlation is found between sedaDNA preservation and paleoclimatic changes reconstructed by fossil pollen records. It should be noted that our results may also be influenced by the ability of the DNA extraction protocols we used to recover DNA from different types of sediments. To conclude, sedaDNA preservation may be highly influenced by sediment type and catchment erosion rate, and glacial lakes appear to be promising for sedaDNA studies in the future. Further analyses of sedaDNA records from other lakes are pending and will be finalized and presented at EGU 2022.