Ancient Lake DNA Reveals Microbes Evolving Faster Than Their Functions Under Climate Change

Buried in the mud at the bottom of a lake in northern China lies a 500-year archive of microbial life, and it is now telling a story that climate scientists have never been able to read before. By extracting and sequencing sedimentary ancient DNA from Lake Daihai, a closed deep-water lake in Inner Mongolia, researchers have reconstructed how microbial communities changed as the planet swung from the cold of the Little Ice Age into the human-driven warmth of the Current Warm Period. Their findings, published in the journal iScience, reveal a striking asymmetry: the identity of the microbes in the lake evolved dramatically over the centuries, but the functions those microbes performed changed far more slowly, buffered by a phenomenon known as functional redundancy.

The research team, led by Xinjie Shi and Hai Xu, analyzed a 187-centimeter sediment core collected from the central deep basin of Lake Daihai in August 2018. Using cesium-137, lead-210, and radiocarbon dating, they assigned seven sediment layers to two distinct climatic epochs: the Little Ice Age, spanning roughly 1517 to 1850, and the Current Warm Period, from 1850 to 2018. Shotgun metagenomic sequencing of the DNA preserved in these layers allowed the team to identify which organisms were present, from bacteria and archaea to viruses and microbial eukaryotes, and to annotate the functional genes they carried using the Kyoto Encyclopedia of Genes and Genomes database.

The taxonomic picture that emerged was one of continuous turnover. Bacteria dominated the sedimentary microbiome throughout the record, accounting for roughly 83 percent of relative abundance, while eukaryotes made up a mere 0.05 percent. But the proportions shifted over time, particularly after 1850, when the percentage of bacteria declined and archaea, viruses, and eukaryotes became more prominent. This timing is significant: 1850 marks the end of the Little Ice Age, when the East Asian Summer Monsoon intensified and northern China transitioned from cold-dry to warm-wet conditions, coinciding with the onset of accelerating human disturbance in the lake’s watershed through land-use change and nutrient loading.

To quantify how the community was being assembled, the researchers applied a statistical framework that separates deterministic processes, such as environmental selection, from stochastic ones, such as random dispersal and drift. Deterministic processes dominated throughout, contributing about 77 percent of community assembly, but their share declined slightly after 1850 as stochastic processes gained ground. The team attributes this shift to weakened competitive interactions among microbes during the adaptation phase, which opened the door to random birth and colonization events, as well as to dispersal limitation of nutrient-sensitive taxa and increased drift in eutrophic microhabitats created by anthropogenic nutrient inputs.

Perhaps the most consequential finding concerns the relationship between who was there and what they were doing. Microbial community similarity showed a clear temporal decay, meaning that communities became progressively more different from one another as the centuries passed. Functional pathways, by contrast, showed no such decay. When the researchers plotted structural similarity against functional similarity, the slope of the relationship was far below one, indicating that community structure evolved nearly ten times faster than function. This disparity is the signature of functional redundancy: multiple phylogenetically distinct lineages carrying identical or similar genes, so that when one species declines or disappears, another can step into its metabolic role.

Functional redundancy is more than a statistical curiosity; it is a mechanism that stabilizes ecosystems under stress. Because microbial functional traits are typically non-monophyletic, with unrelated organisms performing the same jobs, the loss or succession of particular taxa does not necessarily translate into lost ecosystem function. The Lake Daihai data show this buffering in action across a centennial timescale. The researchers also found that the intensity of functional redundancy was stronger during the Little Ice Age, when stochastic processes were weaker, than across the full 500-year record, suggesting that human disturbance has begun to erode this safety net even as it accelerates functional innovation through processes such as horizontal gene transfer.

At the heart of the community’s response were its keystone taxa, the highly connected hubs of the microbial co-occurrence network that hold the whole system together. The team identified twenty-seven keystone taxa and twenty-nine taxa whose abundances differed at least twofold between the two climatic periods, with seven taxa overlapping between the two groups. Notably, all of the distinct keystone taxa were bacteria, and their abundances rose significantly toward the present. These keystone organisms responded to variable temperature and rainfall, and they showed opposite relationships with community structure and function, suggesting that they mediated the divergent evolutionary trajectories of the two. As keystone taxa proliferated, cooperative interactions strengthened, enabling more targeted resource use, while annotated functional pathway proportions declined, hinting at the emergence of unassigned, potentially novel functions.

This trade-off strategy came with costs. As the Current Warm Period progressed, the microbial community’s genomic GC content dropped from an average of 55.79 percent during the Little Ice Age to 51.53 percent, and its niche width narrowed, indicating reduced resource utilization and a modest decline in resistance to environmental disturbance. In other words, long-term microbial persistence under climate change may come at the price of reduced ecological resilience. The community traded breadth of capability for specialization, evolving novel functions to counter potential environmental changes while attenuating traditional genetic and metabolic processing at the cellular level.

The sediment’s chemistry tells a parallel story, but with a telling delay. Stable carbon and nitrogen isotopes in the sedimentary organic matter track biogeochemical cycling at the ecosystem level, and the researchers found that dissimilarity in microbial community absolute abundance tracked dissimilarity in organic carbon isotope values, while functional pathway dissimilarity tracked organic nitrogen isotope values. Yet while the microbial community and its functions shifted visibly after 1850, the isotopic signals did not change markedly until after 1950, nearly a century later. The team attributes this lag to the hierarchy of biological organization: gene-level redundancy allows microorganisms to maintain core functions such as metabolism, carbon mineralization, and nitrogen fixation even as their composition changes, stabilizing ecosystem-level processes. After 1950, anthropogenic nitrogen compounds with relatively low nitrogen-15 values flooded into the lake and were assimilated by microbes into proteins, dragging sedimentary nitrogen isotope values downward, while wider ranges of carbon isotope values reflected enhanced phytoplankton-derived carbon inputs as eutrophication accelerated.

The implications reach beyond one lake in Inner Mongolia. The study demonstrates that microorganisms are inherently sensitive to natural climate variability, responding within decades, whereas ecosystem-level biogeochemical signals require prolonged accumulation of both climatic and anthropogenic forcings before they register in the sediment record. That decoupling means the full consequences of today’s human pressures on lake ecosystems may not become apparent in conventional geochemical monitoring for generations. The authors also acknowledge limitations: the discrete nature of sediment sampling constrains temporal resolution, and sequencing depth may have under-sampled rare taxa that could serve as cryptic seed banks for long-term resilience. Even so, the work establishes sedimentary ancient DNA as a powerful window into microbial eco-evolution, showing that the microscopic engineers of lake ecosystems have been quietly rewriting their own genetic scripts for five centuries, and that their ability to keep doing so may determine how freshwater ecosystems weather the accelerating changes of the Anthropocene.

Subject of Research: Centennial-scale evolution of lake microbial community structure and function reconstructed from sedimentary ancient DNA under climatic and anthropogenic forcings

Article Title: Evolution of lake microbial structure and function under climatic and anthropogenic forcings: Insights from sedimentary ancient DNA

Article References: Shi, X., Xu, H., Li, W., Wang, J., Song, Y., Yang, M., Wang, B., & Liu, C.-Q. (2026). Evolution of lake microbial structure and function under climatic and anthropogenic forcings: Insights from sedimentary ancient DNA. iScience, 29(10), Article 117760. https://doi.org/10.1016/j.isci.2026.117760

Image Credits: AI Generated

DOI: 10.1016/j.isci.2026.117760

Keywords: sedimentary ancient DNA, Lake Daihai, microbial ecology, climate change, functional redundancy, keystone taxa, Little Ice Age, metagenomics, biogeochemical cycling, stable isotopes, eutrophication, lake ecosystems