Decadal lake genome-resolved metagenomics reveals trait-structured microbial responses to climate change

Freshwater lakes host complex, vertically structured microbial communities whose functional traits and interactions underpin key biogeochemical cycles. However, we still lack a temporal, genome-resolved view of how these traits, particularly those involved in host-dependent relationships, respond to long-term environmental change. Here we present the Long-Term Monitoring–Lake Zurich (LTM-LZ), a 10-year, time- and depth-resolved dataset from a temperate, seasonally stratified lake experiencing ongoing disruption of its regular seasonal turnover. LTM-LZ couples biweekly depth-resolved metagenomic sampling with high-resolution climatic and physicochemical measurements, offering an exceptionally detailed perspective on freshwater microbial community and trait dynamics in time and space.

Based on these data, we reconstructed a catalogue comprising over 72,000 freshwater bacterial and archaeal metagenome-assembled genomes and compiled trait matrices describing carbon, nitrogen, and sulphur cycling, redox and resource-use strategies, host association potential, and other genomic properties. Within the vast diversity of lineages, we focused on the Candidate Phylum Radiation (CPR, also known as Patescibacteria), an uncultured yet widespread group in freshwaters that can contribute 15% or more of bacterial diversity and is characterized by extremely reduced genomes, host-associated lifestyles, and limited but specialized metabolic capacities. We then framed CPR lineages and their putative hosts within a joint trait–environment space, using co-occurrence patterns, phylogenetic relatedness, and metabolic complementarity to infer candidate symbioses and ecological niche partitioning along depth, redox, and seasonal gradients in the lake. This trait-resolved framework revealed how CPR–host assemblages reorganize in response to climate-driven shallowing of lake mixing depth and identified repeated interaction modules that link epilimnetic primary production to hypolimnetic nutrient transformations in Lake Zurich.

By integrating decade-scale climate change records, high-resolution environmental measurements, and genome-encoded traits, LTM-LZ turns a single lake into a model system for tracking genomic trends and ecological shifts in microbial populations and communities. Taken together, this approach reorients the field from describing community composition to predicting genome-resolved responses to ongoing environmental change, enabling trait-based analyses of bacterial interactions and the emergence of host-associated networks.