Rethinking evolutionary inference in metagenomic time series.

As ecologists increasingly use metagenomic time series to track evolution in the wild, there is a risk of misinterpreting ecological dynamics as rapid adaptation. This Perspective identifies methodological limitations that generate misleading signatures of microbial evolution. A primary issue is confusing evolutionary change (driven by de novo mutation or horizontal gene transfer) with ecological lineage turnover, such as seasonal oscillations or the reactivation of dormant lineages. Current metagenome-assembled genomes can collapse micro-diverse lineages and decouple adaptive mobile elements, creating inaccurate genomic signatures of sweeps or stasis. To address these issues, I propose a framework integrating long-read sequencing, pangenome graph theory, and forward-time simulations to model populations as temporal genetic networks and better resolve microbial evolutionary dynamics.