Yeast evolution experiment: mutations helping in early stationary phase trade off with late-phase fitness
In serial-transfer evolution of S. cerevisiae with up to 10 days between transfers, early and late stationary phase favored different, trade-off mutations, the authors report.

Jason Tarkington, Aditya Mahadevan, Angelina G. Chan and Gavin Sherlock experimentally evolved the yeast Saccharomyces cerevisiae in batch culture and measured how adaptive mutations perform in different parts of the growth cycle. The work was published in PLOS Genetics on 16 November 2026 (doi.org) plos.org
doi.orgplos.orgOpen source ↗. Conditions spanned up to 10 days between transfers, of which up to 8 days were spent in stationary phase.
Stationary phase begins when a limiting nutrient is exhausted and cell division stops. Most cells then enter quiescence and lose viability over time. Metabolism nevertheless continues in spent medium, and the environment changes, so the relative fitness of genotypes can differ during stationary phase.
Experimental design
The authors varied the time cells spent in stationary phase between growth cycles and isolated adaptive clones. They then measured the relative fitness of these clones in a range of environments, defined by different durations of stationary phase and by two carbon sources.
By comparing the performance of each mutant with that of its ancestor during a given period of the growth cycle, the authors estimated the effect of each observed mutation on performance in different phases plos.org
doi.orgplos.orgOpen source ↗.
Trade-off between early and late stationary phase
According to the authors, when an adaptive mutation arises in growth cycles that include stationary phase, its effect on stationary phase performance is largely independent of the carbon source provided. The carbon source used before stationary phase had little impact on subsequent performance improvements.
For the same group of mutants, effects on performance in early stationary phase were negatively correlated with effects in late stationary phase, which the authors interpret as a trade-off. Longer stationary phase intervals produced larger fitness effects of adaptive mutations and distinct routes of adaptation.
The authors conclude that stationary phase comprises more than one distinct fitness-related phenotype. Phenotypes that give high performance in the first days of stationary phase trade off with those that give high performance in later stationary phase plos.org
doi.orgplos.orgOpen source ↗. They note that this supports earlier evidence that trade-offs across the growth cycle are common.
Sequencing data are deposited in the NIH Short Read Archive under BioProject PRJNA1394608.



