The tempo and mode of evolutionary dynamics in wild bacterial communities

All species live in diverse assemblages with many hundreds of other species. A key challenge for environmental science is to understand complex ecosystems in enough detail to predict how they will respond to fluctuating environments. This is difficult because whole-ecosystem responses depend on traits of all constituent species and interactions among them. Critically, species traits are not just constant but evolve over time. Moreover, the way that traits evolve interacts with ecological changes in abundance and distribution. Accounting for evolution in diverse communities is therefore vital for managing natural systems over policy-relevant timescales.

This project will tackle this problem in bacteria. Bacteria live in exceptionally diverse communities and underpin many processes that human populations depend upon, such as nutrient cycling, decomposition, waste treatment and human nutrition. We know that they evolve rapidly over days, weeks and months in the laboratory and when exposed to strong selection pressures, such as antibiotics. The typical speed and drivers of evolution in wild bacterial communities remains largely unknown, however, due to challenges in tracking evolution in such diverse systems.

We will take advantage of recent advances in genome sequencing, robot handling of laboratory assays, and manipulative experiments to document evolution in wild bacterial communities. Our system comprises the bacteria living in ephemeral pools formed by the roots of beech trees, called tree-holes. We have previously developed experimental evolution for these communities in the laboratory and surveyed diversity in the field - here, we will integrate these approaches and quantify evolutionary dynamics in wild tree-holes. We will compare control tree-holes in normal conditions with tree-holes perturbed by an increase in pH above the natural range implemented by liming.

We will quantify the tempo of ecological and genetic changes over a hierarchy of time-scales (weeks, months, to a year) from metagenome and targeted genome sequencing data and estimate key parameters that determine evolutionary rates such as generation times, effective population size, and the frequency of positive selection - currently unknown in wild bacteria. The work will develop new methodology applicable to other systems in future, including both bacteria and eukaryotes.

To characterise the mode of evolution, we will use time-shift assays to test whether the bacterial communities are continually coevolving (as proposed by the Red Queen paradigm of evolution) or primarily adapting to changes in abiotic conditions (as proposed by the Court Jester paradigm) or coexist at an evolutionary equilibrium (as proposed by a Static paradigm). Further tests for local co-adaptation will determine whether communities coevolve locally within patches. This approach will involve running over 50,000 growth assays on robotic systems, an approach we have trialled in earlier studies to demonstrate feasibility here.

To determine the importance of dispersal for community dynamics further, in tandem we will run experiments with whole communities in bottles in the laboratory, comparing responses between bottles that regularly receive an influx of bacteria from a regional pool and closed bottles that do not after initial inoculation. In combination, the results will reveal the relative importance of local dynamics versus dispersal, and biotic versus abiotic conditions, in setting the mode of evolution in wild bacteria. Our findings on tempo and mode of bacterial evolutionary dynamics will provide key data for future prediction and management of microbial systems across a wide range of applications.