In our recent paper that was published in The ISME Journal (https://www.nature.com/articles/s41396-020-00754-4), we measured the variability that is caused by random births and deaths in a bacterial population as it grows. In community ecology this is called “ecological drift”, in analogy to the respective term “drift” in population genetics that describes the random changes in the frequency of allelic genes in a population.
Measuring ecological drift is not a trivial task, especially for populations in bacterial communities. To measure ecological drift, one needs to make sure that what they measure is not due to any other factor that could possibly alter the dynamics of the population, like an environmental factor such as temperature or bacterial cells migrating in and out of the community. This is so hard to achieve that ecological drift is currently estimated indirectly in microbial ecology, by using the unexplained deviation from certain ecological models as a proxy for it.
Here we thought of a different way to solve this apparent difficulty in measuring drift; if one had initially identical populations that grow under identical conditions and in isolation, they could quantify drift by comparing these communities through time because the observed differences among them could only be due to drift.
But to do that one needs to measure the size of a bacterial population cell-by-cell accurately and this is far from trivial, especially when several populations coexist within a bacterial community. For example if one would use sequencing, the introduced technical variability would surely mask the variability caused by ecological drift. Luckily, we had recently developed a method that is based on flow cytometry with a very high precision to measure bacterial populations in communities composed by up to three specific bacterial strains1.
With this tool and experimental concept in hand, we went on and performed the experiments to quantify drift. And we did! We monitored initially identical triplicate communities containing one to three bacterial populations, subtracting the observed variability across the replicates from the known experimental variability to calculate drift. After having to perform each assay at least twice in order to ensure that we start with as identical communities as possible, we found that the populations across the replicate communities could be 1.4-2% different on average due to ecological drift. This variability was not related to any of the experimental parameters, further corroborating that what we measured was indeed caused by a random process such as ecological drift.
While this was already very important to know because such a quantification of drift did not previously exist, we also asked how this would affect bacterial communities in nature. To estimate that, we made ecological simulations using complex bacterial communities of up to 10,000 species under various conditions. We estimated that under harsh environmental conditions, i.e., when the environment fluctuates highly and when there is minimal migration of cells from community to community, the variability caused by drift can result in significant changes in the structure of the communities and in many bacterial species getting extinct. In specific, ecological drift can change communities by up to 15% and cause 5% of the total species to become extinct. The extinct species are those with the smallest populations to begin with, but under harsh conditions species with large populations can get extinct as well.
Overall, in this work we got the first insights regarding the expected variability in the size of a bacterial population due to ecological drift and, using that, we predicted that ecological drift can account for a significant part of the unexplained variability in natural bacterial communities. Based on our simulations, we also expect that drift will be an even more important driver of assembly in bacterial communities in the prospect of abrupt environmental fluctuations under global climate change.
1 Fodelianakis, S. et al. Dispersal homogenizes communities via immigration even at low rates in a simplified synthetic bacterial metacommunity. Nature Communications 10, 1314, doi:10.1038/s41467-019-09306-7 (2019).