Mortality causes universal changes in microbial community composition

Abstract

6All organisms are sensitive to the abiotic environment, and a deteriorating 7 environment can lead to extinction. However, survival in a multispecies 8 community also depends upon inter-species interactions, and some species may 9 even be favored by a harsh environment that impairs competitors. A 10 deteriorating environment can thus cause surprising transitions in community 11 composition. Here, we combine theory and laboratory microcosms to develop a 12 predictive understanding of how simple multispecies communities change under 13 added mortality, a parameter that represents environmental harshness. In order 14to explain changes in a multispecies microbial system across a mortality gradient, 15we examine its members' pairwise interactions. We find that increasing mortality 16 favors the faster grower, confirming a prediction of simple models. Furthermore, 17if the slower grower outcompetes the faster grower in environments with low or 18 no added mortality, the competitive outcome can reverse as mortality increases. 19We find that this tradeoff between growth rate and competitive ability is indeed 20 prevalent in our system, allowing for striking pairwise outcome changes that 21 propagate up to multispecies communities. These results argue that a bottom-up 22 approach can provide insight into how communities will change under stress. 23

Community composition of microbial microcosms follows simple assembly rules at evolutionary timescales

“…In order to first evaluate whether incorporating data from evolutionary timescales is required, we used the fractions of pairs at generation~70 to predict the trios' composition throughout the experiment. These predicted the trio compositions at the first~70 generations with a prediction accuracy of 0.84 (similar to 0.86 that was found by Abreu et al 13 , Figs 4B, S7), significantly better than the uninformed guess (0.73 accuracy, Methods, Mann-Whitney p-value = ). 3 * 10 −5 However, as anticipated by the fact that the composition changes significantly during these timescales ( Fig.…”

“…It has been previously established that two-species competition experiments could often predict the composition of multispecies communities 12,13 . While these experiments suggest that most ecological outcomes could be explained by pairwise interactions, it is still unknown whether similar predictability holds at evolutionary timescales.…”

“…While pairwise competitions predict the composition of multispecies communities, measuring the pairwise compositions of multiple pairs over evolutionary timescales is laborious and in many cases infeasible. Monoculture data, such as species' growth rates and carrying capacities, can be acquired more readily and was previously shown to correlate with the species' competitive ability [12][13][14]35 . Therefore we next tested whether such monoculture data may provide a more accessible alternative for predicting the structure of multispecies communities.…”

“…Currently, predicting the structure and dynamics of microbial communities commonly relies on fitting specific ecological models, or using genome-scale metabolic models [8][9][10] . Challenges in fitting such models and concerns regarding their validity 11 have recently led us to develop an alternative non-parametric predictive framework, which was validated using laboratory experiments in microbial microcosms 12,13 , and in the gut of C. elegans 14 . In this bottom-up framework, the composition of multispecies communities is inferred using the measured pairwise interactions between the members of the community.…”

Fine sediment and flow velocity impact bacterial community and functional profile more than nutrient enrichment

“…Bacterial communities have fast turnover rates with shifts in abundance, community composition and activity occurring within days (Abreu et al 2019, Romero et al 2019 a ), even in response to minor changes in abiotic conditions. Moreover, stream biofilms are the first to come in contact with dissolved substances in the free‐flowing water column and have been successfully used as "early warning systems" for detecting effects of toxicants (Sabater et al 2007).…”