Colonization-persistence trade-offs in natural bacterial communities

Abstract

Fitness equalizing mechanisms, such as trade-offs, are recognized as one of the main factors promoting species coexistence in community ecology. However, they have rarely been explored in microbial communities. Although microbial communities are highly diverse, the coexistence of their multiple taxa is largely attributed to niche differences and high dispersal rates, following the principle 'everything is everywhere, but the environment selects'. We use a dynamical stochastic model based on the theory of island biogeography to study highly diverse bacterial communities over time across three different systems (soils, alpine lakes and shallow saline lakes). Assuming fitness equalization mechanisms, here we newly analytically derive colonization-persistence trade-offs, and report a signal of such trade-offs in natural bacterial communities. Moreover, we show that different subsets of species in the community drive this trade-off. Rare taxa, which are occasional and more likely to follow independent colonization/extinction dynamics, drive this trade-off in the aquatic communities, while the core sub-community did it in the soils. We conclude that equalizing mechanisms may be more important than previously recognized in bacterial communities. Our work also emphasizes the fundamental value of dynamical models for understanding temporal patterns and processes in highly diverse communities.

Keywords: colonization–extinction dynamics; fitness equalization; natural bacterial communities; neutral theory; species coexistence; species sorting.

Figure 1.

Conceptual summary of this article. The left panel shows a schematic view of colonization-extinction dynamics in microbial communities. In a local community, species can colonize (circle), get extinct (triangle), or maintain their presence (squares). However, detection levels can produce taxa to get apparently extinct (pentagon) or apparently colonize (diamond), which can be a confounding factor. Still, colonization–extinction events allow for the estimation of colonization and persistence in a consistent way. In the right panel, fitness equalization produces a relationship between colonization and persistence with a slope of −1 in logarithmic space. The solid line indicates a perfect persistence–colonization trade-off, where equalizing mechanisms such as trade-offs lead to similar fitness among groups. Any attempt of the satellite taxa (dark circles) to increase their performance would likely result in a corresponding decrease due to life-history constraints. However, in core taxa (light squares) stabilizing mechanisms dominate and niche differences are high (e.g. due to resource partitioning).

Figure 2.

The three communities studied display a signal of fitness equalization. (a) Lakes in the Pyrenees, (b) soils and (c) shallow saline lakes in Monegros. The slope of the relationship of colonization and persistence was close to −1 along the different taxonomic levels, indicating fitness equalization. Significance refers to Spearman’s ρ, * <0.05, ** <0.01, *** <0.001.

Figure 3.

The core members of the community follow a lognormal distribution. (a) Lakes in the Pyrenees, (b) soils and (c) shallow saline lakes in Monegros. Left panels: blue dashed lines represent the linear relationship between the highest abundance and occupancy at the genus level, which presents structural changes, determined by a Chow test with maximum values for the statistic in the grey shaded area. We have considered as core genera (squares) those that presented values of occupancy higher than the mean occupancy of the point with maximum structural change, while those with a lesser occupancy were considered satellite members (circles). Right: the core members of the communities present a lognormal distribution (solid blue line). Pyrenees deviance = 1.063; soils deviance = 0.666; Monegros deviance = 4.042. Lognormal distributions were fitted using function rad.lognormal of the R package ‘vegan’.

Figure 4.

Bacterial communities show a colonization–persistence trade-off at the family level. Three different habitats—(a) alpine lakes, (b) soils and (c) shallow saline lakes—display a linear relationship close to the theoretical expectation under a perfect colonization-persistence trade-off (not shown). The trade-off is maintained throughout the phylogeny, from phylum to genus. However, core (squares) and satellite (circles) members of the community show different relationships between persistence and colonization, being the satellite members closer to the theoretical expectation. The two legs indicate the −1 slope.