Microbial interactions lead to rapid micro-scale successions on model marine particles

Abstract

In the ocean, organic particles harbour diverse bacterial communities, which collectively digest and recycle essential nutrients. Traits like motility and exo-enzyme production allow individual taxa to colonize and exploit particle resources, but it remains unclear how community dynamics emerge from these individual traits. Here we track the taxon and trait dynamics of bacteria attached to model marine particles and demonstrate that particle-attached communities undergo rapid, reproducible successions driven by ecological interactions. Motile, particle-degrading taxa are selected for during early successional stages. However, this selective pressure is later relaxed when secondary consumers invade, which are unable to use the particle resource but, instead, rely on carbon from primary degraders. This creates a trophic chain that shifts community metabolism away from the particle substrate. These results suggest that primary successions may shape particle-attached bacterial communities in the ocean and that rapid community-wide metabolic shifts could limit rates of marine particle degradation.

Figure 1. Marine bacteria form communities on model particles.

(a) Schematic of particle colonization procedure. (b) Particle-attached communities stained with SYBR Green I, a double-stranded DNA strain, and imaged with bright field (top) and fluorescence (bottom) microscopy (Supplementary Methods). Scale bars, 25 μm. Note that different particles are depicted for each time point. (c) Total 16S rRNA V4 copies per particle over time for three colonization replicates (, , ). Symbols in grey (, , ) indicate measurements below the limit of detection of the assay. The grey line (-) indicates the fit to a logistic growth model.

Figure 2. Bacterial communities undergo rapid, highly reproducible successions.

(a) Absolute abundance trajectories for individual taxa from a single colonization replicate (replicate 2). Individual trajectories are normalized to the maximum value. Colour bar indicates order-level taxonomic identities. Line plot above the heat map shows the logistic fit to the total bacterial abundance trajectory. (b) Maximum abundance per particle attained by each taxon. Error bars are s.d.'s (n=3). (c) Absolute abundance trajectories of three representative taxa across colonization replicates (, , ). Grey lines indicate the median trajectories. (d) Histogram of cross-replicate correlations for individual taxa (Methods). (e) Shannon diversity () over time for the three colonization replicates (, , ). Samples for which sequencing coverage was insufficient for the Shannon diversity to saturate have been omitted. The solid grey line (-) indicates the initial Shannon diversity of the seawater.

Figure 3. Differences in functional traits between phase II- and phase III-dominant taxa.

In both subpanels, phases of colonization (I, II and III) are indicated with the grey colour bar. (a) The fraction of read annotations mapped to a given functional category over time. DNA Pol I, DNA polymerase I (EC 2.7.7.7); GH18 family, glycoside hydrolase family 18; CBP, chitin-binding protein (auxiliary activity family 10); Chemotaxis, N-acetylglucosamine-regulated methyl-accepting chemotaxis protein; DeAc, N-acetylglucosamine-6-phosphate deacetylase (EC 3.5.1.25); DeAm, glucosamine-6-phosphate deaminase (EC 3.5.99.6); Chitobiose catabolism, (GlcNAc)₂ Catabolic Operon (SEED Subsystem). (b) Left heat map: absolute abundance trajectories of isolated taxa. Leftmost letter identifiers show order-level taxonomic identities. Right heat map: whether isolates do (blue) or do not (black) display a functional trait (assays described in Methods; Supplementary Methods). Grey: within-taxon isolates differ in their phenotype. The number of isolates surveyed per taxon ranged from 1 to 3.

Figure 4. Phase III-dominant taxa that cannot grow on chitin alone grow in co-culture with chitin degraders.

Fold growth of two phase III-dominant taxa (OTUs 37 and 30) in monoculture (‘Mono', grey) and in co-culture with chitin-degrading partners (‘OTU X', blue/yellow). Blue bars: partners that broadcast extracellular chitinases. Yellow bars: partners that do not broadcast extracellular chitinases. Strains and their co-culture partners were characterized taxonomically with the Ribosomal Protein Database (RDP) classifier. The lowest level of classification with >80% confidence is indicated for each strain. Asterisks: when fold growth in co-culture is significantly different than in monoculture (two-tailed t-test; *P<0.05, **P<0.01, ***P<0.001). Error bars are standard deviations over three biological replicates.