Ecological success of extreme halophiles subjected to recurrent osmotic disturbances is primarily driven by congeneric species replacement

Abstract

To understand how extreme halophiles respond to recurrent disturbances, we challenged the communities thriving in salt-saturated (~36% salts) ~230 L brine mesocosms to repeated dilutions down to 13% (D13 mesocosm) or 20% (D20 mesocosm) salts each time mesocosms reached salt saturation due to evaporation (for 10 and 17 cycles, respectively) over 813 days. Depending on the magnitude of dilution, the most prevalent species, Haloquadratum walsbyi and Salinibacter ruber, either increased in dominance by replacing less competitive populations (for D20, moderate stress conditions), or severely decreased in abundance and were eventually replaced by other congeneric species better adapted to the higher osmotic stress (for D13, strong stress conditions). Congeneric species replacement was commonly observed within additional abundant genera in response to changes in environmental or biological conditions (e.g. phage predation) within the same system and under a controlled perturbation of a relevant environmental parameter. Therefore, a genus is an ecologically important level of diversity organization, not just a taxonomic rank, that persists in the environment based on congeneric species replacement due to relatively high functional overlap (gene sharing), with important consequences for the success of the lineage, and similar to the success of a species via strain-replacement. Further, our results showed that successful species were typically accompanied by the emergence of their own viral cohorts, whose intra-cohort diversity appeared to strongly covary with, and likely drive, the intra-host diversity. Collectively, our results show that brine communities are ecologically resilient and continuously adapting to changing environments by transitioning to alternative stable states.

Keywords: congeneric species replacement; metagenomics; osmotic disturbances; time-series; viral cohort.

Figure 1

Experimental setup of the study. (A) Mesocosm appearance after 813 days, oscillations of (B) temperature (°C), (C) salinity (%), (D) total (DAPI), and (E) bacterial/archaeal (FISH) cell count dynamics (cells/ml) over the 813 days of long-term osmotic disturbance.

Figure 2

Temporal dynamics of prokaryotic and viral communities over 813 days of disturbance. Non-metric multidimensional scaling (NMDS) performed on (A) MASH distances. NMDS using Bray-Curtis dissimilarities based on relative abundances of (B) MAGs, (C) OPUs, and (D) vOTUs. Symbol shapes represent the mesocosm of origin for the samples, while the gradient indicates temporal succession. Black squares indicate time zero (10 June 2020).

Figure 3

Shifts of species abundance and interspecific genus replacement. Relative abundance of Halorubrum, Haloquadratum, and Salinibacter species in the D13 and D20 metagenomes (n = 130) was estimated using sequencing effort (sequencing depth divided by number of metagenomic reads). Darker bubbles show the sequencing effort of each genus across the metagenomic dataset, while the lighter bubble represents the relative abundance of each species within each genus. Phylogenetic reconstruction using the concatenated core orthologous proteins of Halorubrum, Haloquadratum, and Salinibacter species extracted from our experiment. Both phylogenetic trees were reconstructed with a neighbor joining algorithm using the Kimura correction. The bar indicates the 10% sequence divergence in each tree.

Figure 4

Dynamics of viral cohorts (VCs) predicted to infect Hqr.walsbyi, Hqr. sp2, Sal.ruber, and Sal. sp6. Relative abundances were calculated based on sequencing effort normalized by the total abundance of viral cohorts putatively infecting the same host across the metagenomes (n = 130).

Figure 5

Correlation between viral cohort diversity and intraspecific microbial population shifts (ANIr). (A) Pearson’s correlation between the ANIr of MAGs mapped back to their source metagenome versus the Shannon index of the viral cohorts infecting Hqr. walsbyi, Hqr. sp2, Sal. ruber, and Sal. sp6. (B) Pearson’s correlations of ANIr of MAGs mapped back to their source metagenome for Hqr. walsbyi, Hqr. sp2, Sal. ruber, and Sal. sp6 over time in D13 and D20 mesocosms.