Viral community diversity in the rhizosphere of the foundation salt marsh plant Spartina alterniflora

Abstract

Viruses of microorganisms impact microbial population dynamics, community structure, nutrient cycling, gene transfer, and genomic innovation. In wetlands, root-associated microbial communities mediate key biogeochemical processes important for plants involved in ecosystem maintenance. Nonetheless, the presence and role of microbial viruses in salt marshes remain poorly understood. In this study, we analyzed 24 metagenomes retrieved from the root zone of Spartina alterniflora, a foundation plant in salt marshes of the eastern and Gulf coasts of the U.S. The samples span three plant compartments-bulk sediment, rhizosphere, and root-and two cordgrass plant phenotypes: short and tall. We observed differentiation between phenotypes and increased similarity in viral communities between the root and rhizosphere, indicating that plant compartment and phenotype shape viral community composition. The majority of viral populations characterized are novel at the genus level, with a subset predicted to target microorganisms known to carry out key biogeochemical functions. The findings contribute to ongoing efforts to understand plant-associated viral diversity and community composition and to identify potential targets for exploring viral modulation of microbially mediated ecosystem functioning in intertidal wetlands.IMPORTANCESalt marshes are vital coastal ecosystems. Microbes in these environments drive nutrient cycling and support plant health, with Spartina alterniflora serving as a foundation species. This study explores viral communities associated with S. alterniflora, revealing how plant compartments and phenotypes shape viral composition. The discovery of numerous novel viruses, some potentially influencing microbes involved in key biogeochemical processes, highlights their ecological significance. Given the increasing pressures on coastal ecosystems, understanding virus-microbe-plant interactions is essential for predicting and managing ecosystem responses to environmental change.

Keywords: bacteriophages; metagenomics; microbial ecology.

Fig 1

Viral communities in the S. alterniflora root zone are shaped by plant phenotype and microbiome compartment. (a) Metagenomes were obtained from the S. alterniflora rhizosphere at established transects on Sapelo Island, Georgia, from three compartments (root, rhizosphere, and bulk sediment) and two plant phenotypes (short <50 cm and tall >80 cm), for a total of 24 samples. (b) Rarefaction curve displaying the cumulative number of viral operational taxonomic units (vOTUs) found with increasing sample number; 100 permutations of sample order were performed (gray points). The mean cumulative number of vOTUs for each number of samples is represented by the blue line. The saturation plot shows that viral richness is not fully represented by the sampling and viral recovery efforts. (c) Viral richness, depicted by the number of vOTUs per phenotype per compartment. Horizontal lines represent the mean number of vOTUs among samples obtained from the same plant compartment and phenotype combination. Asterisks indicate statistical differences based on pairwise Mann–Whitney tests (two-sided, P-value < 0.05). Significant differences include viral communities of bulk sediment/short and rhizosphere/short samples (P-value = 0.03), rhizosphere/short and rhizosphere/tall samples (P-value = 0.02), and root/short and root/tall samples (P-value = 0.03). (d) Non-metric multidimensional scaling (NMDS) of Bray-Curtis dissimilarity across each sample based on vOTU relative abundance (transcripts per million [TPM]) per sample. The NMDS shows similarities between viral population abundance among the samples, revealing separation between compartments and plant phenotypes. NMDS stress: 0.11.

Fig 2

vOTU co-occurrence across phenotypes and compartments. (a) Number of samples in which a single vOTU is found. Most vOTUs (444 out of 769) are found in two or more samples (black bars), with two sequences found in 19 out of the 24 samples. (b) Comparison of plant phenotypes associated with vOTUs that appear in two or more samples (observed data) to a null model of co-occurrence (expected). Fewer vOTUs are found in samples obtained from plants with different phenotypes (gray section in left bar) than expected (gray section in right bar). Statistical differences based on χ² test (P-value < 0.05). (c) Comparison of microbiome compartments associated with vOTUs that appear in two or more samples (observed data) to a null model of co-occurrence (expected). Fewer vOTUs are found in samples from all three soil compartments (gray section in the left bar) than expected (gray section in the right bar). More vOTUs are found in samples from two different compartments than expected (stripe sections), with the largest increase in vOTUs shared between the rhizosphere and root. No vOTUs are found to be shared exclusively between the bulk and root compartments. Fewer vOTUs are found in samples from the same compartment than expected (solid colors). Statistical differences based on χ² test (P-value < 0.05).

Fig 3

Taxonomic characterization of viral populations. Viral taxonomy of the vOTUs was predicted using vContact2. vOTUs in the same cluster are expected to share genus-level similarities. Of the 769 vOTUs, 421 were assigned to 207 clusters, while 348 were classified as singletons. Among the clustered vOTUs, 102 clusters containing 139 vOTUs matched viral populations previously reported in other environmental studies. The network displayed here represents viral clusters, with edges indicating only the direct similarities to the vOTUs identified in this study. Other potential interactions among vOTUs not described in this study are not depicted.

Fig 4

Host prediction reveals viral targeting of microorganisms implicated in important biogeochemical cycles in salt marshes. Hosts were predicted for the vOTUs associated with S. alterniflora. Fifty-five unique hosts were predicted at the phylum level for 64 vOTUs. Percentages above bars represent the proportion of vOTUs for which hosts were predicted out of all vOTUs found per phenotype and compartment combination. The absolute numbers of vOTUs with hosts predicted for each sample type were 21, 11, 43, 19, 19, and 7 for short bulk sediment, tall bulk sediment, short rhizosphere, tall rhizosphere, short root, and tall root, respectively. Each identified phylum is represented by a different color.