Fig 1.
Five individuals each of four mussel species and three samples of seston were collected between July 28 and September 16, 2016 from each site shown along a ~50 km stretch of the Sipsey River, Alabama, USA.
Temperature, pH, specific conductance (μS/cm), conductivity (μS/cm), dissolved oxygen (mg/L), and ammonia, orthophosphate, nitrate, and nitrite from both the water column and the sediment (μg/L) were collected at three times from each site between June 9 and September 28, 2016. The river flow is from northeast to southwest.
Fig 2.
Major bacterial phyla in the gut microbiome of four freshwater mussel species (in order shown: Cyclonaias asperata, Fusconaia cerina, Lampsilis ornata, and Obovaria unicolor) at six sites in the Sipsey River, AL, USA, and for suspended seston collected from the same sites, as determined by Illumina 16S rRNA gene sequencing.
Stacked bar plots are arranged with the most abundant phylum overall (Proteobacteria) at the top. No O. unicolor samples were found at site 1. The seven most abundant phyla are shown and represent 80% of all sequencing reads. Unclassified sequences made up 16% of the total dataset. All other bacterial phyla are grouped together as “other.” Each of the phyla shown differed significantly in relative abundance between the mussel gut and overlying seston with the exception of Proteobacteria.
Fig 3.
Major bacterial genera in the gut microbiome of four freshwater mussel species (in order shown: Cyclonaias asperata, Fusconaia cerina, Lampsilis ornata, and Obovaria unicolor) at six sites in the Sipsey River, AL, USA, and for suspended seston collected from the same sites, as determined by Illumina 16S rRNA gene sequencing.
Stacked bar plots are arranged with the most abundant genus overall (Clostridium sensu stricto) at the top. No O. unicolor samples were found at site 1. The 24 most abundant genera are shown and represent 73% of all sequencing reads that were identified to the genus level. Sequences classified at the genus level made up 35.7% of the total data. All other bacterial genera are grouped together as “other”.
Fig 4.
Shannon evenness, Chao1 richness, and Inverse Simpson diversity of the gut microbiota of four freshwater mussel species (in order shown: Cyclonaias aspertata, Fusconaia cerina, Lampsilis ornata, and Obovaria unicolor) at six sites in the Sipsey River, AL, USA, and for seston collected from the same sites, as determined from Illumina 16S rRNA gene sequencing.
Seston bacterial evenness, richness, and diversity were higher than those metrics observed in the mussel gut (p < 0.001). L. ornata the lowest evenness and diversity scores of the four mussel species. Error bars represent standard error of the 22–29 samples of each mussel species collected, or the 18 seston samples collected across all sites.
Fig 5.
NMDS ordinations of 16S rRNA bacterial microbiome data collected from the mussel species Cyclonaias asperata, Fusconaia cerina, Lampsilis ornata, and Obovaria unicolor from the Sipsey River, AL, USA, in addition to suspended seston.
Important OTUs driving the ordination are indicated with an arrow, with arrow length proportional to effect size and arrow direction reflecting association with those samples. A. Bray-Curtis ordination of samples categorized by associated environment (mussel vs. seston). B. Bray-Curtis ordination of mussel samples only. Species are represented by shape and sites are represented by colors. C. Bray-Curtis ordination of seston samples only. Sites are represented by colors.
Table 1.
Pairwise species and site comparisons for gut microbiomes of freshwater mussels collected from six sites in the Sipsey River, AL, USA.
Microbiome comparisons were based on the Bray-Curtis dissimilarity metric. Mussel species are Cyclonaias asperata, Fusconaia cerina, Lampsilis ornata, and Obovaria unicolor and the six sites cover a ~50 km stretch of the river, numbered from upstream to downstream. P-values come from a PERMANOVA comparing bacterial community composition by host mussel species and by site. The test excluded seston samples which were significantly different from each mussel species at each site (p <0.01 for all). The R package pairwiseAdonis was used to perform a post-hoc test to determine which pairwise combinations of species and site were significant. Adjusted p-values of this post-hoc test are displayed below.
Fig 6.
A constrained principal coordinates analysis (CAP) of 16S rRNA bacterial microbiome data collected from the mussel species Cyclonaias asperata, Fusconaia cerina, Lampsilis ornata, and Obovaria unicolor as well as seston from the Sipsey River, AL, USA.
Bray-Curtis dissimilarity was used to separate samples by site differences in physicochemical parameters. Three measurements of each environmental parameter were recorded between June 9 and September 28, 2016. CAP identified surface water ammonium, porewater ammonium, surface phosphate, pore phosphate, and surface nitrite as the most critical factors of all the measured parameters and those factors were retained in a reduced model. Surface measurements were collected from the flowing water body and pore measurements were taken from pore water. A. Within the mussel gut samples, these physicochemical parameters explained very little variability, combining 7.9% across each of the first two axes. B. Within the seston samples, environmental chemistry explained 44.5% of bacterial compositional differences across the two axes.
Table 2.
Physicochemical data from six sites along the Sipsey River, AL, USA.
“Surface” measurements reflect readings from the water column and “pore” reflects porewater readings. Sites are ordered moving downstream and the distance from the first site is noted. Three measurements of each environmental parameter were recorded between June 9 and September 28, 2016. Data shown reflects the means of these three measurements and the standard error.