1. Assessment of environmental parameters in the study area
Sampling sites and the studied area are represented in Figure 1. During the sampling campaign, the temperature of superficial seawater was 23 °C whereas at 10-m depth of 10 °C. Data on the concentrations of proteins (PRT), carbohydrates (CHO) and lipids (LIP) as well as chlorophyll-a (Chl-a), phaeopigments (PHEO) and biopolymeric C (PBP) in the sediments of the investigated area are reported in Supplementary Table 3. In the study area, PRT concentrations were the dominant class of the investigated organic compounds, ranging from 1.4 to 7.96 mg/g (mean value of 4.18 ± 0.39 mg/g). CHO concentrations varied from 0.27 to 1.24 mg/g (mean value: 0.66 ± 0.06 mg/g), while LIP ranged from 0.25 to 1.62 mg/g (mean value: 0.74 ± 0.07 mg/g). Chl-a and PHEO concentrations in the sediments were, on average, 1.22 ± 0.22 µg/g and 14.66 ± 1.04 µg/g, respectively (range: 0.35-5.37 µg/g for Chl- and 6.49 to 26.91 µg/g for PHEO). Finally, the range of variability of BPC concentrations was comprised between 0.92 and 5.14 mg/g, with a mean value of 2.86 ± 0.26 mg/g.
2. Composition of pelagic and sediments microbiomes
The V3/V4 regions from the 16s rRNA gene was successfully sequenced from a total of 44 samples (19 waters and 25 sediments), providing 549’318 high quality reads (12’485 ± 3’235 per sample) clustered in 8’271 Amplicon Sequence Variants (ASVs) (206.8 ± 94.4 per sample). None of the detected ASVs have been assigned at the species level, while the assignment rates at the genus, family and order levels were 42, 48 and 48%, respectively. The total diversity at ASVs level was 150 for the water microbiome and 218 for the sediment one. When we assessed for the total assigned diversity at the different phylogenetic ranks, the order level showed the highest value (27.61), respect to family and genus levels scoring 23.8 and 12.96, respectively. In Supplementary Figure 1 the general compositional structure of the water and sediment microbiomes at the order level is provided. For the pelagic microbiome, the dominant orders were: Synechoccus-like Cyanobacteria Subsection I (relative abundance, rel.ab., 13.4%), Flavobacteriales (rel. ab. 12.3%), Rickettsiales (rel. ab. 8.2 %), Oceanospirillales (rel. ab 7.1%), Cellvibrionales (rel. ab. 6.5%) and Rhodobacterales (rel. ab. 5.3%). Among the subdominant fraction, the most represented orders were: SAR11 clade (rel. ab. 4.5%), Vibrionales (rel. ab. 4.2 %), Planctomycetales (rel. ab. 3.4%), Rhodospirillales (rel. ab. 3.3 %), Sphingobacteriales (rel. ab 2.9%) and Verrucomicrobiales (rel. ab. 2.3%). Differently, sediments were dominated by Campylobacterales, Clostridiales, Desulfobacterales, Bacillales and Holophagae-Subgroup 10, showing relative abundances of 10.3, 7.8, 7.4, 5.0 and 5.4%, respectively. For the benthic microbiome, the most represented minor components were: Acidimicrobiales (re. ab. 2.7%), Planctomycetales (rel. ab. 2.7%), Xanthomonadales (rel. ab. 2.6%), Rhizobiales (rel. ab. 2.5%), Vibrionales (rel. ab 2.1%), Holophagae-Subgroup 23 (rel. ab 1.5%), Lactobacillales (rel. ab 1.3%) and Flavobacteriales (rel. ab 1.2%). In order to explore specificities and connections between pelagic and sediment microbiomes, we provide the bubble plots of the distribution of the different microbiome orders in the two ecosystems (Figure 2). According to our findings, Synechoccus-like Cyanobacteria Subsection I, SAR11 clade and Puniceicoccales were characteristic of the pelagic microbiome, while Holophagae-Subgroup 10, Bacteroidales, Bacillales, Lactobacillales, Clostridiales, Desulfobacteriales and Campylobacteriales and Xanthomonadales were almost exclusive for the sediment one. However, several orders were shared between the two ecosystems. In particular, Flavobacteriales, Rickettsiales, Cellvibrionales, Oceanospirillales, Alteromonadales, Rhodobacteriales, and Vibrionales were most abundant in water samples, and Verrucomicrobiales, Rhizobiales, Planctomycetales and Acidomicrobiales were almost equally represented in water and sediment samples.
3. Changes in abundance and diversity of the pelagic and sediment microbiomes at the local scale
To identify change in the microbiome patterns in the study area, we first accounted for changes in alpha and beta-diversity across the 3 transects. Specifically, to highlight alpha-diversity patterns, the area plots of the Shannon index variation in water and sediments samples were computed (Figure 3). For both the pelagic and sediment microbiomes high and low alpha-diversity sectors have been identified, showing significant differences in Shannon diversity values. For the pelagic ecosystem, the North and the South sectors were characterized by microbiomes with higher alpha-diversity, compared with the Central sector (Figure 3A). Similarly, for the sediment microbiome, two high alfa-diversity sectors were identified in the North-East and North-West sectors, while a lower diversity area was detected in the Southern sector (Figure 3B). We subsequently assessed beta-diversity patterns in the study area. To this aim, the PCoA of the ASVs variation in water and sediment microbiome samples was carried out. According to our findings, for the pelagic microbiome, samples belonging to the previously identified alpha-diversity sectors - South, Central and North - significantly segregated in the PCoA plot (Adonis; p = 0.001; Figure 4A). Similarly, for the sediment microbiome, samples segregated according to the corresponding alpha-diversity sectors (North-Est, North-West and South) (Adonis, p = 0.006) (Figure 4B). When we searched for the correlations between PCoA coordinates and water column depth or distance from the coast, significant relationships were detected for both pelagic and sediment microbiomes (Supplementary Figure S2). For the pelagic microbiomes, the MDS1 significantly correlated with depth (R = 0.9, p < 0.005) and distance from the coast (R = 0.9, p < 0.005), while, for the sediment microbiome, we obtained analogous significant correlations but with MDS2, R = 0.25, p < 0.01. Further, when we assessed correlations among samples alpha-diversity and PCoA coordinates, a positive correlation with MDS1 was observed for sediment microbiomes (R = 0.6, p < 0.005), while only tendencies were obtained for the pelagic microbiome. When we accounted for differences in the biochemical composition in sediments corresponding to the sectors, we observed a higher concentration of all biochemical components of sedimentary organic matter (proteins, carbohydrates, lipids and total phytopigments) in the Northern sectors (Figure 5). Finally, as expected, the correlation between the UniFrac distances matrix of sediment microbiomes samples and the correspondent distance matrices of the biochemical composition was significant (Table 1) (Mantel Test in R).
Table 1: Output of Mantel test analyses on the Spearman correlation of the unweighted UniFrac distances of microbiome structure and distance matrix of biochemical components generated with dist function in R (method = “Euclidean”), number of permutations: 9999.
|
Mantel Statistic r
|
Significance (p value)
|
Carbohydrates x UniFrac distances
|
0.2552
|
2.7e-03
|
Proteins correlation x UniFrac distances
|
0.2701
|
6e-04
|
Lipids correlation x UniFrac distances
|
0.3118
|
7e-04
|
Phytopigments correlation x UniFrac distances
|
0.3275
|
1.6e-3
|
4. Variation in the pelagic and sediment microbiomes network structure at the local scale
With the attempt to better identify the community-level implications of diversity patterns observed for the pelagic and sediments microbiomes in the study area, a network-based approach was applied. To this aim, the overall network structure of the pelagic and sediment microbiomes was obtained, then the correspondent declinations in the different sectors have been assessed. For the creation of the overall microbiome networks, the co-abundance associations between orders were assessed, then orders were clustered in co-abundance groups CAGs (Supplementary Figure S3). For both ecosystems, 3 different CAGs were detected and named according to the dominant order. The CAGs composition is provided in the Supplementary Table 4. For the pelagic microbiome, the detected CAGs were Rhodobacteriales CAG, the Vibrionales CAG and the Falavobacteriales CAG, while for the sediment microbiomes the correspondent CAGs were Desulfobacteriales CAG, Clostridiales CAG and Campylobacteriales CAG. In Figure 6, we provided the Wiggum plots of the overall network structure of the pelagic and sediment microbiomes, where the compositional relationships between the correspondent CAGs are represented.
The variation of the pelagic and sediment networks in the different sectors were than assessed. To this aim, for both ecosystems, the specific pattern of over-abundance modules (CAGs) and nodes (orders) for each sector was computed, and the respective over-abundant network plots were created (Figure 7 and 8). The box plot showing the variation in relative abundance of the over-abundant CAGs and orders in each sector are provided in Supplementary Figure 4.
According to our findings, for the pelagic microbiomes, the 3 sectors showed a specific pattern of over-abundance CAGs. Particularly, the Falvobacteriales CAG was most abundant in the North sector, while the Vibrionales CAG and Rhodobacteriales CAG were most represented in the Central and South area, respectively. Focusing on the single orders, each sector showed a specific set of over-abundant components: (i) for the North sector: Xanthomonadales and E01-9C-26 marine group (for the Flavobacteriales CAG), MB11C04 marine group, Vibrionales and KI89A clade (for the Vibrionales CAG) and Clostridiales (for the Rhodobacteriales CAG) resulted over-abundant; (ii) for the Center sector the over-abundant orders were Alteromonadales (for the Flavobacteriales CAG), Planctomycetales, Phycisphaerales, Sphingomonadales, Micrococcales, Puniceicoccales, Burkholderiales and Rickettsiales (for the Vibrionales CAG), Bradymonadales and Bacteroidetes Order II (for the Rhodobacteriales CAG); (iii) for the South sector were over-abundant the following orders all belonging to the Rhodobacteriales CAG: Chlamydiales, Rhizobiales, Sphingobacteriales, Bacteroidetes Orders II and III, Bdellovibrionales, Legionellales, Acidimicrobiales, Parvularculales, SAR11 clade, Myxococcales and Rhodobacterales. For what concern the sediment microbiome, an analogous situation was observed. In particular, at the CAGs level, the North-West sector was enriched in the Campylobateriales CAGs, while depleted in Desulfovibrionale CAGs and Clostridiales CAGs compared to North-East and South sectors. For what concerns the orders, the following site-specific over-abundant pattern was observed: (i) NB1-j, Thiotrichales, Myxococcales, Anaerolineales, Gaiellales, Chlamydiales, Holophagae Subgroup 23 (for the Desulfovibrionale CAGs), Pseudomonadales and Fusobacteriales (for the Clostridiales CAGs) and Nitrospirales, Gemmatimonadales, Bacillales, Campylobacterales, Phycisphaerales (for the Campylobateriales CAGs) were over-abundant in the North West sector; (ii) Bacteroidetes Order II, Cytophagales , HTA4 (for the Desulfovibrionale CAGs), Pseudomonadales Vibrionales (for the Clostridiales CAGs) and Desulfurellales (for the Campylobateriales CAGs) were over abundant in the North East sector; (iii) Alteromonadales, KI89A clade, Rhodobacterales, Spirochaetales, Solirubrobacterales, SAR324 clade(Marine group B), Cellvibrionales, Burkholderiales and NB1-j (for the Desulfovibrionale CAGs), Oceanospirillales, Lactobacillales, Corynebacteriales, Clostridiales, Micrococcales, Flavobacteriales (for the Clostridiales CAGs) and Desulfarculales (for the Campylobateriales CAGs) were over abundant in the South sector. Finally, for both the water and sediment microbiome, site-specific community networks were created for each of the 3 alpha diversity sectors. For each local network, correspondent key parameters in term of modularity, total connectivity, negative to positive cohesions and hubs orders are provided in Table 2.
Table 2: The table represents the parameters of Networks of the single Area, in terms of Negative to Positive Cohesion Ratio (N:P), Modules, Total connectivity and hubbs order of the Network, both for pelagic microbiome (first three rows of the table) and sediment microbiome (last three rows of the table).
Single Networks Parameters
|
|
N:P Cohesion Ratio
|
Modules
|
Total Connectivity
|
Hubbs Species
|
Northern Area -Water
|
1.39
|
22
|
1.06
|
Sphingobacteriales
|
Cytophagales
|
Central Area -Water
|
1.63
|
25
|
0.69
|
Planctomycetales
|
Southern Area -Water
|
0.39
|
20
|
0.77
|
Acidomicrobiales
|
Rhizobiales
|
N/W Area - Sediment
|
0.95
|
25
|
0.58
|
Rhizobiales
|
N/E Area - Sediment
|
1.5
|
39
|
0.71
|
Planctomycetales
|
Southern Area - Sediment
|
0.88
|
26
|
0.80
|
Micrococcales
|
Desulfobacterales
|