Colonization in the photic zone and subsequent changes during sinking determine bacterial community composition in marine snow

Abstract

Due to sampling difficulties, little is known about microbial communities associated with sinking marine snow in the twilight zone. A drifting sediment trap was equipped with a viscous cryogel and deployed to collect intact marine snow from depths of 100 and 400 m off Cape Blanc (Mauritania). Marine snow aggregates were fixed and washed in situ to prevent changes in microbial community composition and to enable subsequent analysis using catalyzed reporter deposition fluorescence in situ hybridization (CARD-FISH). The attached microbial communities collected at 100 m were similar to the free-living community at the depth of the fluorescence maximum (20 m) but different from those at other depths (150, 400, 550, and 700 m). Therefore, the attached microbial community seemed to be “inherited” from that at the fluorescence maximum. The attached microbial community structure at 400 m differed from that of the attached community at 100 m and from that of any free-living community at the tested depths, except that collected near the sediment at 700 m. The differences between the particle-associated communities at 400 m and 100 m appeared to be due to internal changes in the attached microbial community rather than de novo colonization, detachment, or grazing during the sinking of marine snow. The new sampling method presented here will facilitate future investigations into the mechanisms that shape the bacterial community within sinking marine snow, leading to better understanding of the mechanisms which regulate biogeochemical cycling of settling organic matter.

FIG 1

Map of the study area off Cape Blanc, Mauritania. Water samples were collected at depths of 20, 150, 400, 550, and 700 m at station GeoB15704 to quantify the free-living microbial community. Additionally, a surface-tethered freely drifting sediment trap was deployed at station GeoB15704 to collect intact, settling marine snow at 100 and 400 m. The inset shows the track of the surface-tethered drifting sediment trap.

FIG 2

(Top) Drawing of one of the collection cylinders deployed on the sediment trap. The cylinder was filled with the different densities of GF/F-filtered seawater. The lower layer was denser than the layers above. This created three distinct density layers. By adding formaldehyde to the middle layer, we were able to fix the microbial community within marine snow and other aggregates sinking through this layer. The bottom density layer did not contain any formaldehyde and served to wash the fixed marine snow and other aggregates to avoid overfixation of the attached microbes, which prevents the use of fluorescence in situ hybridization. A collection cup filled with a viscous cryogel was placed at the very bottom of the sediment trap cylinder. The viscous gel collected the fixed and washed marine snow and other aggregates and preserved their size and structure. (Bottom) Images of the aggregates collected in the cryogel at 100 m (left) and 400 m (right).

FIG 3

Comparison of relative abundances of the attached bacterial community found within marine snow aggregates collected at 100 m and 400 m. Abundances of Synechococcus (SYN405), Bacteroidetes (CF319a), Roseobacter (ROS537), Gammaproteobacteria (GAM42a), Alteromonas (ALT1413), Pseudoalteromonas (PSA184), and Planctomycetes (PLA46) cells are shown. The asterisks indicate the clades which were significantly different between 100 and 400 m.

FIG 4

(Left) Dendrogram showing hierarchical cluster analysis using the relative abundances of only the microbial groups observed within marine snow. The dendrogram shows the distance between the attached microbial community at 100 and 400 m and the free-living community at 20, 150, 400, 550, and 700 m. Spearman's rank was used to calculate the distances. (Right) Relative abundances, in percentage of attached and free-living microbes, are shown for the groups observed within the aggregates (Table 3).