The dynamics of cable bacteria colonization in surface sediments: a 2D view

pH distributions within sediment microcosms showed distinct spatial and temporal patterns. For the January 2019 experiment series, the strong pH maximum bands developed in the oxic surface sediment after 20 ~ 22 days of incubation. Development was not spatially uniform. Sediment surface pH maxima started to develop from isolated points covering horizontal lengths of 0.6–1 cm at the times of imaging (Fig. 1). Within a week, the pH maximum bands expanded laterally, covering the entire 6 cm long monitoring panel, and were sustained until the end of the experiment (106 days, data not shown). The pH maximum bands were about 2 mm in vertical extent with average pH ~ 8.5. Even after the electrogenic colonization was laterally complete (day 27), the activity intensities of cable bacteria, or the impacts of their activity as reflected by the magnitudes of pH elevations and associated reactions, were still spatially heterogeneous (Fig. 1C). pH values in the underlying anoxic sediment decreased from 7.0 to 6.5 gradually as surface pH maxima formed. However, in the experiment with the sediment collected in August 2019, sediment surface pH maxima started appearing on day 39 and expanded much more slowly, with only a 2.4 cm-long lateral coverage after a week of growth in one of the duplicate microcosms and no development at all in the other (Fig. 2A,B). At the same time, where pH maxima were present, a pH minimum band developed in the anodic zone (> 2 mm depth) just below sediment surface pH maxima and expanded over time (Fig. 2A).

The October 2020 results (Fig. 2C) further resolved the cathodic and anodic zone development patterns. During colonization, the anoxic zone pH minima were evident earlier (day 46 and 64) and were distinctly wider than the sediment surface pH maxima (day 46–71). These differences in cathode and anode detection sensitivity might be caused by more rapid diffusion at the sediment–water boundary (free solution) and rapid neutralization by seawater CO₂. They may also be related to the mode of colonization as discussed below. In addition to pH heterogeneity, the electrogenic activity also resulted in complex topographies of H₂S distributions (Fig. 2D). In the locations where pH hotspots were found, the depths of detectable H₂S are deeper compared with anywhere else, consistent with ongoing electrogenic sulfide oxidation metabolic activity. These data suggest that cable bacteria dynamics can be distinctly different in otherwise similar sediment (e.g. similar concentrations of dissolved H₂S at depth), with variable development controlled by unknown factors.

High resolution cable bacteria abundance data are not available in this study because of the design of the experiment. However, the vertical cable bacteria abundance dynamics, which were retrieved from random locations in the January 2019 microcosms during incubation from day 20 to 43, show that cable bacteria cell abundance in the oxic zone of the sediment did not vary significantly during the primary colonization period. In contrast, one of the microcosm series (Fig. 3B) showed a distinct trend of subsurface, anodic region (0.5–1.5 cm depth) enrichment of filament cells, and importantly, both microcosm series had similar terminal distributions when electrogenic activity had expanded across the entire surface (day 43) (Fig. 3). The first time sample in series A (day 20) (Fig. 3A) is similar in abundance distribution as at day 43. It is possible that the first sample taken in series A was located in an electrogenic colonization patch, that is, locally comparable to what would become the pattern across the entire surface at day 43. These abundance data together with the high resolution pH patterns allow inference of the colonization strategy of cable bacteria, as outlined subsequently.

One noticeable difference among the three sets of incubation series samples was sediment porosity, which impacts the diffusive flux of H₂S causing different cable bacteria colonization dynamics (Figs. 1 and 2) . The porosity did not vary greatly with depth: average 0.89 (January 2019), 0.65 (August 2019), 0.61 (October 2020). In these Mn/Fe-poor carbonate sediments, dissolved sulfide was the only sulfide source for electrogenic metabolism. The H₂S flux into the anodic reaction zone is controlled by H₂S concentration gradients, the whole sediment diffusion coefficient (D_s) and porosity ((varphi )). Similar H₂S distribution profiles were observed between the January 2019 and August 2019 experimental series but it was less sulfidic for the October 2020 series (Fig. 4). Total sulfide fluxes are calculated as 0.64–0.67, 0.17–0.26 and 0.05 µmol cm⁻² day⁻¹ in the anodic zone of January, August and October sediments, respectively. The differences in H₂S availability and flux mainly came from the diffusion coefficient and porosity variations, although gradients are also lower in the October series. Diffusive fluxes are strongly affected by porosity. According to the approximate relationship of tortuosity and diffusion coefficients (D_s ~ D₀ × ({varphi}^2); D₀ = free solution diffusion coefficient; φ = porosity)¹⁸, the diffusion coefficients in January 2019 sediment experiments were about 2 times larger than August 2019 experiments. Based on Fick’s first law (J = − (varphi ) D_s × (frac{dC}{dz})), the diffusion rates in January 2019 sediment microcosms were about 3 times faster than the August 2019 sediment microcosms (i.e., the ratio of φ³ between the incubation series assuming equal concentration gradients). Unlike lithogenic sediments, where the sulfide source can come from both solid phase (FeS) and dissolved phase, lower diffusion rates of sulfide may restrict the development rates of cable bacteria when the source of sulfide is almost exclusively dissolved, such as in carbonate muds from Florida Bay. Here we found that the cable bacteria colonization time is negatively correlated with the H₂S fluxes (Fig. 5). We also propose that the variable cable bacteria dynamics in otherwise similar sediments might have also come from seasonal differences (e.g. DOC sources) in the natural condition of cable bacteria at the different times when sediment was collected.

While the difference in the pH dynamics in the microcosm series are consistent with different sulfide fluxes, remineralization rates, and metabolic rates, they may also reflect differences in diffusive transport of reaction products and rates of neutralization reactions. For example, the low porosities of the August 2019 and October 2020 sediment may result in lower diffusion rates of protons in the sediment and enhance detection of pH minima (Fig. 2A–C), whereas maxima are located at the sediment water boundary where transport is enhanced in free solution. In contrast, the high porosity of January 2019 sediment may tend to obscure detection of proton production or consumption and also minimize differences in transport rates between the sediment–water boundary and deeper in the deposit (Fig. 1A,B).

The manner by which the sediment surface pH maxima and anoxic sediment pH minima spread, together with the cell abundance patterns with time and depth, suggests possible strategies of cable bacteria during colonization. It is known that cable bacteria can be disseminated throughout sediments and utilize multiple redox potentials, implying that they can have multiple modes of metabolism, that the electrogenic metabolic pathway is likely facultative, and that the formation of long filaments is an opportunistic morphology^7,19. We assume that initiation of the electrogenic metabolic mode occurs randomly as a filament fragment becomes located at a stable redox interface and establishes a link between oxygen and sulfide. Once locally initiated, cable bacteria could further divide and grow, or migrate and aggregate, or both. Our data suggest that the spread of the high pH region is not associated with a large increase in bulk abundance of bacterial cells in the surface oxic layer, but do indicate that cell abundance increases in the subsurface, anoxic zone. Recently, Geerlings et al.²⁰ showed that there is a labor division within individual cable bacteria filaments and that apparently only the anoxic–anodic end conserves energy. Assuming growth is largely focused into the anodic end, our cell abundance data would be consistent with a subsurface coiling of filaments⁹ rather than a strictly linear, vertical orientation, that is, a cathodic snorkel coupled to growing subsurface anodic cells.

Cable bacteria are known to be capable of moving through sediment¹⁰. The rapid expansion of the surface pH maximum band and deep sediment minima would also be consistent with migration of filament anodic ends toward randomly established electrogenic sites and coiling of filaments. The anoxic sediment pH minima was initially detected near the oxic sediment surface, and expanded both horizontally and vertically as a surface pH maxima enlarged (Fig. 2C), indicating the enlargement of electrogenic reaction zone in both the vertical and horizontal dimensions. A primary factor driving migration and localized growth in the initially established anodic zone may be low H₂S concentration and high H₂S fluxes. Once a subsurface site of decreased pH is randomly formed by cable bacteria, they may form aggregations through migration and growth in the vicinity of those hotspots. Within those hotspots, the diffusive flux of sulfide is maximum and the lower pH favors speciation of HS^– to H₂S and also dissolves any FeS present, further enhancing H₂S, which may favor the electrogenic metabolic activity of cable bacteria. Due to the expansion of the electrogenic anodic zone in the anoxic sediment, the cathodic zone, which is located in the sediment oxic layer and is characterized by pH elevation, also expands. Considering the positive feedback proposed, hotspots are ideal locations for electrogenic sulfide oxidation. The electrogenic colonization patterns we observed can also be explained by the increase of specific cell electrogenic reaction activity within the filaments already present.

The differences in development times of electrogenic activity in the different experiments demonstrate directly that there is no one timescale that defines cable bacteria establishment at a specific location in deposits, and that a wide range of colonization timescales is possible, in our cases between ~ 20 and 50 days with a direct correlation of colonization time and the estimated diffusive flux of H₂S (Fig. 5). No matter what the colonization strategy actually used by cable bacteria (aggregation, growth of filaments, or specific cell activity stimulation within filaments), electrogenic activity shows spatially heterogeneous patterns during initiation which become more uniformly distributed as patches of cable bacteria and their activity expand both horizontally and vertically, and eventually blend together across larger areas. However, based on 2D pH patterns, cable bacteria activity, or the impacts of their activities on reaction balances, may remain spatially heterogeneous (Fig. 1C). A major implication of these electrogenic colonization patterns is that their impacts on diagenetic reaction balances and sediment–water solute fluxes are similarly highly heterogeneous and strongly time dependent.

Source: Ecology - nature.com