Dataset: Microsocsm Incubation Series II

Microcosms were monitored as a function of depth for pH, O2, H2S, and cable bacteria abundance throughout the 3-month incubation period in order to determine the time points of core sectioning for bulk pore water solute compositions (Ca2+, Na+, Mg2+, SPO43- and total alkalinity). When sampled, the whole rectangular microcosms and cylindrical cores were vertically sectioned (in air; each core processing took less than 15 min; resolution: 0.25 cm above 2 cm depth; 0.5 cm downwards), centrifuged (3000 rpm, 10 min), and the porewater was filtered (VWR polyethersulfone syringe filters; 0.2 μm pore size) for analyses of Ca2+, Na+, Mg2+, SPO43- and total alkalinity. Series II incubation series were sectioned on day 3, 26, 43 (2 cylindrical microcosms each time; A & B), and 106 (duplicate rectangular microcosms, A & B). On day 106, overlying water anoxia was generated and maintained by continuously purging with a mixture of N2 and CO2 gases (CO2, 0.04%), while keeping the overlying water at a pH of 8 (monitored by a pH electrode). Anoxia was confirmed using an oxygen electrode. 2D pH and H2S distribution patterns in the box microcosms were imaged using planar optical sensors before and after the anoxic incubation. The concentrations of other solutes (Ca2+, Na+, Mg2+, SPO43- and total alkalinity) were measured in porewater extracted immediately after the anoxic incubation.

Porewater alkalinity was measured using the bromophenol blue colorimetric method (Sarazin et al., 1999). Reactive inorganic phosphate (H2PO4-, HPO42-, PO43-; abbreviated as SPO43-) was determined colorimetrically using molybdenum blue (Murphy and Riley, 1962). NO2- and NO3- were measured based on the acidic Griess reaction method (LOD<1 µM) (Miranda, et al.,2001).  Ca2+, Na+, and Mg2+ concentrations were determined using a Horiba JY Ultima 2C ICP-OES Spectrometer.

Samples for examination of the microbial community were collected on day 1, 20, 26, 43, and 106 from the rectangular microcosms in series II.  In series II experiments, 3-ml cutoff plastic syringes were inserted 5 cm vertically at random positions into microcosms. The subcores were subsequently sliced and preserved in 2% formaldehyde (resolution: 0.25 cm until 1 cm depth; 0.5 cm from 1 to 3 cm depth; 1 cm downwards). Aliquots of homogenized sediment samples were dried and used to normalize abundances to dry weight (Aller, et al., 2019).  Cable bacteria filament identity was confirmed by FISH with a Desulfobulbaceae-specific oligonucleotide probe (DSB706) (Manz et al, 1992).  Cable bacteria filaments were morphologically distinctive (Schauer, et al., 2014), and we could also readily identify them using acridine orange (AO) using epifluorescence microscopy.   Compared with the FISH method, AO stained samples generally showed longer filament lengths due to less handling, and better precision of cell counts. Additionally, we used AO rather than 4′, 6-diamidino-2-phenylindole (DAPI) for identifying and counting bacteria because it gave better contrast in our natural sediments.

Following the method development with the FISH technique, the lengths of cable bacteria filaments were enumerated primarily from AO stained samples. Abundances were calculated from the summed filament lengths and an average cell size of 2 µm. Sediment subsamples were first diluted 100-fold with 3% saline water and homogenized. AO stained samples were deposited on black GVS polycarbonate track-etched filters (0.2 µm pore size). Cable bacteria filaments were counted within 10 randomly selected fields (1000× magnification; defined by a Whipple reticle) including any that extended beyond the grid field (the same filament only counted once). The inclusion of parts of filaments outside the chosen grid field had the potential to overestimate the cable cell abundance as they sometimes expanded across multiple grid fields. Other filamentous bacteria like Beggiatoa can be easily excluded from erroneous enumeration based on the distinctive cell morphology. No non-filament cells were counted as cable bacteria, although they can apparently occur as isolated, single cells (Muller, et al., 2000).

The pH imaging planar optical sensors were made by immobilizing 8-hydroxy-1,3,6- pyrenetrisulfonic acid trisodium salt (HPTS) into a transparent polyester film (Zhu, et al., 2006). The responses of the sensor films were captured using a commercially available digital camera (EOS Rebel T7i), following ratiometric calibration (emission wavelength: 545 nm; excitation wavelengths: 510 nm and 430 nm). pH data reported here are on the NBS scale. Oxygen sensing films were made with platinum tetrakis (pentafluorophenyl) porphyrin (Lee and Okura, 1997).  The same camera system was used with an excitation wavelength of 395 nm and an emission wavelength of 650 nm. Oxygen concentrations were calibrated based on the Stern-Volmer relationship (Lee and Okura, 1997) using a two-point calibration: in air saturated seawater (100% air saturation) and in anoxic sediment (0% air saturation). Gas permeable H2S sensors were made using the diphenylcarbazone-Zn complex (Yin, et al., 2017). Both the sample and calibration strip responses were recorded using a flatbed scanner (Canoscan 8400F). Standard solutions were made by adjusting the pH of Na2S standards below 4 to convert sulfide into dissolved H2S (> 99.9% conversion). 2D pH O2, and H2S sensor image data were converted to 1D-vertical profiles by horizontally averaging vertically oriented membranes (widths ~ 0.5 cm). In addition to the optical sensors described above, microelectrodes for pH, O2, and H2S (pH, 100 μm tip diameter; O2 and H2S, 50 μm tip diameter; Unisense A.S., Denmark) with a manual micromanipulator and a multimeter (PA2000; Unisense A.S., Denmark) were used to resolve vertical profiles of the series II rectangular microcosms on day 1, 6, 16 and 25 for O2 and H2S, and on day 2, 19 and 27 for pH. The same standards as optical sensor calibrations were used for microelectrode calibrations.