Virus Dynamics Are Influenced by Season, Tides and Advective Transport in Intertidal, Permeable Sediments

Sandy surface sediments of tidal flats exhibit high microbial activity due to the fast and deep-reaching transport of oxygen and nutrients by porewater advection. On the other hand during low tide, limited transport results in nutrient and oxygen depletion concomitant to the accumulation of microbial metabolites. This study represents the first attempt to use flow-through reactors to investigate virus production, virus transport and the impact of tides and season in permeable sediments. The reactors were filled with intertidal sands of two sites (North beach site and backbarrier sand flat of Spiekeroog island in the German Wadden Sea) to best simulate advective porewater transport through the sediments. Virus and cell release along with oxygen consumption were measured in the effluents of reactors during continuous flow of water through the sediments as well as in tidal simulation experiments where alternating cycles with and without water flow (each for 6 h) were operated. The results showed net rates of virus production (0.3–13.2 × 106 viruses cm−3 h−1) and prokaryotic cell production (0.3–10.0 × 105 cells cm−3 h−1) as well as oxygen consumption rates (56–737 μmol l−1 h−1) to be linearly correlated reflecting differences in activity, season and location of the sediments. Calculations show that total virus turnover was fast with 2 to 4 days, whereas virus-mediated cell turnover was calculated to range between 5–13 or 33–91 days depending on the assumed burst sizes (number of viruses released upon cell lysis) of 14 or 100 viruses, respectively. During the experiments, the homogenized sediments in the reactors became vertically structured with decreasing microbial activities and increasing impact of viruses on prokaryotic mortality with depth. Tidal simulation clearly showed a strong accumulation of viruses and cells in the top sections of the reactors when the flow was halted indicating a consistently high virus production during low tide. In conclusion, cell lysis products due to virus production may fuel microbial communities in the absence of advection-driven nutrient input, but are eventually washed off the surface sediment during high tide and being transported into deeper sediment layers or into the water column together with the produced viruses.


Supplementary Figures and
. Net virus and cell production rates and oxygen consumption rates calculated from the average values of the effluents during continuous water flow through the reactors (i.e., excluding start of the experiments and values directly after low tide during tidal simulation).

Supplementary Figures
Supplementary Figure S1. Setup of the February experiment with different sediment column lengths (short, medium, long, long). From the seawater reservoir the water was pumped by peristaltic pumps on the top of the sediment columns of the flow-through reactors. For the experiment, the reservoir and the reactors were covered with black cloths.
Supplementary Figure S2. Extraction efficiency of viruses from Janssand sediment. For counting, viruses were extracted from 1 cm³ of sediment: after adding 4.5 ml 5 mM sodium pyrophosphate buffer, the extract was incubated for 15 minutes at room temperature in the dark and subsequently sonicated for 3 times for 1 minute with interruptions of 30 seconds with manual shaking. After centrifugation at 800 × g for 1 minute, the supernatant was removed and viruses were counted by flow cytometry (1st extract). The remaining sediment was washed three times by adding 5 ml MilliQ water, followed by manual shaking for 1 minute and centrifugation at 800 × g for 1 min. Viruses were counted after each washing step, representing the virus counts for the 2nd, 3rd and 4th extraction, respectively.
Supplementary Figure S3. Virus and cell numbers in effluents of sediment columns with Janssand sediment in June. Tidal cycle simulation experiments were started after 5.5 days with three reactors, where the water flow in the reactors was stopped for 6 hours followed by 6 hours with continuous water flow. Tidal cycles were run six times in total and only three were sampled directly after restarting of the water flow for analyses of virus and cell numbers in the effluents. The water flow was continuous in one sediment column (control). Figure S4. Virus and cell numbers in effluents of sediment columns with beach sediment in July. Tidal cycle simulation experiments were started after 6 days with three reactors, where the water flow in the reactors was stopped for 6 hours followed by 6 hours with continuous water flow. Tidal cycles were run four times and only two were sampled directly after restarting of the water flow for analyses of virus and cell numbers in the effluents. The water flow was continuous in one sediment column (control).

Supplementary
Supplementary Figure S5. Virus and cell numbers in effluents of sediment columns with beach sediment in November. Tidal cycle simulation experiments were started after 6 days with three reactors, where the water flow in the reactors was stopped for 6 hours followed by 6 hours with continuous water flow. Tidal cycles were run six times and only three were sampled directly after restarting of the water flow for analyses of virus and cell numbers in the effluents. The water flow was continuous in one sediment column (control).
Supplementary Figure S6. Virus and cell numbers in effluents of sediment columns of different lengths: short=A, medium=B, long=C, longer=D with Janssand sediment in August. Tidal cycle simulation experiments were started after 7 days with all four reactors, where the water flow in the reactors was stopped for 6 hours followed by 6 hours with continuous water flow. Tidal cycles were run two times and only one was sampled directly after restarting of the water flow for analyses of virus and cell numbers in the effluents.