Diatom-associated bacteria are required for aggregation of Thalassiosira weissflogii

Abstract

Aggregation of algae, mainly diatoms, is an important process in marine systems leading to the settling of particulate organic carbon predominantly in the form of marine snow. Exudation products of phytoplankton form transparent exopolymer particles (TEP), which acts as the glue for particle aggregation. Heterotrophic bacteria interacting with phytoplankton may influence TEP formation and phytoplankton aggregation. This bacterial impact has not been explored in detail. We hypothesized that bacteria attaching to Thalassiosira weissflogii might interact in a yet-to-be determined manner, which could impact TEP formation and aggregate abundance. The role of individual T. weissflogii-attaching and free-living new bacterial isolates for TEP production and diatom aggregation was investigated in vitro. T. weissflogii did not aggregate in axenic culture, and striking differences in aggregation dynamics and TEP abundance were observed when diatom cultures were inoculated with either diatom-attaching or free-living bacteria. The data indicated that free-living bacteria might not influence aggregation whereas bacteria attaching to diatom cells may increase aggregate formation. Interestingly, photosynthetically inactivated T. weissflogii cells did not aggregate regardless of the presence of bacteria. Comparison of aggregate formation, TEP production, aggregate sinking velocity and solid hydrated density revealed remarkable differences. Both, photosynthetically active T. weissflogii and specific diatom-attaching bacteria were required for aggregation. It was concluded that interactions between heterotrophic bacteria and diatoms increased aggregate formation and particle sinking and thus may enhance the efficiency of the biological pump.

Figure 1

Scanning electron microscopy of a T. weissflogii cell with four cells of the attaching bacterial strain, HP15. Bacteria attached to T. weissflogii within 24 h of incubation. Scale bar 1 μm.

Figure 2

Total aggregated volumes (cm³) per rolling tank during 168 h of incubation. Photosynthetic active T. weissflogii cultures were separately incubated with the attaching bacteria strains, HP15, H14, HP10 and Ex5, or the non-attaching bacterial strain, HP2, in duplicates. Axenic T. weissflogii culture served as control. Photosynthetic inactive T. weissflogii cultures incubated with attaching strain, HP15, and non-attaching strain, HP2, as controls. Error bars represent biological replicates (duplicates), s.e.

Figure 3

Sinking velocities of aggregates collected after 168 h of incubation as a function of aggregate diameter (ESD in mm). (a) Aggregates formed by T. weissflogii and the attaching bacterial strains, HP15, H14, HP10 and Ex5, as well as the non-attaching bacterial strain, HP2. (b) Pooled sinking velocities for all aggregate types measured as a function of aggregate diameter. A power equation was fitted to the measurements: sinking velocities=44.11 ESD^1.31; R²=0.81; P<0.0001.

Figure 4

Abundance of TEP and bacterial cell numbers versus time during rolling tank experiments. (a) Treatment with HP15 (b) treatment with H14, (c) treatment with Ex5, (d) treatment with HP10, (e) treatment with HP2 and axenic T. weissflogii cultures as controls. Error bars represent technical replicates (triplicates), s.d.

Figure 5

Correlation of total aggregated volumes and TEP concentrations in rolling tank experiments. Linear regression was fitted to measurements N=30; R²=0.89; P<0.0001.