The Role of Bacteria-Particle Interactions in Marine Snow Dynamics

Marine snow aggregates are one of the primary vehicles for deep-sea carbon sequestration. Bacterial activity on marine snow affects both degradation and aggregation processes that determine the flux of carbon to depth, biogeochemical cycling, and microbial food webs. The microscale processes occurring on aggregates depend on specific interactions between bacteria and particles. In this thesis, I describe two such interactions which have larger scale implications on marine snow dynamics: (1) the effects of starvation on bacterial motility and colonization behaviors on marine aggregates and (2) interactions between bacteria and phytoplankton which could contribute to the production of TEP. Current models describing bacterial colonization on particles do not account for changes in bacterial behavior due to starvation, which may happen between successful encounters with particles and is a possible condition in the oligotrophic ocean. In my first study, I examine the effects of starvation on the colonization and detachment of several bacterial isolates on particles. I also describe the changes in bacterial motility resulting from starvation. Laboratory experiments on model aggregates indicate that responses to starvation are strain-specific, and can result in lower short-term steady-state bacterial abundances on the aggregates. Bacterial detachment from aggregates was unchanged. Motility data indicate that two of three strains tested had reduced swimming velocities, resulting in diffusivities six times lower in starved treatments than in fed treatments. This was corroborated by colonization data. Future models describing bacterial colonization should consider the shifting physiology and behavior of bacteria responding to starvation. In my second study, I investigated the interactions between bacterial isolates and the marine diatom Thalassiosira weissflogii (TW) on the production and characteristics of TEP, a major component of marine snow. One of two bacterial isolates (either Microscilla furvescens or Curacaobacter baltica) was added to jars of TW and incubated on a rolling table for seven days. During the time course, each jar was sampled for TEP length, area, total TEP, and bacterial distribution among the free-living and TEPassociated fractions. The two strains of bacteria showed different responses. Jars inoculated with Curacaobacter baltica had a significantly higher fraction of total bacteria that were associated with TEP, although the number of bacteria per unit area of TEP was lower. These results suggest that the strain-specific interactions between bacteria, phytoplankton, and TEP could impact the population distributions of bacteria. Over seven days, jars inoculated with Curacaobacter baltica produced more TEP; TEP coverage was almost four times higher (~8% of the total filter area) in jars inoculated with Curacaobacter baltica than those inoculated with Microscilla furvescens. Results from both studies stress the importance of strain-specific interactions in describing microscale processes. Integrating our understanding of responses of individual strains with information on the diversity and activities of bacterial communities on aggregates will better determine how these complex interactions may affect the fate of sinking aggregates and the solubilization of particles into dissolved organic matter