Most of us think about the ocean as a uniform vast space, maybe with some temperature dropping at depth, light diminishing in the deeper water perhaps. However, this vast aquatic ecosystem is actually an "ocean of gradients." These gradients, in fact, evolve at various scales, from hundreds of kilometers to micrometers. The distances at the two ends of the scale are so far apart that when trying to resolve ocean-relevant processes, it is extremely difficult, if not impossible, to unify them.
For instance, when we measure biogeochemical depth profiles within the water column, we struggle to achieve the resolution at which the actual source of our target compound is released. It's an old problem of course-graining, yet in the ocean, this often limits our understanding of important processes like the carbon export and its link to other climate-relevant biogeochemical cycles.
Much more carbon on our planet is stored in the deep ocean than in the atmosphere, transported there largely through the biological pump in the form of particles that sink. These particles are colloquially termed “marine snow,” as they somewhat reflect their namesake gently descending through the water column. But importantly, these particles harbor a lot of biological activity (mostly microbial), making them hotspots of biology in an otherwise desert-like ocean. This aggregated material, ubiquitous across the global oceans, and the biological activity taking place within it is at the core of how we translate the small scale to the planetary scale in the ocean.
When I joined Prof. Andrew Babbin's lab in the department of Earth, Atmospheric and Planetary Sciences (EAPS) at MIT, we wanted to explore these small-scale processes to finally reconcile the connection between particle-associated microbial denitrification and carbon export. I was drawn to Andrew's lab by his long-standing experience measuring the most relevant ocean biogeochemical cycles across the planet and that he had constructed a lab that can replicate ocean conditions.
In oxygenated water, there should not be any denitrification happening because it is typically considered an anaerobic process, yet from previous work , we found out that within aggregates in well-oxygenated rich media, denitrification can occur a few hundred micrometers from the edge of a particle. Within a few microns from oxygenated bulk, denitrification takes place at full speed! This experiment was pivotal for what we then developed. I created model marine particles by embedding in agarose, diatom algae as sole nutrient source and marine denitrifying bacteria that Dr. Irene Zhang helped me to identify. The particles were fixed in place between two glass slides, and oxygenated real seawater with no detectable dissolved organic carbon (DOC) was slowly flowed around the particles to simulate sinking. Over the course of the experiment, I simulated slow sinking over ~10 days long experiments and resulted in a few terabytes of microscope images. Each particle's biomass evolution was acquired simultaneously while I sampled the seawater at the outlet of our device. We also measured the DOC, nitrate, and nitrite to evaluate how microscale processes occurring inside marine particles impact the bulk water column. Thanks to a collaboration with Mr. Omar Tantawi and Prof. Desiree Plata in MIT’s department of Civil and Environmental Engineering, we were able to measure extremely low concentrations of DOC released by just a handful of particles of 1.5 mm each.
But to really connect the microscale to the metabolite evolution in the water column, we embedded oxygen nanosensors in the particles to measure highly spatially resolved oxygen changes. I developed a custom-made backtracking algorithm that measured simultaneously the biomass increase and oxygen change with micrometer precision across the three-dimensional particles. This approach enabled us to overlay the metabolite evolution and the oxygen consumption at the scale that matters for bacterial life. The most exciting result for us was to observe an exact match between the slowdown in growth of bacterial colonies and the peak of nitrite in the water column. This was truly a eureka moment; finally, we could pinpoint the temporal origin of nitrite, yet something was still missing. We knew that it was possible to also identify the spatial origin of nitrite within the particles. So, we embedded a bacterial cell with a fluorescence reporter, and with that, could resolve the exact location within the particles where bacteria underwent a switch from nitrate to nitrite respiration.
To complete the story, we compared two endmembers of carbon export regimes, one in which we explored particles with broken diatoms simulating a more mineralized particle and particles with intact diatoms more akin to the phytoplankton aggregates that form following an algal bloom. The latter scenario within intact diatom cells sustained substantially more microbial growth that shielded the particle from losing as much DOC as its broken counterpart.
While we never touched the real ocean to perform these experiments, we were able to finally peer into the many gradients that exist within the ocean.
 Smriga, S., Ciccarese, D. & Babbin, A.R. Denitrifying bacteria respond to and shape microscale gradients within particulate matrices. Commun Biol 4, 570 (2021). https://doi.org/10.1038/s42003-021-02102-4