Microbes Have a Speed Limit for Eating Sinking Carbon — and It Reshapes How We Model the Ocean's Carbon Pump

Shipboard single-cell tracking reveals a chemotactic horizon around marine snow, producing a non-monotonic encounter peak at 14.4 m/day that splits particle fate into overdamped vs. ballistic regimes.

Imagine you're a fly trying to land on a moving car. At low speed, you track the car easily — you smell the exhaust, adjust your flight, and land. At highway speed, the wind shear rips you away before you can orient. There's a sweet spot in between where you intercept most efficiently. This paper says ocean microbes face exactly this problem with sinking particles of marine snow, and the sweet spot dictates how much carbon reaches the deep ocean. Chajwa and Prakash, working from Manu Prakash's lab at Stanford, conducted something genuinely unusual: high-resolution single-cell tracking aboard a research vessel using freshly collected marine snow under controlled flow conditions spanning realistic sinking speeds. This is not a simulation and not a lab surrogate — it's direct measurement of wild microbial behavior around real aggregates in imposed flows. The core finding is a "chemotactic horizon," a spatial boundary around a sinking aggregate within which a microbe's chemical sensing can overcome the rotational shear from the particle's wake. Inside that horizon, microbes reorient toward the aggregate. Outside it, they behave like passive tracers swept along by flow. The encounter rate between microbes and sinking particles is non-monotonic — it peaks at a sinking speed of 14.4 m/day, then drops sharply beyond 78.1 m/day. Below the peak, microbes easily track and colonize aggregates, remineralizing carbon in the upper ocean. Above it, particles sink too fast for chemotaxis to matter, and carbon punches through to the deep. This is the critical insight: the biological carbon pump doesn't just depend on how much stuff sinks, but on the speed distribution of sinking particles relative to this chemotactic threshold. The authors also observe rapid switching between motile and non-motile states, adding stochasticity to the encounter dynamics. The theoretical framework is where the paper gets ambitious. They embed the measured encounter rates into a kinetic transport equation in position-velocity phase space, treating remineralization as a velocity-dependent nonlinear damping term. This produces a separatrix — a dividing line in phase space — that splits sinking particle populations into two fates: overdamped (slow sinking, heavily remineralized, carbon stays shallow) and ballistic (fast sinking, escapes microbial consumption, carbon reaches depth). The separatrix is threshold-induced, meaning small changes in sinking speed near the critical value produce large changes in carbon fate. The experimental methodology is strong for its class. Shipboard measurements with real marine snow under controlled flow are rare and hard to execute. The single-cell tracking captures diverse motility repertoires across natural microbial communities, not lab-cultured monocultures. The flow speeds tested (spanning ecological sinking rates) are well-motivated. The weakness is that the theoretical model, while elegant, is validated against the same measurements that informed it — there's no independent dataset or field observation of the predicted velocity segregation in actual ocean carbon flux profiles. The implications for climate science are significant if the framework holds. Current biogeochemical models of the biological carbon pump typically treat remineralization with depth-dependent decay functions (Martin curve and variants) that don't account for the velocity-dependent threshold dynamics described here. If sinking speed distributions shift — due to changes in phytoplankton community composition, aggregation dynamics, or mineral ballasting under ocean warming and acidification — the nonlinear threshold means carbon export could change abruptly rather than gradually. The separatrix framework gives modelers a mechanistic target to test against. What's missing is the scaling question. The measurements are from specific ocean regions and specific microbial communities. Whether the chemotactic horizon parameters and the 14.4 m/day peak generalize across ocean basins, seasons, and microbial diversity is unstated. The phase-space model is analytically tractable but its predictions about velocity segregation in real carbon flux profiles haven't been tested against sediment trap data or radionuclide-based flux estimates. That's the obvious next experiment — and it would require collaborators with very different instrumentation.