Award: OCE-1851078

Hot hydrothermal vent fluids released to the seafloor are enriched in metals, notably iron (Fe) and manganese (Mn), which are ~1 million-fold higher in concentration than in the deep ocean. In the last ~30 years, studies of Fe and Mn from hydrothermal vents concluded that all of the iron released from hydrothermal vents precipitated as solids that settled out as metal rich deposits on the seafloor around hydrothermal vent sites. In our study prior to this one, however, we discovered that, at least in the South Pacific Ocean, a significant proportion of hydrothermal Fe could be detected thousands of kilometers from their hydrothermal vent source. This Fe traveled so far that it was predicted, upon its upwelling to the surface, to play an important role in photosynthesis-driven biological activity in the Southern Ocean where Fe otherwise limits primary productivity.  What our prior work could not explain, however, is what was special about this Fe that allowed it to travel so far, while the rest of hydrothermal Fe was deposited out to the sediments. We hypothesized that these special Fe transformations occur on the mesoscale – the 1-100km distance – of the hydrothermal plume and are responsible for setting the amount of Fe that can stably persist to the open ocean.

In this project, we aimed to study hydrothermal Fe and Mn transformations on the mesoscale in order to learn what sets the hydrothermal flux of Fe and Mn to the deep sea. The mesoscale is a particularly difficult scale on which to study hydrothermal plumes because this length-scale is poorly predicted and subject to challenging transport pathways. To solve this, we combined a novel set of approaches in a relatively high-risk sampling scenario. First, we chose a hydrothermal system – Endeavor Segment on the Juan de Fuca Ridge (Northeast Pacific Ocean) equipped by Canada with a deep-sea observing system to measure its deep-sea current velocities, temperatures, etc. We fed those data into a state-of-the-art model (like a “hurricane path prediction” but for a deep sea hydrothermal plume) and used that to predict where the plume should go. At sea, we then used a free-swimming robot called Sentry to follow the plume for 10-20km downstream; we then fed the real-time Sentry data back into the model in order to improve our plume model to match observations. This allowed us to predict exactly where to put our water sampling device on tidal (~hours) timescales, improving our ability to sample the exact locations of plume transport downstream as the plume evolved along non-linear trajectories. In our prior study, all we knew was that whatever processes regulated the escape of Fe and Mn from venting to the ocean happened somewhere on the mesoscale. With our successful novel sampling strategy, we were able to fill in missing data points at 0.25km, 0.5km, 1km, 2km, 5km, 10km and 20km from the source, along its exact circuitous transit path. 

What we found was surprising. The pattern that the plume followed through the ocean changed daily, influenced by large-scale ocean weather patterns and the topography of the seafloor. By the end of our project the plume had followed a clockwise-turning spiral pattern with a ~10km diameter.  Within that spiraling plume we found that, following an immediate 10% decrease within the first 1km or less, there was no further dilution of vent-supplied material over the next 10-20km downstream. Dissolved Mn showed the same behavior: high concentrations of vent-sourced Mn were found directly above the vents, and those same high concentrations persisted as far as we sampled. Essentially all Mn was present in the most soluble component throughout, which shows that Mn was not precipitating to the particle phase that would settle to the sediments. For Fe, the behavior was very different. The juvenile Fe born from the hydrothermal vents was in a +2 oxidation state that was stable for 1-2 days in the smallest soluble size fraction. Upon mixing with seawater containing oxygen, it was transformed over these 1-2 days to the +3 oxidation state and formed particles that were in the tiny colloidal (nanoparticle) (<0.2µm) size fraction. As it reached 5-10 km downstream, these nanoparticles aggregated into large particles that accumulated downplume. Only because we had excellent observations of the spatial distribution of this plume, using our cutting-edge sampling approach, were we able to capture the ideal samples to analyze the kinetics of these metal transformations. We observed that Fe2+ is lost with pseudo first-order scavenging kinetics, meaning that Fe2+ is chemically lost by reacting with things other than itself. In contrast, the total dissolved Fe including nanoparticulate Fe as well, is lost by a combination of first- and second-order scavenging kinetics, meaning that self-aggregation (of presumably colloidal Fe with itself) is also responsible for dissolved Fe loss from the plume, with a half-life of 5-10 days.

 

Last Modified: 05/11/2026
Modified by: Jessica N Fitzsimmons