The oceans manage to absorb about half the greenhouse-gas CO₂ produced by humans, but how long this state of affairs will last depends on many unknowns. The role of phytoplankton — tiny marine plants that absorb atmospheric CO₂ and form the first link in most ocean food chains — poses some of the most intriguing mysteries.

Wherever it grows, phytoplankton absorbs carbon dioxide from the atmosphere.

One is the long-standing paradox of phytoplankton growth in the subpolar regions, which remains roughly constant all year round — despite the fact that plants don't usually grow well in the dark, and in winter the polar regions are very dark indeed.

In 1996 oceanographer Jim Bishop — now at Berkeley Lab's Earth Sciences Division and an adjunct professor of marine geochemistry at UC Berkeley, but in 1996 a professor of ocean sciences at the University of Victoria — came upon a spectacular piece of the polar phytoplankton puzzle.

Data from Papa

Well out in the middle of the Gulf of Alaska, at 50 degrees North latitude, 145 degrees West longitude, is an empty patch of water called Ocean Station Papa. Weather ships posted there from December 1949 through June 1981 established one of the longest time-series of detailed oceanographic data on record, and data gathering has continued at the site even though weather ships have long given way to weather satellites.

Until 1996, few had ever obtained water samples from deep in the ocean during the winter, because stormy conditions make extended sampling efforts difficult. One way Bishop collects samples is with the Multiple Unit Large Volume Filtration System (MULVFS), an array of collectors lowered over the side by cable that samples the water at regular intervals down to depths of a kilometer. MULVFS filters much larger volumes of water than other samplers and captures larger amounts of particles, both plankton and other matter, which provide a fingerprint of the ocean chemistry as it varies with depth.

"I fought tooth and nail to get MULVFS aboard the research vessel that was going to Papa in the winter, betting that we'd have the eight-hour weather window we needed to get the samples," Bishop says. "Fortunately, nature cooperated. Usually you don't know what goes on in the ocean before the lights go on in the spring."

	Jim Bishop (right) files MULVFS filter samples from deep in the sea.

Yet what Bishop found at Papa in February, 1996 was an anomaly that remained unexplained for years.

The North Pacific is one of three regions of the world's oceans (the others being the Eastern Equatorial Pacific and the Southern Ocean) known as high-nutrient, low-chlorophyll (HNLC) regions, where phytoplankton doesn't grow well at any time of year; while rich in other nutrients, the plants lack the small increments of iron needed to stimulate growth.

Even in these regions, however, there are occasional plankton blooms. Traditionally, oceanographers have ascribed these events to an influx of iron carried out to sea by dust storms — and indeed, in 2001 Bishop himself acquired the first direct measurements of a short-lived plankton bloom in the North Pacific after a passing storm deposited iron-rich dust from the Gobi Desert.

A similar bloom was evident in the particles Bishop obtained from Ocean Station Papa in February, 1996. Yet there was no evidence of a dust storm that could have deposited iron there in the weeks before the measurements were made.

The anomalous samples went into the collection Bishop maintains and has kept with him. "We have a unique sample archive," he says. "No one else collects such large-volume samples."

A new way to investigate

The 1996 samples stayed in the archive until 2001, when Phoebe Lam, a graduate student at UC Berkeley, arrived at Berkeley Lab to work on her Ph.D. "Jim had an observation he couldn't explain, and he asked me to work on it," says Lam.

Phoebe Lam with MULVS, the multiple-unit large-volume filtration system.

Bishop says, "I've generally turned to microscopy to look at anomalous samples, but Glenn Waychunas had recently responded to a Department of Energy call for research on carbon dynamics with a proposal for synchrotron studies. We had started a collaboration when Phoebe showed up." 

Waychunas, who heads the Earth Sciences Division's Molecular Geochemistry and Nanoscience Group, had a research program on nanoparticles and mineral interfaces using a beamline with x-ray microprobe capability at Argonne Laboratory's Advanced Photon Source near Chicago. He requested a few days of microprobe time, and Lam flew there with some of the odd 1996 samples from Ocean Station Papa. Initial results suggested that useful chemical analyses might be obtained, but it was hard to focus the high-intensity synchrotron beam on the samples without damaging the particles; for the smallest particles, beam position stability was also an issue.  

"Then we discovered that Matthew Marcus at beamline 10.3.2 of our own Advanced Light Source at Berkeley Lab had an ideal set-up for our purposes, with an exceptionally stable beam position, excellent mapping of the target, and a good user interface," Lam says. Combined with additional studies done at ALS beamline 11.0.2, the researchers were able to map the spatial distributions of calcium, titanium, chromium, manganese, and iron in the particles in the water samples.

"The ALS made it possible to see the distribution of iron 'hot spots,' micron-sized regions of concentrated iron" — a micron is a millionth of a meter — "which were embedded within aggregates of biological origin," says Lam. "Their visual impact was a spur to solving the mystery." 

	Synchrotron studies revealed micron-sized "hot spots" of iron compounds associated with biological activity at substantial depths.

The five-year-old samples had faithfully preserved evidence of a vigorous bloom of phytoplankton reaching to a depth of 110 meters, dominated by plankton of the kind seen in polar waters to which iron has deliberately been added. But while the iron hot spots occurred throughout the water column, from the surface down to 900 meters deep, they were not the dominant fraction of the total iron. Moreover, spectroscopy done at beamline 10.3.2 showed that the hot spots were rich in iron hydroxides, particulate forms of iron that may not be readily available to phytoplankton.

Yet the phytoplankton bloom itself was strong evidence that there'd been a recent source of available iron in the region. Could the iron hot spots be a clue to where this bioavailable iron was coming from?  

Modeling the flow of iron

"There are three sources of iron in the open ocean," Lam says, "dust from the atmosphere, iron carried from the continental margins, and upwelling from below. But there was no evidence of dust storms in Asia that could have carried iron to Ocean Station Papa in February, 1996. Not only that, the ratio of titanium to iron in the compounds wasn't typical of Asian dust. And if the iron came from upwelling, the concentration of hot spots should have been greater as you looked deeper, which was not the case. That left only the continental margins."

The plentiful iron that washes down rivers and streams or erodes from the shoreline is largely in the form of insoluble iron silicates and hydroxides. The sediments of the continental shelf are low in oxygen, however, which helps dissolve the particulate iron and makes it more available to stimulate biological activity in coastal waters.  

The nearest continental margins to Ocean Station Papa were the coasts of Western Canada and Alaska, 900 kilometers (560 miles) away, with the Aleutian Islands a little farther. How could soluble iron from such a source make the long trip and still be available to stimulate growth? Soluble iron should have been consumed by plankton long before it reached the open ocean.

Inez Fung is a staff scientist in the Earth Sciences Division and a professor of atmospheric science in UC Berkeley's Department of Earth and Planetary Sciences. For years she has been refining global climate computer models, with particular attention to the carbon cycle.

Riding the pycnocline, continental iron reaches Ocean Station Papa (marked X) from the Aleutian Islands.

"I had previously dealt with iron supply and demand in the upper ocean, looking at issues like storms and upwelling, but not at the continental shelves," says Fung. "There is so much going on in the coastal zones — turbulence and bioactivity — that I didn't know I had a problem until I saw new data from several sources, including the spectacular results of Phoebe's synchrotron studies."

To test the plausibility of Lam and her colleagues' idea that particulate iron was being transported from the continental margin all the way out to Ocean Station Papa, Fung and postdoctoral researcher Cara Henning introduced a "hypothetical inert particulate tracer" to her general-circulation model of the North Pacific — or as Fung puts it, "We painted the continental shelf and watched how the water moved it."

Fung says, "The surprise was, we couldn't get it to Papa."  Although they ran the model for a simulated 15 months, the colored tracer representing particles from the east (the coasts of Canada and Southeast Alaska) got nowhere near Ocean Station Papa. "Lo and behold," she says, "we got it from the Aleutians."

A submarine conveyor belt

In the subarctic Pacific the pycnocline — "a region of high density gradient," Phoebe Lam explains, "which prevents mixing of water above and below it" — averages about 150 meters deep, roughly the depth of the outer continental shelves. "In the North Pacific this density gradient is stable throughout the year," she adds. In effect, the dense water layer provides a surface on which particles and soluble iron can glide out to sea without sinking.

Close to shore, dissolved iron is available to phytoplankton all year round. But in the northern winter, very little light penetrates far below the surface, and soluble iron 150 meters down is far out of reach. Riding the pycnocline, soluble iron can survive the thousand-kilometer trip from the Aleutians to Ocean Station Papa and other regions of the midocean.

In winter the storms hit, roiling the waters and mixing their contents. The deep soluble iron churns to the surface and the phytoplankton blooms, even in February. This extra supply of iron in winter, brought near the surface by storms, makes up for the relative lack of light.

In spring and summer, the combination of fewer storms and solar heating make deep mixing impossible, and the deep source of iron is inaccessible. Thus the production of plankton — although never high in HNLC regions — remains roughly constant all year round. "We think we have uncovered the mechanism that explains this," says Lam.

	Phoebe Lam, forensic biogeochemist

For solving the puzzle and finding the mechanism he calls "a stealthy route" for continental iron, Bishop names Lam a "forensic biogeochemist." So it's ironic, she says, that the first thing she and Waychunas noticed in their synchrotron studies had little to do with plankton blooms.

"The iron hot spots we observed in the samples are mostly not bioavailable iron," Lam says. "They are particles that were passively captured and concentrated by aggregating biological particles but were probably not involved in growth. Still, their appearance in the samples was the clue that tipped us off to an unsuspected source of iron in the middle of the subarctic Pacific."

Additional information

• "Wintertime phytoplankton bloom in the subarctic Pacific supported by continental margin iron," by Phoebe J. Lam, James K. B. Bishop, Cara C. Henning, Matthew A. Marcus, Glenn A. Waychunas, and Inez Y. Fung, appearing in the 1 February 2006 issue of Global Biogeochemical Cycles, is available online to subscribers.
• More about MULVFS and ocean biogeochemistry at Berkeley Lab
• More about dust storms and phytoplankton
• More about Ocean Station Papa
• The analytical work described in this article was supported by DOE's Office of Science, Office of Biological and Environmental Research. Phoebe Lam is now a postdoctoral fellow at Woods Hole Oceanographic Institution.