Most everyone has heard by now that we should limit our consumption of certain fish because they accumulate high levels of toxic mercury. But nobody—not even scientists—knows how that toxic mercury gets into the ocean in the first place.
Here’s the mystery: Most of the mercury that enters the ocean from sources on land or air is just the element mercury, a form that poses little danger because living things can get rid of it quickly. The kind of mercury that accumulates to toxic levels in fish is called monomethylmercury, or simply methylmercury, because it has a methyl group, CH3, attached to the mercury atom.
The problem is that we don’t know where methylmercury comes from. Not nearly enough of it enters the ocean to account for the amounts we find in fish. Somewhere, somehow, something in the ocean itself is converting relatively harmless mercury into the much more dangerous methylated form. (See interactive of the mercury cycle.)
That’s the puzzle Carl Lamborg, a biogeochemist at Woods Hole Oceanographic Institution (WHOI), is trying to solve. Lamborg got hooked on mercury as a master’s degree student at the University of Michigan and then pursued his Ph.D. at the University of Connecticut with Bill Fitzgerald, one of the foremost experts on mercury in the ocean. Fitzgerald, who was the third student to graduate from the MIT/WHOI Joint Program and the first in chemical oceanography, devoted his career to mercury after seeing photographs in the 1970s of people poisoned by methylmercury dumped from a chemical plant into Minamata Bay, Japan. In one famous picture, originally published in Life magazine, a woman cradles her teenage daughter, who had been deformed by prenatal exposure to methylmercury. (The photographer, W. Eugene Smith, later withdrew this and other searing photos from public display at the request of the subjects and their families.)
Minamata Bay was one of the worst cases ever of methylmercury poisoning, but sadly, it was not unique.
“There was a lot of mercury dumped back in the day when folks were not sensitive to what was going on,” said Lamborg. “The buzzword that people use for that is ‘legacy mercury.’ Coastal sediments tend to be really elevated in mercury that was dumped there 30, 40, 50, 100 years ago as a result of some industry. And that might still be in play, because there’s worms and shellfish and things living in the mud, and they’re always sort of stirring it up.”
At Minamata Bay, the source of the methylmercury was clear. We also know the source of most of the elemental mercury in the ocean. Some comes from natural sources such as volcanic eruptions. About two-thirds comes from human activities. The biggest single source is the burning of fossil fuels, especially coal, which releases 160 tons of mercury a year into the air in the United States alone. From there, rainfall washes the mercury into the ocean.
We also discharge mercury-laden industrial effluents directly into rivers or the ocean. This is not just a scourge of modern life; Lamborg said a mercury mine in Slovenia has been dumping its wastewater into the Gulf of Trieste since Roman times.
But even large discharges such as that wouldn’t pose a major threat to human health if the mercury were not converted to methylmercury, which diffuses into phytoplankton and then passes up the food chain in ever-accumulating quantities. Large predator fish such as tuna, for example, contain about 10 million times as much methylmercury as the water surrounding them.
“Something like a shellfish, which is a filter feeder, that’s very close to the bottom of the food chain, is generally not as high in methylmercury as something like a tuna or a mackerel or swordfish or striped bass—all the fish, actually, that we really like to eat,” Lamborg said.
So where and how does the conversion of mercury to methylmercury take place? Lamborg said the process is probably biotic—done by living things. Beyond that, our knowledge is sketchy. We know that fish don’t methylate mercury, and phytoplankton and zooplankton probably don’t either.
However, some species of bacteria do produce methylmercury, as a byproduct of their respiration. This has been observed in bacteria living in seafloor sediments along coasts and on continental shelves. It might also occur in deep-ocean sediments, but no one has looked there yet.
A few centimeters down into the sediment, there’s so little oxygen that microbes living there must use anaerobic respiration. One common means is a chemical reaction called sulfate reduction, in which they use sulfate (SO42-) in surrounding seawater for respiration and excrete sulfide (S2-) into the water as a waste product. If seawater in porous spaces within the sediment also contains a lot of mercury, the stage is set for the production of methylmercury.
That’s because sulfide helps mercury get into cells. Most forms of mercury can’t pass through a cell membrane because they are bound to large molecules or because they carry a charge. But when positively charged mercury ions (Hg+2), the most common form of mercury in the ocean, meets negatively charged sulfide, the two bond. The resulting compound, HgS, is small and uncharged—just right to be able to pass into microbial cells.
Once inside, the mercury gets methylated. Scientists haven’t yet discovered the chemical reactions involved in this conversion, but soon after HgS enters bacterial cells, the cells release methylmercury. Some of the methylmercury diffuses out of the sediments into the open water. There, it is taken up by phytoplankton to begin its journey up the food chain.
But how much of the methylmercury made by bacteria in sediments finds its way into the water above? Is that the only source of the methylmercury that turns fish toxic?
Lamborg is skeptical of that idea. He thinks there has to be another source of methylmercury adding to the oceanic total.
“What I’ve been chewing on is the possibility that a lot of methylmercury is actually coming from within the water itself,” he said.
Lamborg has found that there’s a layer of water in the ocean, between 100 and 400 meters thick, that contains high levels of methylmercury. It occurs at midwater depths—from 100 to 1,000 meters below the surface, depending on different locations in the ocean. He’s seen the high methylmercury layer in the relatively isolated Black Sea, the open ocean near the western coast of Africa, and the waters near Bermuda. What’s especially intriguing is that peak levels of methylmercury occur at depths where the amount of oxygen in the water drops sharply.
“This drop in oxygen is caused by all the plankton that are growing closer to the surface,” he said. “When they die, or when they’re eaten by other plankton, those dead cells or the poops of the other plankton sink down and rot. That rotting consumes oxygen.”
It’s possible that, like bacteria in sediments, any bacteria living in low-oxygen areas of the ocean also rely on sulfate for respiration and could be generating methylmercury in the midwater low-oxygen zone.
Lamborg is pursuing that hypothesis, but first he tested another possibility: whether methylmercury in the low-oxygen zone came from higher up in the water. Scientists studying phytoplankton have found that 20 to 40 percent of the mercury inside them is methylated. Lamborg wondered: As the phytoplankton or zooplankton that eat them die, sink, and get degraded, does any of that methylmercury get released back into the water and accumulate in midwater depths?
To find out, Lamborg collected tiny particles that were sinking through the water and tested them for the presence of mercury and methylmercury. He caught the particles in sediment traps—polycarbonate tubes about 3 inches across and 2 feet long, that were suspended from a cable at 60 meters, 150 meters, and 500 meters below the surface.
Before deploying the traps, Lamborg filled each one with particle-free seawater. Then he added extra-salty brine that was so dense that it formed a distinct layer at the bottom of the tube, which traps the particles.
He left the traps in place for four days, then hauled them up and ran the brine through flat, round filters a bit bigger across than a quarter. There’s no doubt when a trap is successful at gathering material, said Lamborg; the fine brown residue left on the filters has an air of rotting fish. “They smell pretty bad,” he said. “It’s not like poop, but it’s definitely ‘eww!’ ”
Lamborg collected sinking particles at several locations during a research cruise across the Atlantic from Brazil to the coast of Namibia in 2007, and brought them back to his lab at WHOI for analysis.
To find out how much methylmercury fell into a trap, Lamborg converted all the mercury on the filter to elemental mercury. He then passed the sample over grains of sand that had been coated with gold. Only the mercury sticks to the gold; other chemicals don’t. Then Lamborg heated the gold-mercury amalgam to vaporize the mercury.
“This is the same process that people doing gold mining used to use,” Lamborg said. “You know panning for gold? You would squeeze some mercury in your pan and sluice it around, dump off the sediment, and then you would heat it up and burn off the mercury and leave the gold behind.”
In Lamborg’s version of the process, the gaseous mercury is the valuable product. It gets drawn into wiry Teflon tubes that take it to an atomic fluorescence spectrometer that determines how much mercury was in the sample. On a nearby table, mercury from a parallel sample is run through a gas chromatograph to determine what proportion of it was methylated.
“These are some of the most challenging samples to analyze that I’ve come across, because the samples are very small,” Lamborg said. “There’s very little material. The techniques we’re using can detect methylmercury in the femtomolar range.” One femtomole of methylmercury would be 0.000000000000215 grams per liter of seawater.
The samples contained elemental mercury, but so far, none of the samples from any of the three depths have shown substantial levels of methylmercury. It was present, but at lower levels than are found in phytoplankton—far too little to explain the levels of methylmercury seen in the midwater zone.
If organisms in surface waters are not the source of methylmercury in the midwater layer, where does that methylmercury come from? Lamborg said it could be made by bacteria in sediments on the continental shelf and released into the water. Currents could sweep these methylmercury-rich waters off the shelves and into the open ocean at depths about the same as the midwater layer. Other researchers are exploring that possibility.
Lamborg, though, favors the notion that methylmercury found in midwaters is being made there, just as it is in sediments, by microbes that are reducing sulfate. He recently started working with microbiologist Tracy Mincer, a colleague in the WHOI Department of Marine Chemistry and Geochemistry, to identify the genes that bacteria use to methylate mercury. Their research could identify similar genes to look for in microbes in the low-oxygen midwater zone.
And he’s still interested in those sinking particles and what role they might play. Methylating microbes can’t do their thing unless they have mercury to work with, and Lamborg thinks the particles offer an efficient shuttle service for mercury that enters surface layers of the ocean from the atmosphere, ground water, or rivers.
“Mercury entering the ocean today is reaching that low-oxygen zone somehow,” he said. “These particles are still playing an important role in moving mercury from a part of the ocean where methylation doesn’t occur to a part of the ocean where it does.”
This research was supported by the National Science Foundation and the Andrew W. Mellon Foundation Awards for Innovative Research at WHOI.