Mercury: Grain of Truth, Gram of Nonsense

Published March 1, 2004

You have probably heard or read the oft-repeated statement, “One gram of mercury can contaminate an entire 20-acre lake.” It shows up in the environmental advocates’ literature as well as in EPA and state agency documents and various fact sheets on mercury. The statement is meant to scare us into believing that mishandling a thermometer or emitting even one gram of mercury would have irreversible negative consequences. And you won’t be able to eat the fish, either.

The statement is so definitive and exact there must be scientific proof behind it, right? One gram? A 20-acre lake? Government agencies wouldn’t use a statement like that unless they had actually performed the study or found one that was peer-reviewed, would they? Especially since major policy decisions are being based on it.

“One gram of mercury has the ability to contaminate a 20 acre lake,” says a document on EPA’s Region 9 pollution prevention Web site. The document does not cite the source for this claim.

On EPA’s main Web site we find a 54-page “Economic Analysis of Including Mercury Containing Devices in the Universal Waste System, Notice of Proposed Rulemaking, February 15, 2002.” “One gram of mercury,” the document states, “can foul up to 5 million gallons of water.” A different statement … this time, with a footnote. It referenced a November 16, 2000, article in the Boston Globe, a 400-word piece on mercury thermometers. The statement in fact appeared in the article, but again there was no attribution.

A Google search finds legislator Jon Cooper from New York writing in May 2001, “According to the EPA, a typical fever thermometer contains about one gram of mercury, enough to contaminate a 20-acre lake.”

“One gram of mercury,” according to the Toxics Action Center, “is enough to make the fish in a 20-acre lake unsafe to eat for a year.”

After searching 150 documents through Google, it became evident all were variations on the same theme. Everybody quoted each other; no one offered attribution to anything other than press releases or generic public information documents.

They are all serious misrepresentations of the dangers of mercury, and all try to simplify what is in reality a complicated process.

Contamination Calculation

Scientists know contamination is a function of volume, not surface area. A 20-acre lake would have a surface area of 871,600 square feet. Assume an average depth of 30 feet. That would give us 26,148,000 cubic feet–almost 200 million gallons of water. Can one gram of elemental mercury really contaminate 200 million gallons of water?

Turns out I wasn’t the only one skeptical of this claim. In a letter to Junkscience magazine dated February 23, 2001, Dr. Kevin Wallace also cast aspersions on those claims. Wallace is director of the Occupational & Environmental Toxicology Clinic at Good Samaritan Regional Medical Center in Phoenix, Arizona. He led me to a scientist at the Minnesota Pollution Control Agency.

Ed Swain, Ph.D., of the Minnesota Pollution Control Agency, in an effort to determine the rate of atmospheric deposition of mercury in Minnesota, performed a series of core samples on several remote lakes in the state. The article he wrote, “Increasing Rate of Atmospheric Mercury Deposition in Midcontinental North America,” appeared in the August 7, 1992 issue of Science magazine.

Based on the mercury in the core samples, he determined that in Minnesota the rate of atmospheric deposition (from rain, primarily) of mercury to a 20-acre parcel (lakes in this case) is approximately one gram per year. Swain then calculated that based on the size of Minnesota, approximately 6,000 pounds of mercury are deposited in a one-year period throughout Minnesota in order to reach the one-gram-per-20-acre conclusion. That’s the data.

Nowhere does he make the conclusion that one gram of elemental mercury (like that from a thermometer) can contaminate a 20-acre lake. In fact, the study had nothing to do with the effects of mercury on water or fish.

Not only did the advocates misunderstand the study completely, but in their haste to grab headlines, they glossed over a few points in Swain’s report:

  • mercury deposited from natural sources (i.e., volcanic activity, ore leachate) accounts for 25-30 percent of all mercury;
  • mercury deposition is a global problem;
  • it is the soluble form of mercury found in rain that is the problem (not the elemental form found in thermometers). Soluble mercury, through a complex transformation, goes up into the atmosphere and comes down in rain, converting to a form known as methylmercury that is absorbed by fish. Not all mercury converts to methylmercury–only about 5 percent, according to Swain, and even that is subject to some variability. And its effect on fish would be further conjecture, since that would depend on types of fish, reproduction rates, etc. In other words, it’s a guess. Elemental mercury is rarely introduced to water bodies, despite the activists’ thermometer analogy.

Swain’s study has been misinterpreted and misquoted so many times he actually released a clarification of the study in October 2002.

Mercury in Fish

In the May 17, 2003 edition of The Lancet, one of the most respected medical journals in the world, Dr. Gary J. Myers and others from the University of Rochester Medical Center in New York present the findings of their study of mercury exposure in 776 mother-child pairs in the Seychelles islands (in the Indian Ocean) for 14 years.

The level of mercury in the ocean fish in the Seychelles is nearly identical to that in the United States (remember, mercury is a global problem). However, the women there eat fish up to 12 times per week, far more than we do in the U.S. Consequently, concentrations of mercury in indicator hair samples are many times higher among Seychelles women and their children than in the U.S.

Myers’ conclusion (verbatim): “These data do not support the hypothesis that there is a neurodevelopmental risk from prenatal MeHg [methylmercury] exposure resulting solely from ocean fish consumption.”

To my knowledge, EPA has never mentioned the results of this study in any of its literature. The May 17 report is the fifth follow-up Myers has published (others were published when the children were six months, 19 months, 29 months, and 66 months old).

We might not expect environmental advocates to have read and understood the work by Swain or Myers–but scientists at EPA and state agencies should have, and they have an obligation to present a balanced picture.


Robert Rio is vice president of environmental programs at Associated Industries of Massachusetts. His email address is [email protected].