One of the nation's most widely-used herbicides has been found to exceed federal safety limits in drinking water in four states, but water customers have not been told and the Environmental Protection Agency has not published the results.

One of the nation's most widely-used herbicides has been found to exceed federal safety limits in drinking water in four states, but water customers have not been told and the Environmental Protection Agency has not published the results. (AP photo) Records that tracked the amount of the weed-killer atrazine in about 150 watersheds from 2003 through 2008 were obtained by the Huffington Post Investigative Fund under the Freedom of Information Act. An analysis found that yearly average levels of atrazine in drinking water violated the federal standard at least ten times in communities in Illinois, Indiana, Ohio, and Kansas, all states where farmers rely heavily on the herbicide.

In addition, more than 40 water systems in those states showed spikes in atrazine levels that normally would have triggered automatic notification of customers. In none of those cases were residents alerted.

In interviews, EPA officials did not dispute the data but said they do not consider atrazine a health hazard and said they did not believe the agency or state authorities had failed to properly inform the public. "We have concluded that atrazine does not cause adverse effects to humans or the environment," said Steve Bradbury, deputy office director of the EPA's Office of Pesticide Programs.

Officials at Syngenta, the Swiss company that manufactures atrazine, declined requests for interviews about the testing results. In a statement on its Web site, the company says that atrazine "poses no threat to the safety of our drinking water supplies. In 2008, none of the 122 Community Water Systems monitored in 10 states exceeded the federal standards set for atrazine in drinking water or raw water."

Atrazine has become an issue of concern for environmentalists and consumer groups as the use of the herbicide has soared in the United States over the past few decades. Some scientists who have studied atrazine said the information about its higher levels in drinking water should be made public.

"This is an issue of the EPA not being forthright about what they know," said Robert Denver, a neuroendocrinologist at the University of Michigan who has served on two of the EPA's scientific advisory panels on atrazine.

"It is the responsibility of the EPA and Syngenta to inform the public of accurate levels of atrazine in their drinking water," said Jason Rohr, a specialist in ecotoxicology at the University of South Florida who studies the effects of atrazine in animals, and who served on the EPA's atrazine panel this past spring.

Atrazine is sprayed on cornfields and other major crops during the summer months and can run off into rivers and streams that supply drinking water. It is also commonly used on golf courses.

Studies of atrazine's potential links to prostate and breast cancer have been inconclusive. Based on the recommendations of its scientific advisory panels in 2000 and 2003, the EPA has listed atrazine as "not likely" to be a carcinogen but does officially consider it to be a potential hormone disruptor - a risk factor explored by researchers testing animals.

In recent years atrazine has been the subject of intensive debate among scientists about its effects on the reproductive systems of frogs and other vertebrate animals. In some studies, male frogs that were exposed to high levels of atrazine have been documented to grow eggs.

In 2004, the European Union banned atrazine because it was consistently showing up in drinking water and health officials, aware of ongoing studies, said they could not find sufficient evidence the chemical was safe.

State regulators in the U.S. test their local water systems for atrazine a maximum of four times a year, under the federal Safe Drinking Water Act. In 2003, the EPA again approved atrazine for use in the United States but it made some demands of Syngenta for the re-registration.

The EPA and Syngenta negotiated a deal for more extensive monitoring of about 150 vulnerable watersheds. Under that arrangement, the company pays for weekly monitoring and sends the results to the EPA, as well as to the local water companies and most state regulators.

The Natural Resources Defense Council, an advocacy organization, is expected to release a report on Monday that fully analyzes a smaller set of Syngenta's weekly testing results -- from 2003 through 2006 -- and reaches conclusions similar to the Investigative Fund's analysis of all five years of data. The group supplied an advance copy of its report to The New York Times, which today published an article about the tests and other safety questions about atrazine.

Misleading Water Bills

The EPA plans to revisit its rules for atrazine in 2011. Presently the agency requires water systems to notify their customers if the quarterly state tests average higher than 3 parts per billion (ppb) annually. According to the EPA data obtained by the Investigative Fund, cities in four states -- Illinois, Indiana, Ohio, and Kansas -- had yearly averages of atrazine violating that standard from 2003 to 2008.

In addition, more than 40 water systems in those states showed spikes of atrazine over 12 ppb - which if found in the state quarterly tests would have required the water system to notify the public within 30 days.

In none of those cases were residents notified of the high levels. In fact, the brochures in their water bills - reviewed for this report -- contained misleading numbers based on the state testing.

For example, based on the quarterly tests, residents of Mt. Olive, Ill., were told that the highest level of atrazine in their drinking water last year was 2 ppb. However, the EPA data shows a spike in June of 16.47 ppb. The same year, residents of McClure, Ohio, were told that the highest level of atrazine in their drinking water was 3.4 ppb. The EPA data shows a spike in June 2008 of more than ten times that amount -- 33.83 ppb.

Both of these cities' water utilities received the weekly EPA data directly from Syngenta, but did not report it. Legally, they didn't have to. The drinking water act only requires cities to report data collected by the state. State tests are performed infrequently, so they are vulnerable to missing the chemical spikes that consistently occur around the time the weed-killer is being applied. With weekly tests, such as those ordered by the EPA, it is all but impossible to miss these spikes.

Asked why the results of the weekly tests had not been published, the EPA's Bradbury said "no data is withheld from the public." Bradbury said the information has been posted on the agency's electronic public docket. In fact, the weekly test results are one of the only items on the docket that are not posted on the site.

Instead they are listed as available only through the Freedom of Information Act.

In an on-camera interview with the Investigative Fund in June, Bradbury also said that the weekly monitoring had found no spikes in any watershed over 3 ppb. "It's these spikes that we're focusing on," he said. "There have been no exceedances." In fact, the EPA's data recorded more than 130 spikes over 3 ppb during 2008 alone -- not only in Illinois, Ohio, Indiana and Kansas, but also in Missouri, Louisiana, and Texas. Bradbury declined to elaborate on the apparent contradiction.

The EPA does not consider one-time spikes of atrazine to be dangerous, but several peer-reviewed scientific studies suggest that the chemical may be harmful, particularly to developing fetuses, in doses as low as 0.1 ppb. One study, published this year in the medical journal Acta Paediatrica, found that birth defect rates in the United States were highest for women who conceived during months when atrazine levels were spiking.

"If you happen to become pregnant in June, you care about the levels [of atrazine] in June, not in January," said Shanna Swan, an epidemiologist at the University of Rochester who has studied atrazine's effect on semen quality and development.

"For pregnant women, you have a critical period of a couple of weeks to a couple of months," Swan said. "If you have a peak exposure in that period, that's what's relevant to the pregnancy."

"The annual average might be relevant for [measuring the risk of] cancer, but it's obviously not okay if they [the EPA] care about regulating for reproductive toxicity," she said.

Had the EPA, the state or the local water companies made the weekly testing results public, residents could have made different choices about their water consumption, such as using inexpensive household carbon water filters or bottled water.

Asked about the discrepancies between the state and weekly EPA data, an EPA spokeswoman, Deb Berlin, said in an e-mail, "Consumers need accurate information to make health decisions for themselves and their families. EPA and state authorities would be interested in knowing about any situation where a public water system is not reporting accurate information to their customers as required by the Safe Drinking Water Act."

'I'd Do More Testing'

Under the terms of its 2003 agreement with the EPA, Syngenta for the past five years has been monitoring water weekly in 10 states, with special emphasis on Illinois, Ohio, and Kansas.

This is how the EPA's testing program generally works: Syngenta sends boxes containing two tubes to about 150 water utilities. During the summer growing season when atrazine levels are likely to spike, water operators at these utilities take samples on a weekly basis. Every week, they fill one test tube with river water and one test tube with drinking water. They ship these samples to Syngenta labs, where the company analyzes them. Syngenta then reports the data to the EPA, as well as to the water utilities themselves and the state regulators.

Testing at the state level is much more modest. Up to four times a year, but as infrequently as once a year, water utilities ship one test tube filled with drinking water to their state regulator. The state analyzes the water and reports the data back to the water utility. This limited data is reported to the public, as required by federal right-to-know laws.

There are vast discrepancies between the two data sets. The Huffington Post Investigative Fund contacted water plant operators to see if they had noticed.

Some local water officials said they provided weekly samples to Syngenta but did not realize the company was acting under a requirement from the EPA intended to supply more data as a safeguard for their drinking water. They indicated they paid little attention to the results of the tests.

Robert Leonhardt, the water plant manager in Mt. Olive, Ill., received the weekly EPA data but said he was not aware of any of the spikes during the last five years, including a high reading of 16.47 ppb. He said the weekly testing was not a central part of his work. "This is a side thing," he said.

Steve Kubler, the water plant manager in Chanute, Kan., initially said of the state and weekly tests: "The numbers match up pretty well. I've never noticed a discrepancy." He added, "If I did, I'd do more testing."

According to that data, his town of Chanute recorded one reading of 6.51 ppb last year. The city reported a high of 1.4 ppb to the public. Asked about the numbers, Kubler said, "Look, what I do with Syngenta -- it's in excess of what I have to do. I don't know even know why they're testing."

In Illinois, Roger Selburg of the state's Environmental Protection Agency said that he looks at the weekly data. But he said he does not use it to determine violations, nor does he report any of it to the public, because he does not know if the data are reliable or accurate. "We are only required to report the state data," he said.

Other water officials expressed some surprise and dismay about the levels of atrazine that showed up in the weekly tests. Osawatomie, Kan., showed a spike of 8.70 ppb in May 2008, although the city reported to the public a high of 0.89 ppb for the year. "That's a pretty good spike," said Marty Springer, water plant manager at Osawatomie's plant. "And no one knows about it."

McClure, Ohio, showed a spike of 33.83 ppb in June 2008, but the town told its residents the highest level that year was 3.4 ppb. "If we had been using Syngenta's data, obviously we would have hit the maximum contaminant level," said Christopher Diem, superintendent at McClure's water utility.

In Baxter Springs, Kan., atrazine spiked above 11 ppb in May 2008 while the town told its residents the highest level during the year was 1.3 pbb.

"We may have passed the quarterly tests for the state, but we're not passing them weekly or daily," said Stan Schafer, a water plant operator in Baxter Springs. "Somebody's got to do something," he said. "I live here. I drink the water. My parents drink the water. My kids drink the water. I just try to keep it clean."

Schafer said he regularly receives atrazine testing data from Syngenta, along with the results from the state, but he doesn't think he is allowed to report it to the public.

That fits with the impression that Kansas state health officials gave Lloyd Littrell, director of utilities in Beloit, about the weekly test results from Syngenta.

"I kept track of those numbers for a couple of years, but I stopped," Littrell said. "The state of Kansas would not let us report the results. We had several conversations about it. They said it wasn't certified by the state or something. I stopped trying. If we can't use it, what's the point of me looking at it?"

According to the EPA data, atrazine spiked above 20 ppb in May 2008, but Beloit reported a high of 2 ppb to the public.

"It concerns me," Littrell said. "If it's an actual health hazard and they know and the EPA knows it's getting in water -- I can't believe they're not doing anything about it."