Frog Study Leaps to Conclusions

This week’s eco-horror claim is that the most commonly used herbicide in North America supposedly deforms the sex organs of frogs.

"Male frogs exposed to very low doses of a common weed killer can develop multiple sex organs, sometimes male and female, researchers in California have discovered," the Associated Press reported this week.

A University of California team led by Dr. Tyrone Hayes reported in the April 16 Proceedings of the National Academy of Sciences that concentrations of the herbicide atrazine as small as 0.1 part per billion caused the deformed sex organs.

But let’s hold off on worrying about kissing a frog and getting a hermaphrodite instead of a prince, and focus for a minute on the scientific procedures and standards that determine whether research has led to valid scientific discovery or has simply produced more junk science.

The hallmark of the time-honored scientific method is the independent replication of experimental results. Hayes’ study is the first to report such findings and has yet to be replicated. Moreover, the write-up of his study is woefully inadequate in terms of providing useful information, such as statistical analysis and data.

To understand the importance of independent replication, consider the case of a 1996 study published in the journal Science by Tulane researchers. The researches reported that certain combinations of chemicals in the environment were potent disrupters of hormonal processes. Tulane’s one-study-wonder - hailed at the time by the EPA as "persuasive" and "clean-looking" - was instrumental in pressuring Congress into passing a law requiring the EPA to test chemicals for their ability to disrupt hormonal systems.

But six months after the law was enacted, independent labs from around the world began to report they could not replicate the Tulane results. A year after the law was enacted, the study was formally withdrawn from publication.

Last fall, the federal Office of Research Integrity determined that the lead Tulane researcher committed scientific misconduct by intentionally falsifying his study’s results and then trying to cover up his misconduct.

Meanwhile, we have a federal law resulting directly from what has been determined to be scientific misconduct.

Now for the Hayes study.

As the researchers increased frog larvae’s exposure to atrazine, supposedly up to 20 percent of the frogs had either multiple sex organs or had both male and female organs.

The why-should-anyone-care component of this claim is that atrazine is often a detected contaminant in water supplies, supposedly sometimes reaching levels as high as 21 ppb in groundwater and 42 ppb in surface water.

However, atrazine has been used in U.S. agricultural production for over 40 years. Annual use reaches 60 million pounds. Despite this substantial use, there has been no prior report of a corresponding increase in hermaphroditic frogs due to levels of atrazine typically found in the environment.

And it’s not like no one is looking for frog problems. Scattered reports of frog deformities - usually involving hind leg problems - have focused a great deal of attention on frogs since the mid-1990s.

Further, in a 1998 study, University of Illinois researchers collected frogs from several different sites in Illinois to assess the effects of environmental contamination on the prevalence of frog hermaphroditism. Of 341 frogs collected in 1993, 1994, and 1995, 2.7 percent were hermaphroditic. But there was no statistically significant relationship between the chemical compounds detected - including atrazine - and frog hermaphroditism.

So there’s no strong basis for assuming that whatever happened in Hayes’ laboratory is happening to any significant extent, if at all, in the real world. I have a feeling that the grim report is more akin to a Brothers Grimm fairy tale than science.

In August, 1997, the manufacturer of atrazine, Syngenta Crop Protection, convened a multi-disciplinary panel of scientists to study the potential effects of atrazine on fish, reptiles and amphibians. Hayes was asked to join the panel.

Hayes contributed a laboratory study reporting a possible association between low levels of atrazine exposure and frog development problems. But the panel of scientists could not validate Hayes’ data and recommended additional studies. Hayes subsequently left the panel. A follow-up study could not replicate Hayes’ results.

At this point, we don’t know whether Hayes’ results can be replicated or not. Hayes claims to have conducted a sophisticated statistical analysis of his data, but neither his analysis nor his data are presented in his study. Certainly, his prior conduct casts suspicion over his claims.

We are simply to take his word for it.

No thanks. I’d rather kiss a frog.

Steven Milloy is the publisher of , an adjunct scholar at the Cato Institute and the author of Junk Science Judo: Self-defense Against Health Scares and Scams (Cato Institute, 2001).

Respond to the Writer