The Errors That Snared Dr. Bruce Ivins

Ed Lake
Originally posted March 14, 2010
(Latest revision: April 28, 2011)

Dr. Bruce Ivins made two critical errors that eventually led to him being identified as the anthrax killer.  The first error involved believing what was commonly believed, the second was a straight-forward mistake of his own.

Error #1: The source of the Ames strain

In many discussions of the Amerithrax investigation it seems very difficult for some people to separate what is known now from what was known in October of 2001.   The answer to a key question about the investigation requires understanding the difference.  Here's the question and the answer:

Question: Why would Dr. Ivins use an anthrax strain that could be traced directly back to his lab?

Answer:  Dr. Ivins used an anthrax strain that he believed could never be traced to his lab.

At the time of the mailings, Dr. Ivins believed what everyone else believed about the Ames strain.  And, as it turned out, everyone was wrong.

In 1985, sixteen years before the anthrax mailings, Dr. Gregory Knudson from USAMRIID co-authored an article with this information about the source of the Ames strain:

TABLE 1. Histories of isolates used this study
Isolates Source and date of isolation
Vollum .........................Cow; ca. 1944
Vollum 1B ...................Derived from Vollum
Ames ............................Cow; Iowa, 1980
Buffalo .........................Buffalo; Iowa, 1979

In 1996, five years before the anthrax mailings, Dr. Ivins was a co-author and the prime contact on a scientific paper titled,
"Efficacy of a standard human anthrax vaccine against Bacillus anthracis aerosol spore challenge in rhesus monkeys."  The article says,

"The virulent Ames strain of B. anthracis was obtained from the U.S. Department of Agriculture, Ames, Iowa."

In 1997, four years before the anthrax mailings, Dr. Ivins contacted with Dugway Proving Grounds to make a very large collection of anthrax of spores.  On October 22, 1997, when filling out the "Reference Material Receipt Record" to keep track of the usage of the spores stored in flask RMR-1029, Dr. Ivins wrote on the form after the word "Vendor:"

"From B. anthracis Ames strain, Ames Iowa"
A day before Bob Stevens died of inhalation anthrax on October 5, 2001, a sample of Bacillus anthracis taken from his body was sent by air to Paul Keim at Northern Arizona State University.   According to the December 16, 2007, issue of The Arizona Republic,

On the afternoon of Oct. 4, 2001, Keim was in his office when the telephone rang. On the other end was an FBI agent, who told him a plane was on its way to Flagstaff from Atlanta with a culture taken from Stevens' spinal fluid. The FBI wanted Keim to analyze the DNA and find out what type of anthrax Stevens had contracted. This could provide possible clues to where the anthrax originated.

And here is what happened when the sample arrived in Arizona:

The agents filled out paperwork, then handed the box to Keim, who placed it in the back of his 4Runner and drove back to his lab. A glass tube nestled in ice held the culture from Stevens' body. Keim and a couple of his key researchers worked through the night, isolating, processing and magnifying the DNA using machines and computers similar to ones found in crime labs. In the early morning, they compared the results with their anthrax database. They found a match: a virulent type called the Ames strain. The U.S. Army developed the lab strain in the 1980s as a test for the anthrax vaccine.

As soon as it was learned that Bob Stevens had been infected by the Ames strain, investigators immediately began to search for the sources of the Ames strain.  It was known that Ft. Detrick used it, but everyone at Ft. Detrick and everywhere else believed that the strain was a common strain that Ft. Detrick had obtained from the USDA in Iowa. 

On October 10, 2001, five days after Bob Stevens died, The Florida Sun-Sentinel reported,

CNN reported Wednesday morning that the anthrax virus that killed a Lantana man and was found in his Boca Raton office appears to be manmade and apparently produced in an American lab about 50 years ago.

The television network reported that the anthrax that was found in a newspaper office in Boca appears to have been made in a lab in Iowa, one of only two in the United States, that made the deadly disease for research purposes.

The report also said the anthrax used in south Palm Beach County was probably manufactured sometime in the 1950s.

On October 12, 2001, Dr. Ivins wrote someone an email in which he described his beliefs about the source of the Ames strain this way:

I can tell you to whom I have sent this so-called "Ames" strain.  Please keep in mind that a) it is apparently 50 years old; b) that USAMRIID received this strain 20 years ago; c) that it is a USDA strain, not a USAMRIID strain, U.S. Army strain, or Department of Defense strain; d) the individuals primarily responsible for determining the location of the strain are located in Ames, Iowa, not in Frederick, Maryland; e) that of any U.S. labs having human pathogenic strains (including B. anthracis), none have higher security than USAMRIID, f) that if we are the only recipients of this "tasker," it is transparently evident that we are being harassed by our regular detractors simply because we are DOD researchersIt is not within the purview of USAMRIID researchers to ascertain where the USDA has sent its strains of Bacillus anthracis or any other organism.

When someone sent Dr. Ivins another email about the Ames strain
on October 18, 2001, Ivins testily responded via an email that is reproduced on page 76 of the FBI/DOJ's summary report of the Amerithrax investigation:

The “Ames” strain of Bacillus anthracis was sent to us in the late 1980-early 1981 time frame from the United States Department of Agriculture, Animal & Plant Health Inspection Services, National Veterinary Services Laboratories, Ames, Iowa. We were told it came from a dead cow. We were not told the specifics of the strain, specifically where it was isolated, or when it was isolated. Basically, we were told it was Bacillus anthracis that had been isolated from a clinical veterinary case.  I’ve read that the strain was originally isolated in the 1950s at Iowa State University, but we were not given that information when we got the strain. I have also read that the strain is very common in veterinary labs, clinical labs, university bacteriology labs and research institutes all over the country, and that doesn't surprise me. From the literature, it seems that many places have the “Ames” strain or its derivatives. The proper place to find out the details of the strain is the USDA, not us. They sent it to us. It’s their strain, and it's their responsibility to know the details about it. Thanks!

Where did Dr. Ivins read  that the strain was isolated in the 1950s at Iowa State University?  Since it is totally untrue, it doesn't seem likely that it was printed anywhere, and there doesn't appear to be any supporting documention which says that the strain ever had anything to do with Iowa State University.  After checking into all the reports I could find from the time, it now appears that the misinformation by Iowa State University was simply a mistake by NBC Nightly News, when some reporter somehow changed Department of Agriculture (where Ivins and everyone at USAMRIID believed the Ames strain came from) to Department of Energy, and the media then descended upon Iowa State University which ran a large lab for the Department of Energy (a lab that evidently had nothing to do with anthrax).

According to the February 1, 2002, issue of The Iowa State Daily which is published by Iowa State University in Ames, Iowa,

The dean of the College of Veterinary Medicine was thrown into a frenzy.

"[October 9, 2001] was a crazy day," said Norman Cheville, dean of the College of Veterinary Medicine office. "In that two-day period, we had over 140 interviews and over nine television interviews."

At first, it was unknown whether Iowa State still had anthrax, said James Roth, distinguished professor of veterinary microbiology and preventative medicine.

"This building has over 1,000 rooms in it and probably a dozen microbiology labs," he said. "It took a bit to find out if we had some [anthrax]."

Eventually, the professor in charge of the lab area where the anthrax was stored affirmed that Iowa State did, indeed, possess anthrax.

But they couldn't find anything about any "Ames stain" anywhere at Iowa State University:

"We didn't know what the Ames Strain was, either," [James Roth, distinguished professor of veterinary microbiology and preventative medicine] said. "We didn't name it the Ames Strain."

When the questions began, the collection was taken out of the drawer where it had been stored.

It was examined in a biological safety cabinet by two members of the ISU environmental health and safety unit. Although some labels were incomplete or cryptic, none of the more than 100 tubes were labeled "Ames Strain."

In a 1985 publication, U.S. Army Medical Research Institute of Infectious Diseases referred to the strain, now known to be from Texas, as the Ames Strain. The report also referred to the strain as being isolated in 1980, another reason Roth had doubts about the Iowa connection.

"It says it came from a cow in 1980. No one here remembers a case of anthrax in 1980, but we do remember a case in 1979," he said.

The Washington Post made public more details about the Ames strain on October 26, 2001:

Microbe Is of Type Commonly Used in Research


Office of Homeland Security Director Tom Ridge said yesterday that the bacterial spores that caused anthrax outbreaks in Florida, New York and Washington belong to the so-called Ames strain -- a subtype of the anthrax bacterium that is commonly used in universities around the world and was a focus of studies by the U.S. military.


That strain was first isolated in Ames, Iowa, and sent in 1980 to Army researchers, who have since distributed it to various academic laboratories.

The strain has spread by other routes to countless research labs around the world, making its identification relatively useless as a tool for tracking the perpetrators, experts have said.

As with so much of the information that was reported in the world media in the early days of the Amerithrax case, this information was totally wrong.  We don't know who told the media that the Ames strain was widely distributed.  The media only talks about what "experts" said.  And, once again, it's totally false information that could very likely have come directly or indirectly from Dr. Bruce Ivins.   It was what he evidently believed to be true.

It wasn't until three months later that the true facts became publicly known.   On January 29, 2002, The Washington Post reported "One Anthrax Answer: Ames Strain Not From Iowa."
  And The New York Times reported on January 30, 2002, "Geographic Gaffe Misguides Anthrax Inquiry."   Here is part of what The Times reported:

Federal investigators have found in recent weeks that the so-called Ames strain was first identified not in Ames, Iowa, its reputed home, but a thousand miles south, in Texas. The strain of the bacteria was found on a dead cow near the Mexican border in 1981, and the geographic gaffe was the result of a clerical error by a scientific researcher.

In late 1980, Dr. Gregory B. Knudson, a biologist working at the Army's biodefense laboratory at Fort Detrick, Md., was searching for new anthrax strains to use in tests of the military's vaccine.  In December 1980, he wrote Texas A&M to see if they had any new anthrax strains.  They didn't have anything at the time, but in early 1981, they received a sample of anthrax that had been extracted from a cow that had recently died.  So, Texas A&M forwarded a portion of that sample to Ft. Detrick

However, because Texas A&M frequently sent such samples to the USDA in Iowa, they had postage-paid labels from the USDA and they used one of the USDA labels, simply pasting the Ft. Detrick address over the USDA address(It was evidently a way of saving a few dollars for Texas A&M.)  And, when the sample arrived at Ft. Detrick, Dr. Knudson called it "The Ames Strain" because the mailing label indicated the sample had come from Ames, Iowa

The photo below was faxed somewhere on December 18, 2001.
Ames mailing label

It was a simple mistake that misled Dr. Knudson, Dr. Ivins and everyone else into thinking that the Ames strain came from Iowa.

Dr. Ivins believed what the news media and everyone else believed: the Ames strain was a common strain used by "countless labs around the world, making its identification relatively useless as a tool for tracking the perpetrators."

That's why Dr. Ivins used the Ames strain in the anthrax letters.  What other strain available to him would have been better?  He believed no other strain was as widely distributed.  Other strains might somehow be tracked back to Dr. Ivins, but the Ames strain wouldn't be tracked back to him because Dr. Ivins believed that his samples were virtually identical to what was stored at the USDA in Iowa and distributed by the USDA to labs all over the world.

So, there can be no doubt as to what Dr. Ivins believed about the source of the Ames strain at the time he used it to create the powders in the anthrax letters.

Error Two: Mutations & The Single Colony Pick

The second error which eventually led to Dr. Ivins' being identified as the anthrax killer was a misunderstanding about biological mutations.   His elementary misunderstanding of mutations caused Dr. Ivins to believe that the contents of flask RMR-1029 were virtually identical to what was stored at the USDA in Ames, Iowa, and which - he believed - had been distributed from there to universities and labs all around the world - thus making it virtually impossible for the attack anthrax to be traced back to him.

Dr. Ivins beliefs about this subject are documented in the 2,700 pages of supplementary information made public on February 19, 2010.   On page 16 and 77 of the 193 page .pdf file identified as "BEI Section 5 (2/2008)" or pdf file 847547, the following information can be found.  It's from a February 18, 2008 FBI interview with Dr. Ivins:

          In a report dated October 18, 2001, IVINS wrote about the concentration and purity of the spores in the Daschle letter.  In that report, he opined the evidence is not 'garage" spores.  When asked what he meant by garage spores, IVINS explained they were refractile spores.  To produce spores of that quality,  not only is very sophisticated laboratory equipment necessary, but one must have extensive experience in the purification process.  Based upon these factors, it is IVINS' opinion that the spores from the mail were grown/created in a laboratory and not in someone's home or garage.  IVINS volunteered that XXXXX "was the spore queen."

          IVINS was shown a diagram he drew and gave to XXXXX on 1/23/02 and asked to explain what he meant by the diagram.  IVINS didn't remember drawing the diagram and asked when he prepared it and what the interview was about.  IVINS was not told the date of the diagram or the context of the interview, and he was asked again to interpret it.  IVINS was non-responsive to the question and would not provide an explanation of what he meant by the diagram.  Rather, he merely identified the names and places on the diagram.

          IVINS was then shown two photographs of spores grown on blood agar plates.  One photograph was labeled "IVIN'S SPORE PREPARATION" while the other was labeled "FTD 1004 FROZEN CULTURE."  IVINS had previously typed captions under each photograph which proportedly explained what was depicted in them.  IVINS was asked if the captions verbalized what was meant by the diagram.

The captions under the photographs read as follows:

"Ames strain - From XXXXX culture collection at USAMRIID.  Similar in appearance to the Bacillus anthracis colonies from mail.  Sent to XXXXX who sent it to XXXXX, who sent it to XXXXXX.  This version of the Ames strain was serially passaged before freezing down and storing in the culture collection."

"Ames strain - from original agar slant from Ames, Iowa, USDA.  This is the version of the Ames strain used by XXXXX and Bruce Ivins.  This version of the Ames strain given to Dugway Proving Ground, Battelle Memorial Research Institute, DRES, and U. of New Mexico."

          After reading the captions and examining the photographs, IVINS explained that he obtained the photographs from XXXXX and typed the captions under them.  Althought IVINS was able to understand what the diagram and captions explained, he could not remember the specific interview which caused him to create them.  Additionally, he would not adopt the statements or explanations as he own beliefs.

          IVINS eventually explained that XXXXX would make several subcultures, or serial passage the organism when growing them.  This caused variants or mutations to appear.  IVINS did not use serial passages.  Rather, his cultures were all grown from the original slant provided to USAMRIID by the USDA, thereby preventing the creation of variants or mutations.  The "BI Culture" depicted in the diagram refers not only to RMR-1029, but also to the spores that were grown in the same fashion as RMR-1029.  So, when IVINS wrote that New Mexico, DRES, Battelle, and Dugway received his culture of Ames, only Battelle and New Mexico actually received RMR-1029.  DRES received some frozen material, and Dugway was given spores from IVINS' culture so they could grow larger lots which were ultimately used to create RMR-1029.

          The spores in the photographs were grown on agar plates which makes it easy to identify mutations or variants.  Ivins never saw the spores from the mail grown on blood agar, so he could not compare them to either his or XXXXX spores.  XXXXX did not make that comparison, however, and told IVINS what XXXXX found.  Namely, that the spores from the mail looked like XXXXX spores and not IVINS'.  The only examination IVINS conducted of the spores from the mail occurred when he grew them on tryptic soy agar (TSA) plates.  When growing spores on TSA plates, however, it is not as easy to identify mutaions or variants.  When growing the spores from the mail on TSA plates, IVINS saw nothing in them which appeared to be an obvious mutation.

The FBI's supplemental pdf document #847551 contains more about what Ivins believed about creating spores.  On page 68 there's this summary of the February 18, 2008 interview:

In IVINS's last interview (with the FBI) he was told the materials in the letters had colony morphology variance.  The way XXXXX lab did it was to pick a single colony so they wouldn't get the morphology variance.  One would swipe over a plate to get the morphology variance.  IVINS claimed his microbiology background taught him to always pick a single colony.

Pdf document #847551, when first released on February 19, 2010, contained an 8-page description of an "off the record" interview with Dr. Bruce Ivins that occurred on June 9, 2008.  That document was on pages 70-77 and has since been removed because it was "off the record."  Fortunately, I downloaded and saved a copy on Feb. 19.    

Here's what it says on the original page 70:

          When growing spores, IVINS and those in his laboratory streak a plate and pick a single representative colony from the plate to innoculate a growth medium.  This ensures all subcultures are identical and do not have morphological variants.   Using the single colony pick also ensures any contaminants present on the plate are not introduced into the growth medium.  If one were to swipe across the plate and use those spores to innoculate the growth medium, the resulting spore growth would have variants.  Therefore, IVINS would expect that all of his subcultures, including RMR-1029, are homogeneous and free of variants.

 And here's what it says beginning at the bottom of page 76 and continuing into page 77:

          At the time of the anthrax mailings, the true history of the Ames strain of anthrax was unknown.  It was widely believed the Ames strain was in so many laboratories it would not be possible to indentify who all had it.  IVINS cannot recall why the FBI was establishing a repository of the Ames strain, or what they hoped to accomplish with the submitted samples.  XXXXX are very tight lipped and did not share information with others regarding the investigation or establishment of the repository.  Ivins attended many meetings regarding the investigation, but he cannot recall what the FBI was doing in the investigation.  Nor can he recall attending any meetings which discussed why the repository was being established or how to prepare the submissions.  IVINS did not receive the protocol or any instructions on the preparation of submissions until May 14, 2002, when he received an email from XXXXX.

And the last paragraph on page 77 is as follows:

If IVINS knew investigators were trying to "follow the genetics" of the Ames strain, he is not sure if he would provide a sample using his single colony pick method, or if he would swipe across the plate.  By swiping across the plate, contaminants and "all kinds of other stuff" are introduced.  Therefore, IVINS would want to talk to the investigators first before preparing the submission.

What all this means is that Ivins had a fundamental misunderstanding about mutations, a.k.a. variants.  He falsely believed that he was able to eliminate variants by using his "single colony pick" method of inoculating growth medium.  And because he used
tryptic soy agar plates when examining his cultures, he was unable to spot mutations as easily as one would when using blood agar plates.  He believed his methods assured that flask RMR-1029 was free of mutations.


When Dr. Ivins chose the Ames strain to create the powders in the letters, he falsely believed it was a strain that could not possibly be traced back to his lab.  He falsely believed it was a common strain used by universities and labs all over the world, and there would be no way to use DNA to trace the spores back to any specific lab.

And he also falsely believed that the spores in his flask RMR-1029 were identical to the supply he believed was stored at the USDA in Ames, Iowa.  He falsely believed the DNA was identical.  He falsely believed that his single colony pick method would assure that there were no differences.  And because he used TSA plates when examining his cultures, he missed the opportunity to notice his error.

Somehow, Dr. Ivins appears to have believed that if he avoided incorporating mutations when starting a growth culture, no new mutations would spontaneously appear in the growth culture.  
In reality, mutations are more a matter of statistics than biology.  Bacterial mutations occur approximately once in a billion generations.  But, that doesn't mean you have to have a billion generations before you see the first mutation.  The very first generation in the billion could be the generation that has the mutation, or it could be any other generation from #2 to #1,000,000,000.

The first error narrowed the search from "countless" laboratories to just 15 or so.  

The second error narrowed the search down from 1,070 samples of the Ames strain in those 15 labs to the single flask of  Ames strain spores controlled and maintained by Dr. Bruce Ivins.

The first error was
as if The Fates had stepped in and provided a miracle for the FBI.  For Dr. Ivins it must have been like being struck by lightning out of a clear blue sky.

The second error was the kind of stupid error investigators always hope that overconfident criminals will make.      

It seems so simple now, but it took seven years for these errors to lead directly to Dr. Ivins. 

List of Revisions and Updates:

March 19, 2010 - Added a comment about Ivins writing on the RMR-1029 "Reference Material Receipt Record" that the Ames strain came from Ames, Iowa.
May 24, 2010 - Added a paragraph about an October 12, 2001 email where Dr. Ivins testily tells people that the source of the Ames stain is the USDA in Iowa.
October 22, 2010 - Added information about NBC Nightly News pointing to the Department of Energy when they meant the Department of Agriculture.
April 28, 2011 - Corrected a quote from the email dated October 12, 2001, by Bruce Ivins. 

© Copyright 2010, 2011 by Ed Lake
All Rights Reserved