Unauthorized Hair Samples Submitted for Analysis (06-MAR-02,	 
GAO-02-488R).							 
                                                                 
The National Interagency Canada Lynx Survey (Protocol) was	 
designed to determine the presence of Canada lynx through	 
deoxyribonucleic acid (DNA) analysis of hair samples recovered	 
from scratch pads in forests in the northern United States.	 
Beginning in January 2002, GAO investigated the submission of	 
unauthorized samples to the University of Montana's laboratory as
part of the National Survey and investigated whether the	 
biologists involved had communications about their submissions.  
GAO found were four instances in which unauthorized hair samples,
not obtained from the Wenatchee and Gifford Pinchot National	 
Forests, were submitted for DNA testing as part of the National  
Survey for those forests. These included one submission on bobcat
hair in 1999, and three submissions of lynx hair in September and
October 2000. The biologists maintain that they submitted these  
samples to test the accuracy of the work done by the laboratory, 
although they knew that the Protocol for the National Survey did 
not provide for such action. They also said that they did not	 
have the authority to make these submissions and that they were  
aware that they had alternatives for testing the laboratory other
than submitting samples as part of the survey. GAO summarized	 
this report in testimony before Congress, see GAO-02-496T.	 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-02-488R					        
    ACCNO:   A02855						        
  TITLE:     Unauthorized Hair Samples Submitted for Analysis	      
     DATE:   03/06/2002 
  SUBJECT:   Biological research				 
	     Endangered animals 				 
	     Endangered species 				 
	     Investigations into federal agencies		 
	     National forests					 
	     Testing						 
	     Ashley National Forest (UT)			 
	     Canada						 
	     Gifford Pinchot National Forest (WA)		 
	     National Interagency Canada Lynx Survey		 
	     National Lynx Detection Protocol			 
	     National Survey for the Ashley National		 
	     Forest						 
                                                                 
	     Wenatchee National Forest (WA)			 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-02-488R
     

United States General Accounting Office Washington, DC 20548

March 6, 2002

Congressional Requesters

Subject: Unauthorized Hair Samples Submitted for Analysis 
This report responds to your December 21, 2001, request and subsequent
conversations with committee staff that we investigate allegations that
biologists with the U.S. Forest Service, U.S. Fish and Wildlife Service, and
Washington Department of Fish and Wildlife improperly submitted hair samples
for deoxyribonucleic acid (DNA) analysis as part of the National Interagency
Canada Lynx Survey (National Survey) during 1999 and 2000. Specifically, you
asked us to (1) investigate allegations that biologists submitted for DNA
analysis lynx hair samples that purported to be from the Gifford Pinchot and
Wenatchee National Forests, but which were actually obtained from other
sources and (2) determine whether the biologists who participated in the
survey communicated about any such submissions. In addition, at the
committee's request, we investigated a separate allegation that "fake" lynx
hair samples were submitted to the laboratory as part of the National Survey
for the Ashley National Forest, in Utah.

The National Survey is designed to determine whether Canada lynx exist in
forests in the northern United States from Vermont to Washington and in high
elevation forests within the Cascade mountain range, such as the Wenatchee
and Gifford Pinchot National Forests, in Washington State. The survey has
recovered samples of animal hair from these forests during specific periods
of time1 in 1999, 2000, and 2001.2 Such surveys are important given that the
Canada lynx is listed as a threatened species and may require certain
actions or land management restrictions under the Endangered Species Act of
19733 in areas where they are found. The Forest Service sponsors the survey,
with assistance from the Fish and Wildlife Service and the Washington
Department of Fish and Wildlife. The University of Montana's Carnivore
Conservation Genetics Laboratory (the laboratory) performed the DNA testing
for the Forest Service.

1 According to the field coordinator for the National Survey, the survey
season primarily runs during

the summer or fall months for approximately 4 weeks, although some surveys
conducted in Midwest

forests are done during the winter.

2 Approximately 800 samples of hair were submitted to the laboratory for the
entire survey in 1999 and

approximately 1,000 such samples were submitted in 2000. According to the
Fish and Wildlife Service,

the number of samples from all of Oregon, Washington, and Idaho in 1999 is
estimated to be no more

than 200.

3 16 U.S.C. sect. 1531 et. seq.

                       GAO-02-488R Canada Lynx Survey

Results in Brief

There were four instances in which unauthorized hair samples not obtained
from the Wenatchee and Gifford Pinchot National Forests were submitted for
DNA testing as part of the National Survey for those forests. These included
submission of bobcat hair in 1999, and three submissions of lynx hair in
September and October 2000. The Forest Service, Fish and Wildlife Service,
and Washington Department of Fish and Wildlife employed the biologists who
made those submissions. These biologists maintain that they submitted these
samples to test the accuracy of the work performed by the laboratory,
although they knew that the protocol for this survey did not provide for
such action. They also stated that they did not have proper authority to
make these submissions.

The survey was conducted pursuant to a National Lynx Detection Protocol
(Protocol), which describes the method for detecting lynx, obtaining lynx
hair samples, and submitting the samples to the laboratory for analysis. The
Protocol did not provide procedures to submit hair samples collected outside
the survey to test the accuracy of laboratory results.

In 2000, one of the participants, a Forest Service biologist with the
Gifford Pinchot National Forest, notified the field coordinator for the
National Survey that a control sample had been submitted, but did not
identify the sample. As a result, the laboratory together with the Forest
Service decided not to analyze any hair samples submitted as part of the
2000 survey for the region that included the Gifford Pinchot and Wenatchee
National Forests until the Forest Service identified the unauthorized
submission. After the unauthorized samples were identified, the laboratory
completed its analysis of the 2000 survey samples, including the three
unauthorized samples. These three samples were determined to be Canada lynx,
and were the only samples submitted for analysis for the Gifford Pinchot and
Wenatchee National Forests that tested positive for Canada lynx. We found
that some of the individuals who participated in the unauthorized
submissions had discussions about submitting unauthorized samples both prior
to and after the submissions.

We found that the assertion that the National Lynx Detection Protocol
permitted submissions of control samples to the Montana Laboratory from the
Ashley National Forest was unfounded. The individual who made this statement
acknowledged that he did so in error.

Background

In 1998, prior to the National Survey, the Forest Service entered into a
contract with Dr. John Weaver of the Wildlife Conservation Society to
conduct surveys, including DNA testing, of whether the Canada lynx was
present in the Cascade Mountain range of Washington and Oregon. In a March
1999, Dr. Weaver reported finding three Canada lynx samples each from the
Wenatchee and Gifford Pinchot National Forests. Dr. Weaver based his
preliminary results on DNA analysis of hair samples recovered from each
forest. However, in June 2001, Dr. Weaver issued a final report to the
Forest Service concluding that hair samples on which the preliminary
findings were based had been contaminated. We previously investigated this
matter at the request of the House Committee on Resources and reported on it
in: U.S. General Accounting

Page 2 GAO-02-488R Canada Lynx Survey

Office, Accidental Contamination of Samples Used in Canadian Lynx Study
Rendered the Study's Preliminary Conclusion Invalid, GAO-01-1018R,
(Washington, D.C.: Aug. 14, 2001). (See app. I.)

The National Survey was conducted pursuant to the Protocol, which was
prepared by the Forest Service and the laboratory. The Protocol describes
the method for detecting lynx. It describes how hair pads containing a scent
lure were to be placed throughout the survey area, and sites for placing the
hair pads would be selected. It also describes how hair samples were to be
collected and submitted to the laboratory for analysis. The Protocol does
not contain any provisions pertaining to the submission of control samples
or tests of the accuracy of the laboratory's work. The laboratory does not
release its analysis of samples until several months after submission. If
the National Survey had detected Canada lynx in an area not previously
recognized as a known lynx habitat, a follow-up snow-tracking survey would
have been conducted in that area to determine whether or not a lynx
population was present.

During 1999 and 2000, hundreds of biologists, who were to collect hair
samples recovered from the forests as part of the survey, were trained in
the procedures set forth in the Protocol. During a training session held on
July 11, 2000, in Portland, Oregon, the field coordinator for the National
Survey announced the findings for the 1999 survey season for Region 6, which
includes the states of Washington and Oregon. Those findings were that the
only lynx hair samples recovered had come from areas surveyed in the
Okanogan National Forest, in Washington. According to the field coordinator,
a discussion took place about how those findings contrasted with the 1998
Weaver study that identified Canada lynx samples from Gifford Pinchot
National Forest. At that time, the results of the Weaver study were believed
to be valid, since the contamination of samples during the Weaver study had
not yet been announced. The biologists attending the session raised
questions about the validity of the National Survey Protocol.

According to a Fish and Wildlife supervisor, although it is standard
scientific procedure to submit control samples to test laboratory results,
such testing is generally provided for in the protocol for a particular
study. However, if submission of control samples were not provided for in
the protocol, a scientist would be expected to notify a lab in advance of
the submission of a control sample. The director of the laboratory at the
University of Montana informed us that, prior to initiating the DNA analysis
for the National Survey, he submitted 20 animal control samples to the Fish
and Wildlife Service's forensic laboratory in Ashland, Oregon. He said that
this laboratory validated his method of DNA analysis by following the
Protocol, and correctly identifying the animal species in all 20 control
samples. He added that the University laboratory also conducted internal
"blind" testing which validated his methods.

In September 2000, a Forest Service biologist who participated in the survey
notified the field coordinator for the National Survey that lynx hair
obtained from an animal held in captivity had been submitted to the
laboratory as a control sample as part of the National Survey for the
Gifford Pinchot National Forest, but did not identify the sample. As a
result, the field coordinator notified the laboratory, which together with
the Forest Service decided not to process any hair samples submitted as part
of the

Page 3 GAO-02-488R Canada Lynx Survey

2000 survey for Region 6, which included the Gifford Pinchot and Wenatchee
National Forests. The Forest Service ultimately determined that two
additional unauthorized submissions had occurred in connection with the
survey for the Gifford Pinchot and Wenatchee National Forests. In February
2001, the Forest Service hired a private investigator to conduct an
investigation. That investigation was completed in June 2001, and its Report
of Investigation summarizes the statements made by the individuals
interviewed by the private investigator.

Biologists Submitted Unauthorized Samples in National Survey

In 1999 and 2000, biologists with the Washington Department of Fish and
Wildlife, Forest Service, and Fish and Wildlife Service submitted
unauthorized hair samples to the Montana laboratory for DNA analysis as part
of the National Survey. The first unauthorized submission was samples of
hair from a bobcat pelt submitted in 1999 by a biologist with the Washington
Department of Fish and Wildlife. Subsequently, in 2000, three additional
unauthorized samples were submitted to the laboratory. First, a biologist
with the Washington Department of Fish and Wildlife submitted a sample from
a captive lynx. Second, a biologist with the Forest Service submitted a
sample obtained from lynx held in captivity at the Northwest Trek.4 Third, a
biologist with the Fish and Wildlife Service submitted a sample also
obtained from lynx held in captivity at the Northwest Trek. Some of the
individuals who participated in the unauthorized submissions had discussions
about submitting unauthorized samples both prior to and after the
submissions. In addition, other employees from these agencies knew of and/or
participated in the unauthorized submissions, including some supervisors.

Submissions of Bobcat Hair in 1999 by a Washington Department of Fish and
Wildlife 
During the 1999 survey season, a biologist (WDFW-1) who was the lead person
on the survey from the Washington Department of Fish and Wildlife for the
Wenatchee National Forest submitted hair samples to the laboratory for
analysis as part of the survey. Some of those samples consisted of hair
obtained from a stuffed tanned bobcat pelt. WDFW-1 informed us that he
submitted the bobcat hair samples to ensure that the results of the
laboratory tests were accurate. He also informed us that he knew that the
Protocol did not provide for the submission of these samples and that he
never informed the laboratory about them. The laboratory analyzed the
samples but could not conduct a valid test of them because they did not
contain

5

sufficient DNA. This information was reported in the official laboratory
results of 1999 DNA testing for the National Survey.

WDFW-1 further informed us that he discussed the submission of the bobcat
hair samples with his supervisor, his office manager, and with another
biologist employed by the Washington Department of Fish and Wildlife
(WDFW-2). WDFW-2

4 The Northwest Trek is a wildlife park that displays North American
wildlife species. Among other
species, the Northwest Trek maintains captured Canada lynx.
5 The director of the laboratory explained that about 20 percent of the hair
samples analyzed do not
contain enough DNA to amplify, and as a result, the hair sample is listed as
"No Qual," meaning no
qualifying DNA. This lack of DNA could be attributed to a number of factors,
such as exposure to
heat, water, chemicals, etc.

Page 4 GAO-02-488R Canada Lynx Survey

subsequently made an unauthorized submission of captured lynx hair during
the 2000 survey season. (See further discussion of WDFW-2's actions below.)

Additionally, WDFW-1 informed us that he subsequently told two biologists, a
Fish and Wildlife Service employee and a Forest Service employee, about his
submission of the bobcat hair. In Spring 2000, WDFW-1 had a conversation
with a biologist for the Forest Service who coordinated the survey results
for the Wenatchee National Forest. The coordinator discussed the fact that
the laboratory was unable to identify the bobcat samples he submitted in
1999. WDFW-1 informed the coordinator that the samples the laboratory could
not identify had been obtained from a bobcat pelt, and that he had submitted
them as "control" samples. The coordinator told us that she did not take any
action because the samples were from a bobcat pelt, the laboratory could not
identify them, and she believed that WDFW-1 was in fact testing or
validating the laboratory's work. In July 2000, WDFW-1 also told the lead
person on the National Survey from the Fish and Wild Life Service for the
Wenatchee National Forest (FWS-1) about the 1999 bobcat hair submissions. As
set forth below, FWS-1 subsequently submitted a lynx hair sample taken from
captive lynx during the 2000 survey season.

Submissions of Hair Samples from Captive Canada Lynx during 2000 Survey
Season 
During the 2000 survey season, three biologists submitted hair for DNA
testing from lynx in captivity as part of the survey of the Wenatchee and
Gifford Pinchot National Forests. These biologists were employees of the
Forest Service, the Fish and Wildlife Service, and the Washington Department
of Fish and Wildlife. In addition, other employees working for these same
agencies became aware of these submissions. Some of these individuals had
discussions about these submissions either before and/or after the actual
unauthorized submissions.

The director of the laboratory confirmed information for 2000 that the
laboratory received one unauthorized submission from the Gifford Pinchot
National Forest and two from the Wenatchee National Forest. He stated that
all three of these control samples tested positive for Canada lynx. In fact,
these were the only samples that tested positive for Canada lynx hair from
the Gifford Pinchot and Wenatchee National Forests. He added that if someone
wished to submit a control sample to the laboratory for DNA analysis he
would have accepted it, so long as it was labeled a control sample, so it
could not be mistakenly included with the survey's published laboratory
report. He also said that there was no procedure whereby the biologists who
submitted samples would receive preliminary results, so that they could
subsequently notify the laboratory of their unauthorized submissions.

Submissions of Captive Lynx Hair by a Biologist with the Washington
Department of Fish and Wildlife

As previously discussed, in late October 1999, WDFW-1 told WDFW-2, a fellow
biologist at the Washington Department of Fish and Wildlife, that he
submitted bobcat hair samples as part of the 1999 survey season. WDFW-1 also
asked that WDFW-2 help with the National Survey. On September 18, 2000,
WDFW-2 submitted hair from a captive lynx to the laboratory as part of the
survey for the Wenatchee National Forest. WDFW-2 informed us that on
September 14, 2000, his supervisor

Page 5 GAO-02-488R Canada Lynx Survey

told him that local authorities had captured a pet Canada lynx and were
keeping it temporarily in the office until it could be returned to its
owner. WDFW-2 asked and received permission from his supervisor to obtain
hair from this animal to submit to the laboratory as a test sample for the
National Survey. The supervisor assisted WDFW-2 in collecting the hair
samples. WDFW-2 also informed us that he informed WDFW-1 that he had
submitted the captive lynx sample to the laboratory.

Neither WDFW-2 nor his supervisor notified the laboratory concerning the
submission of unauthorized samples. WDFW-2 first disclosed his submission
when he contacted the private investigator for the Forest Service and
informed her of his actions.

Submissions of Lynx Hair by Biologists with the Forest Service

A Forest Service biologist with the Gifford Pinchot National Forest (FS-1)6
provided Canada lynx hair samples he obtained from the Northwest Trek to a
Forest Service colleague (FS-2), and asked him to submit them to the
laboratory. On September 26, 2000, FS-2 submitted them as part of the
Gifford Pinchot National Forest survey. FS-1 acknowledged that he was not in
a position of authority to take such action and that the Protocol did not
provide for it. FS-1 also said that he could have submitted a sample labeled
as a control to test the laboratory. FS-1's supervisor told us that FS-1
informed him that he was going to submit a sample to test the laboratory.
FS-1's supervisor added that, at that time, he was acting for FS-1's actual
supervisor while that individual was on a detail. He added that he was not
familiar with the National Survey and did not know about the Protocol. On
September 29, 2000, FS-1 left a telephone message with the field coordinator
for the National Survey stating that he had sent a sample, which he called a
control sample. The field coordinator told us that in this message, FS-1 did
not identify which sample from the Gifford Pinchot National Forest contained
the control.

The second Forest Service biologist (FS-3) who went to the Northwest Trek
was not involved in the unauthorized submissions to the laboratory; however,
she was involved in the discussions prior to the submissions.

 Submissions of Lynx Hair by biologists with the Fish and Wildlife Service

On October 18, 2000, FWS-1 submitted hair samples from captive Canada lynx
to the laboratory as part of the survey for the Wenatchee National Forest.
These samples had been obtained from lynx held in captivity in the Northwest
Trek. FWS-1 told us that he submitted the lynx hair to test the ability of
the laboratory to identify Canada lynx through DNA testing and that he did
not notify the laboratory. FWS-1 also told us that he was aware at the time
he submitted the captive lynx hairs that WDFW-1 had submitted unauthorized
samples to the laboratory in 1999; that he knew the Protocol did not provide
for the submission of control samples; and that he did not have proper
authorization to take such action. FWS-1 also said that he could have
submitted a sample labeled as a control to test the laboratory.

                   6 FS-1 retired on September 30, 2000.

                   Page 6 GAO-02-488R Canada Lynx Survey

FWS-1 obtained the captive lynx hair samples from another Fish and Wildlife
Service biologist (FWS-2). FWS-2 had traveled to the Northwest Trek on
September 19, 2000, with two biologists from the Forest Service. During that
trip, both FWS-2 and one of the Forest Service biologists, FS-1, who
collected captive lynx hair samples discussed submitting them to the
laboratory and subsequently provided them to other biologists who made the
submissions. When FWS-1 submitted the lynx hair samples to the laboratory,
he knew that FS-1 had also obtained lynx hairs during the trip to the
Northwest Trek and was sending them to the laboratory.

The Fish and Wildlife Service division manager for the Northwest Forest Plan
where FWS-1 and FWS-2 work stated that these biologists believed that
scientists who focus on DNA analysis have a narrow view of the lynx habitat
and that the description of the lynx habitat should be broadened. As a
result of hearing these concerns, these biologists were instructed to
prepare a paper laying out the basis for their concerns. Given the lack of
knowledge about lynx abundance and distribution on the west slope of the
Cascades, the issues were (1) how to characterize possible lynx habitat, and
(2) to what extent possible lynx habitat should be protected when lynx are
not known to currently be using the habitat. This paper was completed and
submitted to the Interagency Lynx Steering Committee at its October 2001
meeting. The Steering Committee reviewed the paper and concluded that it
lacked substance in which to change or modify either the lynx conservation
assessment and strategy or the written direction for lynx habitat mapping.

Claim that the National Lynx Detection Protocol Permitted Submissions of
Control Samples from the Ashley National Forest Was Unfounded

At the committee's request, we investigated another allegation that there
were submissions of "fake" lynx hair samples to the University of Montana
laboratory, as part of the National Survey for the Ashley National Forest in
Utah. We interviewed the Forest Service biologist who authored an e-mail
dated December 18, 2001, that included the following statement:

"Here in Utah we're [sic] added several `fake' lynx hairs to our snare
surveys. This was done by the US Forest Service, USFWS [U. S. Fish and
Wildlife Service] and the Utah Division of Wildlife to test the validity of
the lab sampling techniques. On all "fake" samples the lab in Montana
correctly identified the sample as a lynx. The lab then informs us of the
positive lynx results and we inform them of our test. I can't say what the
Washington forests were doing, but I think this is the same sort of validity
test."

The biologist said that his e-mail message, which he sent to a number of
different entities mistakenly, implied that under the National Survey
Protocol control samples had been sent to the Montana laboratory from Utah.
He added that he regrets this mistake.

Scope and Methodology

Beginning in January 2002, we investigated the facts and circumstances
surrounding the unauthorized submission of hair samples to the Montana
laboratory during the 1999 and 2000 survey seasons. We interviewed the
private investigator and reviewed

Page 7 GAO-02-488R Canada Lynx Survey

her Report of Investigation, which included signed declarations of many of
the individuals who participated in those unauthorized submissions. We also
interviewed current and former employees of the Forest Service, Fish and
Wildlife Service, and Washington Department of Fish and Wildlife, including
the biologists involved in the submissions. In addition, we interviewed
employees from the University of Montana and the Northwest Trek. This
investigation did not review the Protocol or determine the impact of the
actions taken by these biologists. GAO has been requested to conduct a
separate review regarding the Protocol, and generally accepted practices, if
any, that should have been applied by the National Survey.

Agency Comments

The Forest Service, Fish and Wildlife, and Washington Department of Fish and
Wildlife provided comments on a draft of this report, in which they
concurred with the facts and its findings. They also provided technical
corrections and, where appropriate, we have made those corrections.

We will send copies of this report to the Chief, Forest Service; the
Director, Fish and Wildlife Service; and the Director of the Washington
Department of Fish and Wildlife Service. We will make copies available to
others on request. The report will also be available at www.gao.gov. If you
have any questions or need additional information, please contact Assistant
Director Patrick Sullivan at (202) 512-6722. Key contributors to this report
were Senior Special Agents George Ogilvie and Thomas Wiley, Senior Attorney
Barry Shillito, and Assistant General Counsel Robert Cramer.

Ronald Malfi
Acting Managing Director
Office of Special Investigations

Page 8 GAO-02-488R Canada Lynx Survey

List of Requesters

The Honorable James V. Hansen
Chairman
Committee on Resources
House of Representatives

The Honorable Scott McInnis
Chairman
Subcommittee on Forests And Forest Health
Committee on Resources
House of Representatives

The Honorable Richard W. Pombo
House of Representatives

The Honorable Mike Simpson
House of Representatives

The Honorable Greg Walden
House of Representatives

The Honorable John Peterson
House of Representatives

The Honorable Doc Hastings
House of Representatives

The Honorable Tom Tancredo
House of Representatives
*** End of document. ***