Wildlife as biosamplers:
Contaminants in hair of elk harvested
Near the anaconda smelter site
Karen L. Gillespie, Environmental Engineering Department, Montana Tech of The University of
Montana, Butte, MT 59701
Holly G. Peterson, Environmental Engineering Department, Montana Tech of The University of
Montana, Butte, MT 59701
Casey M. Clark, Environmental Engineering Department, Montana Tech of The University of Montana,
Butte, MT 59701
Jennifer S. Black, Environmental Engineering Department, Montana Tech of The University of
Montana, Butte, MT 59701
The purpose of this research was to test a new way of investigating biological uptake of
smelting-related contaminants with a focus on harvested wildlife. Specific objectives were 1)
to collect hair samples from elk (Cervus elaphus) harvested in the vicinity of the Anaconda
Smelter National Priority List Site in Montana, 2) to analyze the samples using inductively
coupled plasma – mass spectrometry (ICP-MS), and 3) to identify potential elements of
concern from the data. Hair samples were collected from 56 elk, and concentration data
were processed using a hazard quotient/index approach based on concepts commonly used
in fields of ecological and human health risk analyses. Arsenic concentrations in the hair
decreased as a function of increasing distance from the Anaconda smelter stack, and 57 %
of the elk sampled were identified as animals of concern. For elk harvested within 25 km of
the stack, elements of concern were aluminum, arsenic, barium, boron, lithium, manganese,
molybdenum, strontium, and vanadium. For elk harvested within 76-101.5 km of the stack,
elements of concern were aluminum, barium, boron, lithium, and manganese. Hazard
indices for uranium, arsenic, cadmium, and lithium were larger by factors of ~17, 9, 7, and 6,
respectively, for elk harvested within 25 km of the stack compared to hazard indices for elk
harvested within 76-101.5 km.
Key Words: arsenic, pollution, biomonitoring, hair samples.
uptake using domestic pets as bioindicators
For more than a century, smelting
of environmental conditions in Butte and
activities in Anaconda caused wide-spread
Anaconda (Peterson and Madden 2006).
contamination in Montana. The United
The technique involved sampling the hair of
States Environmental Protection Agency
domestic dogs (Canis lupus familiaris) and
(USEPA) listed the Anaconda Smelter
cats (Felis catus), analysis by inductively
Site on the Superfund National Priorities
coupled plasma-mass spectrometry (ICP-
List (NPL) in 1983 (USEPA 1998). As a
MS), and identification of elements of
part of the Superfund activities, several
concern with a hazard index approach
contaminants have been characterized,
similar to methods employed in the field
risks to human health have been estimated,
of risk analysis. More than 400 samples
and some cleanup has taken place.
from the domestic pet population identified
Unfortunately, few data have addressed
eight elements of concern (aluminum,
biological uptake of contaminants by human
arsenic, boron, lead, lithium, manganese,
or wildlife populations.
molybdenum, and selenium) in residential
neighborhoods of Butte and Anaconda
During the past decade, we developed
(Madden 2006, Barry 2006, Peterson and
and tested a new way to study contaminant
Barry 2006, and Robertson 2007).
Similar to our previous field campaigns,
as evidence of illegal drug and alcohol use
the overall goal of the research presented
(Pragst and Balikova 2006).
in this paper was to improve understanding
Elements in the bloodstream of
of biological uptake of environmental
mammals are transferred from the root cells
contaminants. Instead of domestic pets,
into the hair shaft during growth stages
however, we addressed harvested wildlife.
(Beernaert et al. 2007). Hair consists of
Specifically, we targeted the local elk
keratin with cysteine sulfhydryl groups
(Cervus elaphus) population, and objectives
capable of binding to metals and other
were 1) to collect hair samples from elk
elements (Mandal and Suzuki 2002).
harvested in the vicinity of the Anaconda
Siedel et al. (2001) and others presented
Smelter NPL Site, 2) to analyze the samples
uncertainties about external contamination,
with ICP-MS, and 3) to identify potential
but Hinwood et al. (2003) concluded hair
elements of concern from the data.
sampling to be a good “screening-level”
Uptake Of Contaminants By
technique for studying environmental
exposure if care is taken to properly handle,
rinse, and analyze the specimens.
As part of remedial investigation/
feasibility studies of the Anaconda Smelter
In addition to our research in Butte
Site, numerous sampling campaigns were
and Anaconda, field campaigns elsewhere
conducted to characterize risk and burden of
have been advancing the legitimacy of
pollutants on the surrounding environment
hair sampling as a research tool. Rashed
(USEPA 1998). Few studies, however,
and Soltan (2005), for example, analyzed
were performed to characterize exposure
hair of goats (Capra hircus), sheep (Ovis
and uptake of these contaminants for
aries), and camels (Camelus) in Egypt, and
resident wildlife species, nor to monitor the
concentrations of cadmium, cobalt, iron,
efficacy of environmental cleanup. Initial
lead, manganese, and nickel in the hair
assessment of ecological risk used a simple,
correlated to contaminants in vegetation
predictive food chain model (USEPA 1998)
consumed by the animals. D’Have et al.
without direct consideration of wildlife.
(2009) linked concentrations of lead and
Following the initial assessment, a handful
cadmium in hair of European Hedgehog
of projects addressed contaminants in small
(Erinaceus europaeus) to contaminant
mammals and avian species (Hopper et
concentrations in the soil. Mercury
al. 2002). From recreational and wildlife
concentrations were studied in hair of
management viewpoints, however, large
wild boars (Sus scrofa) by Sobanska
mammal populations in the vicinity of the
(2005), in hair of deer mice (Peromyscus
Anaconda Smelter Site were neither sampled
maniculatus) by Waring and Douglass
(2007), and in hair of sled dogs by Dunlap
et al. (2007). Beernaert et al. (2007)
Hair samples As Biosamplers Of
found linear relationships of lead and
cadmium among hair, kidney, and liver
samples in the Wood Mouse (Apodemus
We were the first to propose domestic
sylvaticus). McLean et al. (2009) also
pet hair as a unique tool for studying
linked concentrations of lead and cadmium
residential exposure to mining-related
in soil with hair concentrations from small
contaminants (Peterson and Madden 2006).
mammals residing near a decommissioned
In human populations, however, hair and
lead and zinc smelter in Australia. Finally,
toenails have been used for many years in
pollution in Nairobi, Kenya, was studied
the field of forensics to determine possible
using hair samples from residential pets and
cause of death by ingestion of toxic metals
wildlife (Mwaniki 2007). Prior to results
and/or medicines (Chatt and Katz 1988).
summarized here, however, no data were
Likewise, human hair samples have been
available for wild game species residing on
used by law enforcement and by employers
40 Gillespie et al.
or near contaminated Superfund sites in the
A hazard quotient (HQ ) of element i for
animal j was calculated as:
As described in more detail by Gillespie
(2011), we conducted field campaigns in the
where C was concentration of element i in
Anaconda, Montana area during two hunting
the hair sample of animal j; and RfC values
seasons (October-November of 2009 and
were the same reference concentrations used
October-November of 2010). In 2009, we
in our other research projects (TEI 2005,
collected hair samples from wild game at
Peterson and Madden 2006, Madden 2006,
the Montana Fish, Wildlife and Parks (FWP)
Barry 2006, and Robertson 2007).
hunting check station located along Mill
Creek. In 2010, we obtained samples at
In addition to hazard quotients,
the Mill Creek check station and at another
two hazard indices were examined. A
FWP check station in Divide, Montana.
normalized animal hazard index (HI ) was
Both stations were selected based on elk
calculated by summing the hazard quotients
populations commonly harvested in the
across the elements:
vicinity of the Anaconda Smelter NPL Site.
Regarding experimental protocol, we
completed a questionnaire for each animal
in our study. Specimens were assigned
identification numbers. Hunters were
also asked in which hunting districts and
where N was the number of elements.
drainages the animals were harvested. Other
Likewise, a normalized element hazard
information, such as sex and approximate
index (HI ) was calculated by summing
age of the animal, was documented
the hazard quotients across the number of
Hair samples, ~ 150 milligrams (mg)
in size, were removed from the harvested
animals’ coats with clean stainless steel
scissors. When possible, the hair sample
where M was the total number of animals
was collected from the region between the
sampled. As per the method of Peterson and
shoulder and neck of the animal. Samples
Madden (2006), the target value was 1.0 for
were sealed in contaminant-free envelopes
both HI and HI . Animals with HI values ≥
and stored until the end of each hunting
1.0 were defined as animals of concern, and
season when they were sent to Trace
elements with HI values ≥ 1.0 were defined
Elements, Incorporated (Addison, Texas).
as elements of concern.
Hair was examined with a microscope and
rinsed repeatedly with de-ionized water
to remove external soil particles prior to
analysis by inductively coupled plasma-
During field campaigns in 2009 and
mass spectrometry. Trace Elements,
2010, we collected hair samples from 56 elk
Incorporated, is a licensed, certified clinical
harvested in the vicinity of the Anaconda
Smelter NPL Site (Fig. 1). Harvest locations
of the elk in the study corresponded to
We analyzed the concentration data
distances ranging 7.5-101.5 km from the
using the hazard index technique of Peterson
Anaconda smelter stack, and our dataset
and Madden (2006). The method is based
consisted of hair samples from 31 adults, 25
on concepts commonly used in the fields of
sub-adults, 28 males, and 27 females (Table
ecological and human health risk analyses.
1). Adult elk in this project were defined
Wildlife as Biosamplers: Contaminants in Hair of Elk Harvested Near the Anaconda Smelter Site 41
Figure 1. Map of harvest locations (circles) for 56 elk sampled during field campaigns
in 2009 and 2010. The Anaconda smelter stack is represented by a diamond, and harvest
locations are labeled with sample identification numbers. As a scale of reference, the distance
between the stack and harvest location is 19.1 km for Sample 2010-28, and the corresponding
distance is 101.5 km for Sample 2010-59.
as 3 years of age and older, and sub-adults
Using Equation (1) to calculate hazard
were younger than 3 years.
quotients, Figure 3 depicts data for 14
Concentrations of arsenic in the elk
elements in Samples 2010-28 and 2010-59.
hair decreased as a function of increasing
The elk for Sample 2010-28 was harvested
distance from the stack (Fig. 2). Thirty-six
~19.1 km southeast of the smelter stack, and
(~ 64 percent) of the samples contained
the elk for Sample 2010-59 was harvested
arsenic concentrations greater than a
~101.5 km southwest of the stack. Thirteen
reference concentration of 0.20 parts per
elements in Sample 2010-28 exceeded
million (ppm), and based on the best-fit
a hazard quotient of 1.0, and elements
equation in Figure 2, arsenic concentrations
with the highest HQ values were lithium
did not fall below 0.20 ppm until harvest
(42.9), manganese (25.6), and arsenic
distances were greater than ~ 58 km from
(24.5). In contrast, the only elements in
Sample 2010-59 with HQ values ≥ 1.0 were
42 Gillespie et al.
0.6 0.7 0.7 0.9 0.8 1.2 5.0 10.7 1.1 1.2 1.4 1.1
0.6 1.7 2.0 0.7 0.7 1.7 0.6 0.5 0.5 1.1 1.8 0.6 0.7 0.5 1.1
), arsenic j
Senic hq j 0.4 0.4 0.3 1.5 2.0 2.2 2.6 24.5 4.1 1.8 1.8 0.6 0.6 1.0 0.8 0.7 1.1 1.0 9.9 0.4 0.2 0.4 6.7 4.1 1.3 1.0 0.5 2.5
0.08 0.07 0.05 0.30 0.40 0.43 0.51 4.89 0.82 0.35 0.35 0.12 0.5 0.20 0.15 0.14 0.22 0.19 1.98 0.07 0.04 0.07 1.33 0.82 0.26 0.20 0.09 0.50
Ar (ppm)c j
X j (km) 93.1 93.1 59.1 85.6 83.5 51.6 51.4 19.1 19.1 25.5 45.9 45.9 0.09 54.0 74.8 62.2 41.1 41.1 7.5 81.0 101.5 101.5 19.1 27.4 32.1 36.5 32.1 21.4
Anaconda smelter stack (X
29 30 31 32 33 34 35 36 37 38 39 40 51.6 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56
A A A A
S S S S 41
FS FA M FA M M M M FA M M M M FA FA M FA M M FS FA M FS M FA FS FA M
lt, U = U
2010-20 2010-21 2010-22 2010-23 2010-24 2010-26 2010-27 2010-28 2010-29 2010-30 2010-32 2010-33 2010-35 2010-36 2010-37 2010-43 2010-44 2010-45 2010-48 2010-57 2010-58 2010-59 2010-60 2010-61 2010-62 2010-63 2010-64 2010-67
) for 56 elk harvested during field campaigns in 2009-2010.
S = F
1.9 1.4 1.7 2.0 1.3 1.2 1.2 1.0 0.8 1.1 1.3 0.7 1.2 0.5 1.8 0.8 0.7 4.1 1.8 0.7 0.7 1.4 1.6 1.4 0.8 0.4 1.1 2.2
Senic hq j 4.2 5.6 3.5 4.1 1.4 3.5 1.5 5.3 1.5 2.2 5.8 0.3 2.2 0.2 0.5 0.6 0.6
6.5 2.0 0.6 1.7 2.0 2.9 0.9 0.3 0.6 1.5
A = F
0.83 1.11 0.69 0.81 0.27 0.70 0.29 1.06 0.29 0.43 1.16 0.06 0.44 0.04 0.09 0.11 0.12 2.06 1.30 0.40 0.12 0.33 0.39 0.57 0.17 0.05 0.12 0.30
), and animal hazard index (HI j
Ar (ppm)c j
, sex-age class, animal number (j), distance between harvest location and
18.3 11.7 49.5 25.0 83.4 35.3 47.0 27.9 23.8 25.0 35.1 45.9 36.5 85.9 85.9 97.5 50.8 54.4 30.9 34.9 40.8 40.8 49.2 49.2 97.5 97.5 97.5 85.9
S = M
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
A = M
), arsenic hazard quotient (HQ
A S S A S
S A A
M M M M M FS FS FA FS FA FA FS M M US M FA FA FA M FS M M M FS FA M FA
Sample identification number
2009-01 2009-02 2009-03 2009-04 2009-05 2009-06 2009-07 2009-08 2009-09 2009-10 2009-11 2010-01 2010-02 2010-03 2010-05 2010-06 2010-07 2010-09 2010-10 2010-11 2010-12 2010-13 2010-14 2010-15 2010-16 2010-17 2010-18 2010-19
T concentration (C
Wildlife as Biosamplers: Contaminants in Hair of Elk Harvested Near the Anaconda Smelter Site 43
Figure 2. Arsenic concentration in hair samples from elk harvested at distances ranging
7.5-101.5 km from the Anaconda smelter stack. A best-fit trendline of the data and an
arsenic reference concentration of 0.20 ppm are also shown.
Figure 3. Hazard quotient values for Samples 2010-28 and 2010-59. These hair samples
were collected from adult elk harvested 19.1 and 101.5 km from the Anaconda smelter
stack, respectively. For 14 elements, the corresponding animal hazard indices (HI ) are
10.7 and 0.5.
44 Gillespie et al.
aluminum (1.5) and boron (1.3). For these
Zone 1 to Zone 4 for most of the elements.
two samples, the animal hazard indices
In Zone 1, HI was ≥ 1.0 for aluminum,
from Equation (2) were 10.7 and 0.5,
arsenic, barium, boron, lithium, manganese,
respectively. Based on the HI value ≥ 1.0,
strontium, and vanadium. Of these elements
the elk corresponding to Sample 2010-28
of concern, the largest hazard indices in
was identified as an animal of concern, as
Zone 1 were for arsenic (6.5), lithium (6.0),
were 32 (~ 57 percent) of the 56 elk sampled
and manganese (5.3). For elk harvested
in Zone 4, however, the only elements of
As per concentration statistics
concern were aluminum, barium, boron,
of the data set, no samples contained
lithium, and manganese, with manganese
concentrations exceeding the reference
exhibiting the largest hazard index (2.8).
concentration for zinc (Table 2), and average
Zinc showed no spatial variation among the
and median concentrations were lower than
zones, but the element hazard indices were
reference concentrations for cadmium, iron,
dramatically larger in Zone 1 compared to
lead, molybdenum, strontium, uranium,
Zone 4 by factors of 16.7 for uranium, 8.9
vanadium, and zinc. However, because
for arsenic, 6.7 for cadmium, and 5.6 for
the main source of contamination was the
smelter, we divided the samples into four
zones according to distance (X) between
harvest location and the Anaconda smelter
In this project, we proposed and tested
stack: Zone 1 (X<25 km), Zone 2 (26-50
a novel way to study contaminant uptake
km), Zone 3 (51-75 km), and Zone 4 (76-
using elk as biosamplers of environmental
101.5 km). Element hazard indices were
conditions near the Anaconda Smelter
calculated with Equation (3) for each of
NPL Site. Even though the Environmental
the four zones (Table 3), and elements of
Protection Agency has been directing
concern were identified.
cleanup activities in the area for many years,
While Zones 2 and 3 revealed more
large mammal populations have not been
variability, HI values decreased from
addressed. Based on our data, elk in the
Table 2. Element (i), reference concentration (RfC ), minimum (Min), maximum (Max),
average (Avg), median (Med), and standard deviation (Stdev) concentration (C ) measured in
hair samples of 56 elk harvested during field campaigns in 2009-2010.
Med C stdev c
(ppm) (ppm) (ppm) (ppm) (ppm) (ppm)
Wildlife as Biosamplers: Contaminants in Hair of Elk Harvested Near the Anaconda Smelter Site 45
Table 3. Element (i) and element hazard index (HI ) values for Zone 1 (X <25 km), Zone 2
(26-50 km), Zone 3 (51-75 km), and Zone 4 (76-101.5 km). Also shown is the ratio of the
hazard indices for Zone 1:Zone 4.
Zone 4 zone1:zone4
Hi hi hi hi ratio
0.5 0.5 0.5 0.5 1.0
vicinity of the site are still being exposed to
Specifically, studies should address the
significant amounts of contamination. While
impact of contaminants on the health of
results from this campaign are site-specific,
the game animals; however, hunters and
our technique could be used at other sites
their families are also at risk of developing
where anthropogenic pollution is of concern,
health problems if they routinely ingest wild
and where efficacy of remediation is in
meat contaminated with arsenic and other
contaminants. This latter topic will be the
Concentrations for many contaminants
focus of a follow-up paper by our research
in our dataset increased for elk harvested
closer to the Anaconda stack. Elk are
migratory animals, however, and we know
that contaminant concentrations in hair
We conducted the first known field
are not solely dependent on environmental
campaign using hair samples to investigate
conditions at the harvest locations. To
uptake of environmental contaminants for
advance our fundamental understanding of
harvested wildlife residing in a Superfund
variability within and among the samples,
area. Based on 56 elk harvested in the
we recommend future research to merge
vicinity of the Anaconda Smelter NPL Site
hair sampling with radio-collar tracking
during hunting seasons in 2009 and 2010,
for a subset of elk during the growth
~57% of the elk sampled were identified as
period of the hair (i.e., for several months
animals of concern. Manganese, arsenic,
prior to hunting season). With subsequent
and lithium were identified as elements of
environmental sampling along the migratory
most concern, especially for elk harvested
path, uptake of contaminants into hair could
within 25 km of the smelter stack. In
be correlated to pollution concentrations
addition, hazard indices for uranium,
in the soil, vegetation, and water within
arsenic, cadmium, and lithium were larger
a specific habitat. In addition, in-depth
for elk harvested within 25 km of the stack
medical research should scrutinize health
by factors of ~17, 9, 7, and 6, respectively,
effects associated with uptake of these
compared to elk harvested within 76-101.5
pollutants by the local elk population.
46 Gillespie et al.
Gillespie, K. 2011. Harvested wildlife as
The authors thank Ms. Vanna Boccadori
biosamplers: elk and deer in the vicinity
and Mr. Braden Burkholder of the Montana
of the Anaconda smelter national priority
Fish, Wildlife and Parks (FWP) for their
site. Master of Science Thesis. Montana
interest in our project and for letting us
Tech of The University of Montana.
obtain samples at the Mill Creek and
Butte, MT. 333 pp.
Divide game stations. We also recognize
Hinwood, A.L., M.R. Sim, D. Jolley, N. de
the hunters who were willing to participate
Klerk, E.B. Bastone, J. Gerostamoulos,
in the campaign. In addition, Mr. John
and O.H. Drummer. 2003. Hair
Helfrich is acknowledged for sharing his
and toenail arsenic concentrations
invaluable knowledge about hunting and
of residents living in areas with high
trapping. Finally, we are grateful to Dr.
environmental arsenic concentrations.
David Watts and the laboratory personnel at
Environmental Health Perspectives
Trace Elements, Incorporated for agreeing
to analyze our somewhat-unconventional
Hopper, M., G. Cobb, and S. McMurry.
2002. Wildlife biomonitoring at the
Anaconda smelter site Deer Lodge
county, Montana. The Institute of
Barry, S. 2006. Domestic pets as
Environmental and Human Health Texas
biosamplers of mining-related
Tech University, Lubbock, TX. 380 pp.
contaminants. Master of Science Thesis.
Madden, M. 2006. Development of a
Montana Tech of The University of
biomonitoring technique using domestic
Montana. Butte, MT. 279 pp.
pets as sentinel species in a mining
Beernaert, J., J. Scheirs, H. Leirs, R.
impacted community. M.S. Thesis.
Blust, and R. Verhaagen. 2007.
Montana Tech of The University of
Non-destructive pollution exposure
Montana, Butte, MT. 141 pp.
assessment by means of wood mice hair.
Mandal, B. and K. Suzuki, 2002. Arsenic
Environmental Pollution 145:443-451.
around the world: a review. Talanta
Chatt, A. and S.A. Katz. 1988. Hair
analysis: applications in the biomedical
McLean, C., C. Koller, J. Rodger, and G.
and environmental sciences. VCH
MacFarlane. 2009. Mammalian hair as
Publishers, Inc., New York, NY. 134 pp.
an accumulative bioindicator of metal
D’Have, H., F. Vermeulen, V. Mubiana, N.
bioavailability in Australian terrestrial
Van den Brink, R. Blust, L. Berveots, and
environments. Science of the Total
W. De Coen. 2009. Relevance of hair
and spines of the European hedgehog
Mwaniki, G. 2007. Domestic pets and
(Erinaceus europaeus) as biomonitoring
wildlife as biosamplers of environmental
tissues for arsenic and metals in
contaminants in Nairobi, Kenya. Master
relation to blood. Science of the Total
of Science Thesis. Montana Tech of The
University of Montana, Butte, MT. 198 pp.
Dunlap K., A. Reynolds, P. Bowers, and L.
Peterson, H. and M. Madden. 2006.
Duffy. 2007. Hair analysis in sled dogs
Development of a new biomonitoring
(Canis lupus familiaris) illustrates a
technique using domestic pets as sentinel
linkage of mercury exposure along the
species. Intermountain Journal of
Yukon River with human subsistence
food systems. Science of the Total
Wildlife as Biosamplers: Contaminants in Hair of Elk Harvested Near the Anaconda Smelter Site 47
Peterson, H. and S. Barry. 2006. Final
Sobanska, M. 2005. Wild boar hair (Sus
report: domestic pets as biosamplers of
scrofa) as a non-invasive indicator of
mining-related contaminants. United
mercury pollution. Science of the Total
States Environmental Protection Agency,
National Risk Management Research
Trace Elements, Incorporated (TEI). 2005.
Laboratory. 118 pp.
Reference ranges and HTMA materials
Pragst, F. and M.A. Balikova. 2006. State
and methods. Trace Elements, Inc., Box
of the art in hair analysis for detection of
514, Addison, TX 75001.
drug and alcohol abuse. Clinica Chimica
United States Environmental Protection
Agency (USEPA). 1998. Superfund
Rashed, M. and M. Soltan. 2005. Animal
record of decision: Anaconda regional
hair as biological indicator for heavy
water, waste, and soils operable unit
metal pollution in urban and rural
Anaconda smelter NPL site. EPA/ROD/
areas. Environmental Monitoring and
R08-98/096. EPA Region VIII, Montana
Assessment 110: 41-53.
Office, Helena, MT.
Robertson, T. 2007. Domestic pets as
Waring, T. and R. Douglass. 2007. Mercury
biosamplers of toxic elements at a
in mouse hair: a monitoring tool for
Superfund site. Master of Science
environmental exposure. Intermountain
Thesis. Montana Tech of The University
Journal of Sciences 13:110-115.
of Montana, Butte, MT. 250 pp.
Siedel, S., R. Kreutzer, D. Smith,
S. McNeel, and D. Gliss. 2001.
Assessment of commercial laboratories
performing hair mineral analysis.
Journal of the American Medical
Received June 16, 2011
Accepted August 3, 2012
48 Gillespie et al.