Mink as a sentinel species in environmental health. Environ Res

National Wildlife Research Center, Canadian Wildlife Service, Environment Canada, Ottawa, Ontario, Canada.
Environmental Research (Impact Factor: 4.37). 02/2007; 103(1):130-44. DOI: 10.1016/j.envres.2006.04.005
Source: PubMed


The concept of "sentinel species" is important in the environmental health sciences because sentinel species can provide integrated and relevant information on the types, amounts, availability, and effects of environmental contaminants. Here we discuss the use of mink (Mustela vison) as a sentinel organism by reviewing the pertinent literature from exposure- and effects-based studies. The review focuses on mercury (Hg) and polychlorinated biphenyls (PCBs), as they are persistent, ubiquitous, and bioaccumulative contaminants of concern to both humans and wildlife. Mink are widely distributed, abundant, and regularly trapped in temperate, aquatic ecosystems, and this makes them an excellent model to address issues in environmental pollution on both temporal and spatial scales. As a high-trophic-level, piscivorous mammal, mink can bioaccumulate appreciable concentrations of certain pollutants and have been shown to be sensitive to their toxic effects. The husbandry and life history of mink are well understood, and this has permitted controlled dosing experiments to be conducted using animals reared in captivity. These manipulative studies have yielded important quantitative information on exposure-response relationships and benchmarks of adverse health effects, and have also allowed the cellular mechanisms underlying toxic effects to be explored. Furthermore, the data accrued from the laboratory continue to validate observations made in the field. Research derived from mink can bridge and integrate multiple disciplines, and the information collected from this species has allowed environmental health scientists to better understand and characterize pollution effects on ecosystems.


Available from: Laurie Hing Man Chan, Sep 04, 2015
This article was originally published in a journal published by
Elsevier, and the attached copy is provided by Elsevier for the
author’s benefit and for the benefit of the author’s institution, for
non-commercial research and educational use including without
limitation use in instruction at your institution, sending it to specific
colleagues that you know, and providing a copy to your institution’s
All other uses, reproduction and distribution, including without
limitation commercial reprints, selling or licensing copies or access,
or posting on open internet sites, your personal or institution’s
website or repository, are prohibited. For exceptions, permission
may be sought for such use through Elsevier’s permissions site at:
Page 1
Author's personal copy
Environmental Research 103 (2007) 130144
Mink as a sentinel species in environmental health
Niladri Basu
, Anton M. Scheuhammer
, Steven J. Bursian
, John Elliott
Kirsti Rouvinen-Watt
, Hing Man Chan
National Wildlife Research Center, Canadian Wildlife Service, Environment Canada, Ottawa, Ont., Canada
Department of Natural Resource Sciences, McGill University, Ste. Anne de Bellevue, Que., Canada
Center for Indigenous Peoples’ Nutrition and Environment (CINE), McGill University, Ste. Anne de Bellevue, Que., Canada
Department of Animal Science, Michigan State University, East Lansing, MI, USA
Pacific-Yukon Region, Canadian Wildlife Service, Environment Canada, Delta, BC, Canada
Department of Plant and Animal Sciences and Canadian Centre for Fur Animal Research (CCFAR), Nova Scotia Agricultural College, Truro, NS, Canada
School of Dietetics and Human Nutrition, McGill University, Ste. Anne de Bellevue, Que., Canada
Community Health Program, University of Northern British Columbia, Prince George, BC, Canada
Received 20 January 2006; received in revised form 29 March 2006; accepted 3 April 2006
Available online 23 May 2006
The concept of ‘‘sentinel species’’ is important in the environmental health sciences because sentinel species can provide integrated and
relevant information on the types, amounts, availability, and effects of environmental contaminants. Here we discuss the use of mink
(Mustela vison) as a sentinel organism by reviewing the pertinent literature from exposure- and effects-based studies. The review focuses
on mercury (Hg) and polychlorinated biphenyls (PCBs), as they are persistent, ubiquitous, and bioaccumulative contaminants of concern
to both humans and wildlife. Mink are widely distributed, abundant, and regularly trapped in temperate, aquatic ecosystems, and this
makes them an excellent model to address issues in environmental pollution on both temporal and spatial scales. As a high-trophic-level,
piscivorous mammal, mink can bioaccumulate appreciable concentrations of certain pollutants and have been shown to be sensitive to
their toxic effects. The husbandry and life history of mink are well understood, and this has permitted controlled dosing experiments to
be conducted using animals reared in captivity. These manipulative studies have yielded important quantitative information on
exposure–response relationships and benchmarks of adverse health effects, and have also allowed the cellular mechanisms underlying
toxic effects to be explored. Furthermore, the data accrued from the laboratory continue to validate observations made in the field.
Research derived from mink can bridge and integrate multiple disciplines, and the information collected from this species has allowed
environmental health scientists to better understand and characterize pollution effects on ecosystems.
r 2006 Elsevier Inc. All rights reserved.
Keywords: Mink; Sentinel species; Mercury; Polychlorinated biphenyls; Wildlife
1. Introduction
The protection of environmental and human health from
pollutants requires an assessment strategy that can relate
the exposure and bioaccumulation of toxicants to possible
biological effects. The status quo approach in environ-
mental assessments has been to measure the concentrations
of pollutants in key ecological compartments (e.g., air,
water, or biota) and then make toxicological judgments
based on the known or suspected health effects according
to various bench marks, such as no observed adverse effects
level (NOAEL), lowest observed adverse effects level
(LOAEL), or tissue residue value (TRV). Certain wild
animals represent excellent models, or sentinels, to address
issues related to environmental pollution, as they can
provide integrative data on both exposure (i.e., informa-
tion on the type, amount, and availability of contaminants)
and effect (i.e., information on sublethal and clinical health
responses). This has been recognized and discussed
0013-9351/$ - see front matter r 2006 Elsevier Inc. All rights reserved.
Corresponding author. National Wildlife Research Center, Canadian
Wildlife Service, Environment Canada, Ottawa, Ont., Canada.
Fax: +1 613 998 0458.
E-mail address: (N. Basu).
Page 2
Author's personal copy
previously (Beeby, 2001; Fox, 2001; LeBlanc and Bain,
1997; O’Brien et al., 1993; Van der Schalie et al., 1999),
including a report by the US National Research Council’s
Committee on Animals as Monitors of Environmental
Hazards that defined a sentinel species as an ‘‘animal
system to identify potential health hazards to other anima ls
or humans’’ (National Research Council, 1991). Clearly,
sentinel species are of great ben efit and importance for
studies of environmental health.
For an animal to be considered a key sentinel species, it
must satisfy certain requirements (Table 1). Many animal
groups have been hailed as valuable sentinels, including
birds (Furness and Nettleship, 1991; George, 1999), marine
mammals (Colborn and Smolen, 1996; Wells et al., 2004),
aquatic organisms (O’Conner, 2002; Zelikoff, 1998), and
domestic pets (Bukowski and Wartenberg, 1997; National
Research Council, 1991). Terrestrial mammalian wildlife,
in particular, possess multiple characteristics that are
favorable to their inclusion in studies concerning en viron-
mental and human health (O’Brien et al., 1993). Mamma-
lian wildli fe have phy siological systems that are similar to
those of humans and rodents in mediating the uptake,
distribution, meta bolism, and elimination of toxicants.
Furthermore, humans and many species of mammalian
wildlife inhabit similar ecosystems and are exposed to
common climates, food sources, and pollutants. Free-
ranging wildlife can integrate ecological factors and real-
world complexities (e.g., stresses associated with disease,
human disturbance, or temperature) that are not attainable
in controlled laboratory bioassays. As such, the toxicolo-
gical information derived from mammalian wildlife may
have more applicability to hum ans than data collected
from nonmammalian models, such as fish, reptiles, and
Mink (Mustela vison), a member of the weasel family, are
abundant in temperate aquatic ecosystems and are easily
reared in captivity (Lariviere, 1999). As fish-eating
mammals that occupy a high trophic status, mink can
bioaccumulate many notable pollutants. Mink satisfy all of
the criteria presented in Table 1, and their utility as a
sentinel organism has been recognized by many organiza-
tions, including Environment Canada, the United States
Environmental Protection Agency (USEPA), the United
States National Academy of Sciences, and the Swedish
Environmental Protection Agency. While descriptive and
comprehensive publications are available concerning the
toxic effects of pollutan ts on mink (Aulerich and Bursian,
1996; Calabrese et al., 1992; Wren, 1991), the goal of this
paper is to expand upon these reviews by discussing the
advantages of mink as a sentinel species in the field of
environmental health. To achieve this, we provide the
relevant biologic characteristics of mink, summarize the
key results from ecotoxicological studi es in the field and
laboratory, and discuss the importance of this species as an
integrative tool for environmental healt h asses sments. We
conclude by providing a series of recommendations for
future areas of research to improve the utility of mink as a
key sentinel species for studies in the environmen tal health
sciences. The present review will focus primarily on the
mercury (Hg) and polychlorinated biphenyl (PCB) litera-
ture, as these persistent, ubiquitous, and bio accumulative
chemicals are of concern to both human and environ-
mental health.
2. Biological factors
Mink have been raised in captivity since 1866, and the
accumulation of pertinent biological information has been
encouraged by the economic importance of mink to the
global fur industry, which is valued at greater than $11
billion USD (IFTF, 2003). A vast body of information
exists on the natural and life history of mink (Aulerich et
al., 1999 ; Eagle and Whitman, 1987; Hunter and Lemieux,
1996; Lariviere, 1999), including an extensive encyclopedia
compiled by Sundqvist (1989) containing nearly 7000
references. The review by Calabrese et al. (1992) is
noteworthy for its comparisons of nutrition, disease,
physiology and biochemistry between mink and humans.
The taxonomic classification of the American mink is:
phylum (Chordata), class (Mammalia), order (Carnivora),
family (Mustelidae), genus (Mustela), and species (vison
(Lariviere, 1999). Other members of the Mustela genus
include the weasel (Mustela nivalis), ferret (Mustela
putorius), and European mink (Mustela lutreola).
2.1. Distribution
The mink is one of the most widespread carnivores in
North America and is generally found throughout forested
regions across the continent (except for the high Arctic and
arid regions of the southern USA), especially those
containing wetlands (Arnold and Fritzell, 1990)(Fig. 1).
They have been introduced across Europe as an animal to
support local fur industries because they can readily adapt
to new habitats (Lariviere, 1999). Mink have also been
brought to South America, but their current distribution
there is unknown (Medina, 1997). Their occurrence across
wide geographical areas ensures their presence in both
polluted and nonpolluted regions, thus permitting eco-
epidemiological studies. There are no reliable estimates of
Table 1
Characteristics of a sentinel species (adapted from Beeby, 2001; Fox, 2001;
LeBlanc and Bain, 1997; National Research Council, 1991; Van der
Schalie et al., 1999)
Widespread distribution
High trophic status
Ability to bioaccumulate pollutants
Maintained and studied in captivity
Captured in sufficient numbers
Restricted home range
Well-known biology
Sensitive to pollutants
N. Basu et al. / Environmental Research 103 (2007) 130–144 131
Page 3
Author's personal copy
mink population numbers and most predictions are based
on the analysis of fur harvest records (Eagle and Whitman,
1987). It has been estimated that approximately 10,000,000
mink inhabit North America (EPA, 1997), and given that
400,000–700,000 mink are trapped annually (Eagle and
Whitman, 1987), a large sample size can be obtained for
biomonitoring purposes.
Mink are solitary, yet active mammals. Their linear
home range is between 1 and 5 km, and population
densities range between 0.1 and 0.7 animals per square
kilometer (Lariviere, 1999). Adults and males tend to have
larger home ranges than juveniles and females, respectively
(Gerell, 1970). Their home ranges are variable and can be
influenced by multiple factors, such as habitat quality, food
supply, and season (Eagle and Whitman, 1987).
2.2. Growth and reproduction
Mink are monoestrous and usually mate between
February and April (Hannson, 1947). The gestation period
varies from 40 to 75 days because mink, similarly to rabbits
and cats, are induced ovulators and the developing
blastocytes remain in the uterus for many days before
implantation. Offspring are produced in late April or early
May, and the average litter size is four kits. Kits are
weaned at 6–8 weeks of age, and both males and females
reach sexual maturity during their first year. Males are
generally larger than females in size (10%) and mass (50%)
(Hall and Raymon d, 1981). Life span is approximately 3 yr
in the wild, but in captivity mink can live up to 8 yr. This
longevity ensures that they can bioaccumulate pollutants to
appreciable levels.
2.3. Feed and ingestion
Mink are monogastric animals with a digestive tract
more similar to that of humans than that of rodents,
as there is no cecum at the junction of the small and
large intestine (Calabrese et al., 1992 ). Daily requirements
are 140–200 g/kg body weight (b.w.)/d of feed and
75–100 ml of water per day (Aulerich et al., 1999), and
their nutritional requirements have been reviewed by the
US National Research Council’s Subcommittee on
Furbearer Nutrition (National Research Council, 1982)
and others (Aulerich et al., 1999; Rouvinen-Watt
et al., 2005). Optimal diets consist of 18–30% fat,
25–40% protein, 20–50% carbohydrates, and 6–12% ash
(National Research Council, 1982). Females need slightly
more energy than males during the growth period (i.e.,
weaning to maturity), but this dimorphic difference does
not exist in mature animals, as both sexes require
approximately 140 kcal metabolic energy/kg b.w./d (Na-
tional Research Council, 1982). It should be noted that
mink require more energy, on a daily per-kg basis, than
Studies on the feedi ng habits of mink are based on the
analyses of gastrointestinal tracts from trapped animals. In
the wild, mink are opportunistic predators that consume a
range of prey items available in their local habitat,
including small mammals, frogs, snakes, and birds (Wise
et al., 1981). As mink typically forage in close proximity to
aquatic habitats, fish account for approximately 50% of
their diet and represent the primary route by which
persistent chemicals, such as Hg and polychl orinated
biphenyls (PCBs), are accumulated (Chan et al., 2003;
Wiener et al., 2003).
Fig. 1. Global distribution of Mustela vison as indicated by shaded regions (adapted from Lariviere, 1999).
N. Basu et al. / Environmental Research 103 (2007) 130–144132
Page 4
Author's personal copy
3. Field studies
For over 300 yr, wild mink have been trapped for their
fur, usually during the winter months when pelt quality is
highest. Trapping activities are regulated by governmental
agencies (e.g., the Nova Scotia Department of Natural
Resources) and supported by private organizations (e.g.,
the Ontario Fur Managers Federation) (IFTF, 2003). Since
carcasses are discar ded following skinning, research groups
often collaborate with local trappers to obtain biological
samples. These partnerships represent a valuable approach
to obtaining a steady supply of tissues. Besides providing
samples, trappers can offer important anecdotal informa-
tion regarding population trends and individual behavior
(Baker, 2000; Fimreite and Reynolds, 1973; Wren, 1985).
Trappers are generally eager to collaborate with research-
ers, as they are interested in the biology of the animal and
want to ensure that populations remain stable for both
ecological and economic reasons.
3.1. Exposure assessment in the field
Mink can provide relatively steady-state information on
the types, bioavailability, and concentrations of pollutants
in specific regions owing to their high troph ic status and
limited home range. It is not surprising that pollutants
measured in mink tissues reflect levels in the local prey base
(Cumbie, 1975 ; Foley et al., 1 988; Kucera, 1983; Wren et
al., 1986), and tissue burdens of Hg and PCBs can be
upwards of 30 times greater than levels measured in their
food source (Hornshaw et al., 1983; Kucera, 1983).
Because many persistent chemicals can cause adverse
effects in target organs in the part per billion and part
per million ranges, their bioaccumulation is of concern.
For example, Hg concentrations in the tissues of wild mink
approach values associated with toxicity (EPA, 1997).
Laboratory trials have demonstrated that neuronal lesions,
behavioral deficits, and even death are likely outcomes
when concentrations of Hg in the brain are between 4 and
5 mg/g (wet weight (w.w.); Aulerich et al., 1974; Wobeser et
al., 1976a; Wren et al., 1987a, b). Furthermore, significant
changes in neurochemistry may be apparent for concentra-
tions of Hg in the brain as low as 1 mg/g w.w. (Basu et al.,
2006a). A review of field studies over the past 35 yr shows
that Hg burdens in the brains of wild mink are often within
one order of magnitude of levels known to impair animal
health (Fig. 2). Because 35% of American mink popula-
tions reside in regions that have been classified with high
Hg deposition rates (EPA, 1997), the subclinical and subtle
health effects associated with chronic Hg ingestion need to
be addressed.
As with Hg, the concentrations of PCBs in the tissues of
wild mink can equal, or exceed, values known to cause
reproductive impairment or lesions (Fig. 3). The threshold
concentration of hepatic PCBs associated with ad verse
reproductive effects has been estimated as approximately
3.1 mg/g (w.w.), 8.7 mg/g (lipid weight (l.w.)), or 60 pg /g
TEQ (toxic equivalents to 2,3,7,8-tetrachlorodibenzo-p-
dioxin) following the results of multiple feeding trials
(Bursian et al., 2006a; Halbrook et al., 1999 ; Heaton et al.,
1995b; Tillitt et al., 1996). Additionally, subclinical effects
have been observed at lower hepatic PCB co ncentrations,
including proliferation of squamous cells (1.7 mg/g w.w.;
Bursian et al., 2006b), alterations in thyroid hormone levels
(1.6 mg/g w.w.; Restum et al., 1998), decreased hepatic
estrogen receptors (0.89 mg/g w.w.; Shipp et al., 1998a), and
induction of hepatic CYP1A1 activity (0.17–0.64 mg/g w.w.;
Shipp et al., 1998b). An analysis of PCB concentrations in
the livers of wild mink across North America illustrates
that many populations have tissue burdens associated with
adverse reproductive effects (Fig. 3).
Mink can also provide data on spatial and temporal
trends in environmental pollution because they are
abundant, frequently trapped, and found over a wide
geographical area. For example, assessment of Hg con-
centrations in livers (Evans et al., 2000; Gamberg et al.,
2005; Klenavic, 2004; Poole et al., 1998; Yates et al., 2005)
and brains (Basu et al., 2005a; Yates et al., 2005) of wild
mink collected across Canada show that tissue residues
increase in an easterly manner, with highest burdens in
animals inhabiting Atlantic Canada. While a complete
assessment is not possible for PCBs since residue data is
not available from numerous regions, mink residing in
eastern North America generally have the highest concen-
trations of hepatic PCBs on the continent (Fig. 3). Trends
can also be inferred at the regional level (Foley et al., 1988;
Fortin et al., 2001; Haffner et al., 1998; Klenavic, 2004;
Martin et al., 2006a; Poole et al., 1998; Yates et al., 2005).
For example, temporal analyses of the data have shown
that Hg concentrations in mink tissues from Ontario have
not changed over the past 20 years (Basu et al., 2005a;
Evans et al., 2000; Wren et al., 1986), but have decreased
by ap proximately 37% in the state of New York between
the periods 1982–1984 and 1998–2000 (Yates et al., 2005).
Klenavic (2004) compared tissue Hg burdens in mink
inhabiting coasta l and inland regions of Nova Scotia and
found that hepatic concentrations were 200% higher in
the inland group (6.1 vs. 18.4 mg/g dry weight).
3.2. Effects assessment in the field
Ecotoxicological risk assessments are dependent upon
information from both exposure and effects studies
(National Research Council, 1991; Van der Schalie et al.,
1999). While exposure data are available for wild mink
(Section 3.1; Figs. 2 and 3), much less is known about the
associated biological effects in free- ranging populations.
For example, of 13 publications on Hg exposure in wild
mink (listed in Fig. 2), only one study (Basu et al., 2005a)
examined parameters potentially related to health.
Despite the fact that most field studies focus on
exposure, there is evidence that wild mink are highly
susceptible to many pollutants (Aulerich and Bursian,
1996; Calabrese et al., 1992). During the 1960s, fur
N. Basu et al. / Environmental Research 103 (2007) 130–144 133
Page 5
Author's personal copy
ranchers observed that mink consuming fish collected from
the Great Lakes failed to reproduce (Aulerich et al., 1971),
and subsequent feeding trials implicated contaminants
present in the tissues of fish collected from the region as
the root cause. During the 1970s it was observed that wild
mink inhabiting areas in close proximity to industrial point
sources (e.g., pulp and paper mills, chlor-alkali plants)
suffered from intoxications (Fimreite and Reynolds, 1973;
Wobeser and Swift, 1976).
There is some evidence that pollutants can affect mink at
the population level. However, conclusions from these
studies should be drawn cautiously, as the strength of
association between exposure to pollutants and changes in
population dynamics is generally weak because many other
factors, such as habitat loss, disease, and natural cycles, are
involved (Fox, 2001 ; Wren, 1991). Estimates of population
trends are usually deduced from the harvest yield of
trappers, and historical declines in mink populations have
been reported from Ontario, Ohio (Wren, 1991), Oregon
(Henny et al., 1981), Georgi a, North Carolina, and South
Carolina (Osowski et al., 1995), and France (Lode
, 2002).
The presence of pollutants in these regions has been
implicated as a possible factor in decreased harvest returns.
At the whole-animal level there is some evidence that
pollutants can affect physiological systems in wild mink. A
survey of mink inhabiting the PCB-contaminated Kalama-
zoo River Superfund site (Michigan, USA) revealed a
positive correlation between concentrations of hepatic
PCBs and the presence of a jaw lesion (proliferation of
mandibular and maxillary squamous epithelium), with
44% of mink trapped in the Superfund site having this
particular phenotype (Beckett et al., 2005). These findings
were verified by laboratory experiments in which exposure
of mink to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
(Render et al., 2000a), PCB 126 (Render et al., 20 00b), and
PCB-contaminated fish (Bursian et al., 2006b) resulted
in the same lesion. In another correlative study, Harding
et al. (1999) reported a negative relationship between
Lab Study
(Outcome, Ref #)
Brain H
0.001 0.01 0.1 1 10 100
Field Study Location
(Year, Ref #)
Normal (stuvw)
Biochemical (t)
Pathology (uv)
Behavioural (uv)
Death (uvwx)
Maine (2003,a)
Yukon (2002,b)
Nova Scotia (2002,c)
Ontario (2002,c)
Nova Scotia (2001,a)
Ontario (2000,d)
New Hampshire (2000,a)
Quebec (1999,e)
Tennessee (1993,f)
Georgia (1990,j)
Connecticut (1988,k)
Ontario (1986,l)
E. New York (1983,m)
W. New York (1983,m)
N. Quebec (1994,g)
Northwest Territories (1992,h)
Minnesota (1991,i)
South Carolina (1990,j)
North Carolina (1990,j)
Manitoba (1983,n)
Manitoba (1983,o)
NE United States (1980,p)
Wisconsin (1975,q)
Georgia (1973,r)
Fig. 2. Brain Hg concentrations in wild mink (gray bars) and laboratory-exposed mink (black bars). The lowest observable effects level (LOEL, 5 mg/g
brain Hg wet weight) and 10% LOEL are indicated. Letters refer to the following studies: a—Yates et al. (2005);bBasu et al. (2005a);cKlenavic
(2004);dEvans et al. (2000);eLangis et al. (1999)
;fStevens et al. (1997)
;gFortin et al. (2001)
;hPoole et al. (1998)
*; i—Ensor et al. (1993)
j—Osowski et al. (1995)^; k—Major and Carr (1991)
;lWren et al. (1986);mFoley et al. (1988)
;nKucera (1983);oManitoba Summary Report
;pO’Connor and Nielsen (1981)
;qSheffy and St. Amant (1982)*; r—Cumbie (1975)
;sWobeser et al. (1976b);tBasu et al. (2006a);u
Wobeser et al. (1976a);vWobeser and Swift (1976);wWren et al. (1987a);xAulerich et al. (1974). Brain concentrations for some studies (as
indicated by symbols) were estimated from published concentrations in the liver (
), fur (
), and kidney (
) based on interorgan relationships of Hg
provided by Kucera (1983), Wren et al. (1986), Evans et al. (2000), and Klenavic (2004). For studies that only listed a mean Hg value, the range was
calculated as two standard deviations and indicated by an asterix (*).
N. Basu et al. / Environmental Research 103 (2007) 130–144134
Page 6
Author's personal copy
concentrations of hepatic PCBs and baculum length in
juvenile mink collected from British Columbia, though a
much larger study reported no significant associations
between PCB exposure and baculum size or weight in 160
juvenile mink collected from British Columbia and
Ontario, Canada (Elliott et al., 2004). Furtherm ore, no
relationships between PCB expo sure and baculum size were
found in a laboratory experiment ( Aulerich et al., 2000)
and a subsequent field study (Millsap et al., 2004).
Additional research is recommended because a significant
relationship was found between baculum size and DDE
exposure in adult mink (Elliott et al., 2004). Neuronal
lesions in the cerebellum and occipital cortex related to Hg
toxicity have been measured in wild mink (Wobeser and
Swift, 1976), but it should be noted that pollutants may not
always be the causative factor for observed pathologies in
wild mink populations (Friedman et al., 1977). Despite the
above-mentioned reports, it is apparent that information
concerning toxicant-induced physiological or whole-animal
effects in wild mink is generally sparse, and when effects
are reported, many of the studies are hampered by low
sample sizes and inadequate statistical power.
Field and ecological studies are increasingly exploring
the effects of pollutants at the molecular and cellular levels,
and recent research reveal that neurochemicals (Stamler et
al., 2005), DNA (Smith and Burgoyne, 2004), and mRNA
(Doyon et al., 2003) can be assayed from carcasses of free-
living animals. Research on trapped wild mink demon-
strated a significant correlation between the concentrations
of brain Hg and levels of neurochemical receptors (i.e.,
muscarinic cholinergic and dopamine-2 receptors) (Basu et
al., 2005a). Some of these findings have been verified in the
laboratory (Basu et al., 2006a), and taken together with
research on river otters (Basu et al., 2005b, 2006b), they are
the first to demonstrate that exposure to ecologically
relevant levels of Hg can be associated with significant
changes in brain chemistry in piscivorous mammals.
4. Laboratory studies
The ability to raise mink in captivity makes them a
particularly useful model in toxicology, as quantitative
exposure–response relationshi ps can be derived. These data
can stre ngthen environmental risk characterization, as
toxicity data can be compared between wild and captive
animals. In this regard, mink are an excellent sentinel
because such an approach is not available for endangered
species, animals not easil y raised in captivity, or humans.
Laboratory bioassays using ranch mink were initiated in
the mid- to late 1960s following concern regarding
chemical pollution of the Great Lakes region (Hartsough,
1965). One of the earliest controlled dosing studies
demonstrated that dietary exposure of neonatal mink to
pesticide-contaminated fish collected from the Miramichi
River (New Brunswick, Canada) caused physiological
changes (e.g., blood chemistry, tissue weights) and even
death (Gilbert, 1969). Many subsequent trials were
completed at the Michigan State University Experimental
Fur Farm under the direction of Dr. Richard J. Aulerich
between the 1970s and 19 90s. Typically, studies were
0 7.0 µg /g
0 0.1 µg /g Harding
0 0.1 µg /g
0 0.1 µg /g
0 −3 .5 µg /g
0.8 −16 .6 µg /g l.w.*
0.1 − 6.0 µg /g
1.6 − 6.0 µg /g l.w.
2.4 − 11.5 µg /g l.w.
0 − 12.8 µg /g
0.6 − 2.4 µg /g
0.9 − 10.0 µg /g
0 − 7.9 µg /g
0 − 2.8 µg /g l.w.
0.8 − 117.7 µg /g l.w.*
3.4 − 171 pg /g TEQ
0.4 − 4.1 µg /g
Poole et al. (1998)
et al. (1999)
Elliott et al. (1999)
Henny et al. (1980)
Ensor et al. (1993)
Millsap et al. (2004)
Tansy et al. (2003)
Osowski et al. (1995)
Tansy et al. (2003)
O' Shea et al. (1981)
Major & Carr (1991)
Major & Carr (1991)
Foley et al. (1988)
Proulx et al. (1987)
Haffner et al. (1988)
Champoux (1996)
Martin et al. (2006a)
Fig. 3. Heptic PCB concentrations (wet weight) in wild mink collected across North America. Boxes that are shaded indicate populations in which hepatic
PCB concentrations equal, or exceed, levels known to cause reproductive health effects. l.w. refers to concentrations expressed on a lipid weight basis, and
* refers to whole body analyses. (Champoux, 1996; O’Shea et al., 1981; Proulx et al., 1987; Tansy et al., 2003).
N. Basu et al. / Environmental Research 103 (2007) 130–144 135
Page 7
Author's personal copy
conducted to address the reproductive and lethal effects of
toxicants of concern (i.e., pesticides, organ ochlorines, and
heavy metals) in kits and adults (reviewed by Aulerich and
Bursian, 1996; Calabrese et al., 1992). The findings from
these experiments have been invaluable to the field of
environmental toxicology and risk assessment, and mink
continue to play an important role in this area of research
as newly emerging chemicals (e.g., polybrominated diphe-
nyl ethers and perfluorooctane sulfonate) and research
paradigms (e.g., antagonist effects of selenium on Hg
toxicity, genomic technologies) are starting to be explored
in this species.
4.1. Organochlorines
TCDD is one of the most toxic chemi cals found in
nature. The calculated LC
for mink is 4.2 mg TCDD/
kg b.w. following a 28-d exposure (Hochstein et al., 1988).
Furthermore, intraperitoneal injection of mink kits with
0.1 mg TCDD/kg b.w./d caused significant reductions in
body weight within 2–3 weeks of exposure, and all anima ls
dosed with 1 mg TCDD/kg b.w./d died within 2 weeks of
exposure (Aulerich et al., 1988). In another study, exposure
of mature female mink to 1.4 mg/kg TCDD for nearly 4.5
months resulted in a range of significant changes to blood
chemistry (e.g., white blood cell counts, serum iron,
phosphorus, albumin, total CO
, cholesterol, osmolality,
and alanine aminotransaminase activity) and to tissue
function (e.g., deformed toenails, ascite s, gastric ulcers,
intestinal hemorr hages, and depletion of fat reserves)
(Hochstein et al., 2001). Collectively, these toxicity da ta
rank mink as one of the most sensitive animals to TCDD
poisoning among those tested (species name is followed by
30-day adult LC
): Syrian hamster (41571 mg/kg b.w.),
New Zealand rabbit (115 mg/kg b.w.), Wistar rat (110 mg/
kg b.w.), and Rhesus macaque (50 mg/kg b.w.) (reviewed by
Geyer et al., 1997).
Of all the toxicants screened in mink, perhaps none have
received as much attention as PCBs. PCBs are complex
mixtures of 209 different congeners that were primarily
used as coolants and lubricants in electrical capacitors and
transformers prior to being banned by most developed
nations (ATSDR, 2000). Because they bioaccumulate in
the food chain and resist degradation, PCB contamination
is ubiquitous in the environment. PCBs were believed to be
one of the causative agents for the reproductive impair-
ment observed in mink that consumed fish from the Great
Lakes (Aulerich and Ringer, 1977; Aulerich et al., 1971,
1973; Ringer et al., 1972). Subsequent laboratory trials
confirmed these hypotheses, as short-term bioass ays (28–35
days) established a dietary LC
range between 47 and
84 mg/g Aroclor 1254 (Aulerich et al., 1986; Hornshaw et
al., 1986). Studies on ecologically relevant concentrations
(i.e., 2 mg/g Aroclor 1254, a commercial PCB mixture,
incorporated into fish) revealed that intake could result in a
series of adverse health outcomes, including reduced
growth, anorexia, deform ed nails, increased liver and
kidney weights, and reproductive impairment (Aulerich
and Ringer, 1977). Further research has shown that dietary
exposure to commercial mixtures of PCBs (0.1 mg Clophen
A50/animal/d, Brunstro
m et al., 2001; 0.64 mg/g Aroclor
1254, Platonow and Karstad, 1973) and PCB-contami-
nated fish (3.7 mg/g, Hornshaw et al., 1983; 0.25 mg/g,
Restum et al., 1998; 0.72 mg/g, Heaton et al., 1995a, and
Tillitt et al., 1996; 1.5 mg/g, Bur sian et al., 2006a) can
impair reproductive performance by, for example, reducing
kit survivability, litter size, and whelping success. These
reproductive effects may partly be due to the degeneration
of the placenta and trophoblast (Ba
cklin et al., 1998; Jones
et al., 1997).
Recently, experiments have been completed to better
understand the mechanisms underlying PCB toxicity in
mink. Characteristic lesions of the mandible and maxilla in
mink exposed to organochlorines result in loosened teeth
and swollen jaws. Laborator y studies linked this phenom-
enon to PCB-induced (Bursian et al., 2006b; Render et al.,
2000b) and TCDD-induced (Render et al., 2000a) osteoin-
vasion of squamous epithelium. The liver is known to be a
major site of PCB toxicity (ATSDR, 2000), and this is
supported by histological examinations on hepatic tissues
from PCB-intoxicated mink (Bergman et al., 1992; Restum
et al., 1998). Effects include iron deposition in Kupffer
cells, lipidosis, and increased frequency of polymorpho-
nuclear and mononuclear cells. Multiple studies have also
shown that PCBs induce hepatic cytochr ome P450 activity
in a concentration-dependent manner (Brunstro
m, 1992;
Shipp et al., 1998b), and this induction is transient with
enzyme activity returning to baseline levels when PCB-
exposed animals were restored to a clean diet (Shipp et al.,
PCBs may also affect the immune and endocrine
systems. Studies on immature females demonstrated that
a single intraperitoneal injection of e ither 3,3
hexachlorobiphenyl (a coplanar PCB; 0.4 mg/kg b.w.) or
-hexachlorobiphenyl (a noncoplanar PCB;
20 mg/kg b.w.) antagonized the 17b-estradiol-stimulat ed
increase of nuclear estrogen receptors in the uterus
(Patnode and Curtis, 1994). These authors also observed
significant increases in the levels of uterine progesterone
receptors in pregnant mink 14 days after a single
intraperitoneal injection of either congener at the indicated
dose. Shipp et al. (1998a) reported that effects on the
estrogen receptor were localized to the liver (i.e., negative
correlation), but no changes were measured in the number
of estrogen or pr ogesterone receptors in the uterus of mink
exposed to PCB-contaminated carp for 18 months. Similar
tissue-specific responses were reported by Ba
cklin et al.
(1998), who observed a negative correlation between PCB
exposure (0.65 and 1.3 mg Clophen A50/d) and insulin-like
growth factor (IG F II) gene in the adult liver, but not in the
placenta or fetus. Exposure of 6-month-old mink to 1 mg
PCBs (as Aroclor 1242) per day resulted in the replacement
of membrane polyunsaturated fatty acids with monounsa-
turated and saturated fatty acids in the mesenteric lymph
N. Basu et al. / Environmental Research 103 (2007) 130–144136
Page 8
Author's personal copy
nodes surrounding the adipose tissue (Ka
and Hyva
inen, 1999).
Some of the PCB-induced effects on animal development
may be related to alterations in the status of vitamin A or
thyroid hormones. Ka
et al. (1999) found that PCBs
can reduce the concentrations of hepatic vitamin A2 and
plasma vitamins A1 and A2. Ha
kansson et al. (1992) found
that reductions in vitamin A were localized to hepatic and
pulmonary tissues, and not renal tissues. However, Martin
et al. (2006b) found reductions of retinol in the kidney and
plasma of mink exposed to a diet of 30% PCB-
contaminated carp (1.69 ppm total PCB or 73.2 pg/g
TEQ). Martin et al. (2006b) also found that plasma total
triiodothyronine (T3) was lower in males exposed to the
30% carp diet, but plasma total T4 and free thyroxine (T4)
were higher in kits fed a 10% carp diet. Other studies have
reported decreased concentrations of T3 (Aulerich et al.,
1987; Heaton et al., 1995b), and increased (Heaton et al.,
1995b; Restum et al., 1998) and decreased (Aulerich et al.,
1987) concentrations of T4 in mink following PCB
exposure. While PCBs appear to impair thyroid hormone
homeostasis, pathological damage of the thyroid glands
has not been found (Aulerich et al., 1987; Martin et al.,
The toxicity of individual PCB congeners is directly
influenced by the locat ion and number of chlorine atoms
on the biphenyl rings. Ingestion of non- an d mono-ortho-
chlorinated PCB congeners by mink resulted in the
induction of hepatic EROD activity and reproductive
impairment (i.e., reduced kit survival) (Brunstro
m et al.,
2001). However, such effects were not observed in animals
fed congeners with tw o to four chlorines in the ortho
position. In a comparative study, Aulerich and Ringer
(1977) demonstrated that the toxicity of PCB mixtures can
be related to the degree of chlorine substitution, as Aroclor
1254 caused greater reproductive impairment than Aroclor
1016, 1221, or 1242 when each mixture was tested at a
dietary level of 2 mg/g for approximately 11 months.
Exposure of female mink to 20 mg/g of Aroclor 1016 and
1242 resulted in 25% and 100% mortality, respectively,
during a chronic feeding trial (Bleavins et al., 1980).
Further studies on the same mixtures showed that hepatic
EROD activity was induced by Aroclor 1242 but not by
Aroclor 1016 (Shull et al., 1982).
4.2. Mercury
Mercury is a naturally occurring heavy metal predomi-
nantly released into the environment as a byproduct of
fossil fuel combustion, mining, smelting, and waste
incineration (ATSDR, 1999). Methyl Hg (MeHg) is the
most relevant form of Hg to which mink are exposed
because more than 95% of the Hg in fish is in the organic
form (Bloom, 1992). MeHg can effectively biomagnify
through the food chain and accumulate in high trophic
wildlife, unlike inorganic Hg, which has a limited capacity
to penetrate biological membranes and is easily excreted by
the body (Chan et al., 2003; Wiener et al., 2003). As a
result, the chemical composition of Hg will influence its
toxicity, as demonstrated in an early dosing study on mink
in which dietary exposure to 5 mg/g MeHg caused death,
but 10 mg/g inorganic Hg (HgCl
) did not evoke any clinical
effects (Aulerich et al., 1974). While mink exposed to
accumulated this metal only in their kidneys, those
exposed to MeHg had significant increases in their liver,
kidney, muscle, spleen, brain, lung, and heart tissues
(Aulerich et al., 1974). Mink, like other mammals, can
demethylate MeHg to the mercuric ion (Hg
), but this
process is limited in the brain, spleen, and muscle when
compared to the liver and kidney (Jernelov et al., 1976).
Within the must elid family, species differences in the
capacity to demet hylate MeHg are apparent in the brain, as
MeHg constituted 88.2% of total Hg in wild mink (Basu et
al., 2005a), and only 73.5% in wild river otters (Basu et al.,
The margin by which MeHg induces toxicity is quite
narrow, as animals ingesting 0.5 mg/g Hg were clinically
normal (Kirk, 1971; Wobeser et al., 1976b), whereas 1 mg/g
Hg caused mortality in adult mink (Dansereau et al., 1999;
Wobeser et al., 1976a; Wren et al., 1987a). Furthermore,
death is imminent once the initial symptoms of Hg
intoxication are observed, and efforts to reverse the toxicity
with chelating agents have proven unsuccess ful (Aulerich et
al., 1974). Within 1 week, incoordination and anorexia can
rapidly progress to ataxia, tremors, convulsions, and high-
pitched vocalizations (Aulerich et al., 1974; Dansereau et
al., 1999; Wobeser et al., 1976a). The progression of Hg
neurotoxicity in mink is quite similar to that documented
in humans and rodents (Watanabe and Satoh, 1996). Even
though Hg has a high affinity for thiol groups and can
disrupt multiple cellular components (Atchison and Hare,
1994), autopsi es on Hg-poisoned humans, rodents, and
wildlife (including wild and captive mink) show that lesions
are present in discrete regions of the brain (i.e., calcarine
region of the occipital cortex and granule layer of the
cerebellum) (Watanabe and Satoh, 1996; Wobeser and
Swift, 1976; Wobeser et al., 1976a).
At ecologically relevant exposure levels, Hg can affect
the reproductive system. Litter size, kit weight, and hepatic
lipid concentrations were all lower in animals exposed to
0.22 mg/g dietary Hg compared to nonexposed controls
(Halbrook et al., 1997), but it should be noted these diets
were also contaminated with organochlorines. Other
studies reported no statistical effects on litter size or kit
growth following maternal exposure to 1 mg/g Hg (Danser-
eau et al., 1999; Wren et al., 1987b). In a two-generation
study, Dansereau et al. (1999) found that a lower
percentage of F1 females gave birth following exposure
to 1 mg/g Hg than controls (33% vs. 94%). Wren et al.
(1987b) also observed a lower whelping percentage in mink
exposed to 1 mg/g Hg, compared to controls.
Based on the laboratory data, a lowest observable
adverse effects level for Hg of 1.1 mg/g in the diet and
5 mg/g (w.w.) in the brain has been established. At lower
N. Basu et al. / Environmental Research 103 (2007) 130–144 137
Page 9
Author's personal copy
exposure levels, neurochemi cal changes are apparent, as
males fed 0.5 mg/g dietary MeHg had higher muscarinic
cholinergic receptors and increa sed cholinesterase activity
in discrete regions of the brain (Basu et al., 2006a). As
discussed earlier, it is not uncommon for wild mink to be
exposed to such levels in nature ( Fig. 2; Kamman et al.,
5. Integrative considerations
Effects on mink from contaminant exposure have been
the subject of several risk assessments over the past decade.
Studies from Ohio, Ontario (Wren, 1991), British Colum-
bia (Elliott et al., 1999; Harding et al., 1999), Michigan
(Giesy et al., 1994 ; Millsap et al., 2004), California (Wolfe
and Norman, 1998), Connecticut, Massachusetts (Bursian
et al., 2006a, b), and Tennessee (Moore et al., 1999; Sample
and Suter, 1999) have utilized this species to address
contaminant issues. Furthermore, many governmental
agencies, including Environment Canada, the USEPA,
the National Institutes of Health, the Swedish Environ-
mental Protection Agency, and the US National Academy
of Sciences, have acknowledged the importance of mink as
a surrogate species.
In their natural environment, mink are potentially
subject to man y confounding risk factors (e.g., extreme
temperature, reduced food availability, and predation) that
cannot be replicated practically under laboratory condi-
tions. Unlike the laboratory, wild mink are not housed in
sterile, temperature-controlled, and disease-free conditions
with continuous access to food, water, and medical
attention. Som e researchers have begun to address these
concerns by incorporating multiple variables into the
design of their laboratory experiments. Wren et al.
(1987a) observed that cold stress and PCBs both poten-
tiated MeHg-related health effects in captive mink. Other
studies have shown that dieldrin exposure can exacerbate
PCB-induced mortality in mink (Aulerich and Ringer,
1977; Ringer et al., 1972). A recent study found that
selenomethionine can mitigate the uptake of MeHg into
tissues (Basu et al., in preparation). By attempting to mimic
ecological conditions in the laboratory with multifactorial
designs, mink bioassays can be developed with the aim of
reducing some of the inherent difficulty when extrapolating
laboratory data to the field.
Mink can also be used in multitiered approaches to
characterize risk, as exposure–response data can be
obtained at the level of cells (in vitro), organisms (whole-
animal bioassays), and field (wild-trapped animals). Such
an approach has been used to address the effects of Hg on
neurochemical systems in mink. While the exact mechan-
isms of Hg-induced alterations of neurobehavior are
unknown, laboratory studies have demonstrated that Hg
can affect cholinergic neurotransmission (Coccini et al.,
2000; Kobayashi et al., 1979). Studies on wild mink
collected from three regions of Canada demonstrated that
animals with high concentrations of brain Hg also had
significantly greater numbers of mu scarinic cholinergic
receptors (Basu et al., 2005a). These findings were verified
with a controlled dosing experiment, as there was no
significant difference in the slopes of the regression plots
relating brain Hg and mACh receptor levels between the
field (y ¼ 118:8 log(x)+629.4; Basu et al., 2005a)and
laboratory (y ¼ 129 :8 log(x)+1507.5; Basu et al., 2006a)
studies. To exp lore the underlying mechanisms, an in vitro
experiment demonstrated that Hg could inhibit the binding
of radioligands to the muscarinic cholinergic receptor
(Basu et al., 2005c). Because Hg can impede cholinergic
neurotransmission at the level of the muscarinic cholinergic
receptor, up-regulation of this receptor in high-Hg-exposed
animals likely represents an adaptive mechanism to ensure
proper signal transduction. Thus, by exploring the effects
of MeHg on cholinergic neurotransmission at multiple
levels of biological organization, the weight of evidence
suggests that Hg can affect the muscarinic cholinergic
receptor not only in the laboratory, but also in the field.
Others have used a similar approach. For example,
researchers at Michigan State University observed that
various compounds and mixtures of organochlorines can
induce jaw lesions in captive mink (Bursian et al., 2006a;
Render et al., 2000a, b), and these findings were verified in
the field (Beckett et al., 2005). The derivation of multiple
lines of evidence is necessary in ecotoxicology and
environmental health studies because causal linkages
between toxicant exposure and biological response are
often difficult to establish (Fox, 2001; Wren, 1991).
Thus, to validate previous laboratory observations that
PCBs affect norepinephrine, dopamine (Aulerich et al.,
1985), vitamin A (Ka
et al., 1999), vitamin E (Ka
et al., 1999), alkaline phosphatase, alani ne aminotransfer-
ase ( Edqvist et al., 1992), T3, and T4 (Aulerich et al.,
1987; Heaton et al., 1995b; Martin et al., 2006b; Restum et
al., 1998), correlative studies from wild animals are
required before cause-and-effect relationships may be
Mink have been used as a surrogate model to address
toxicity issues in endangered species or animals not easily
maintained in captivity. For example, bioassays on mink
were used to address concerns of Hg poisoning in river
otters (Halbrook et al., 1994; Sample and Suter, 1999) and
petroleum exposure in sea otters (Beckett et al., 2002 ;
Mazet et al., 2000; Schwartz et al., 2004). Because they
share common physiological systems and allometries,
mammalian wildlife, laboratory rodents, and even humans
may respond to toxicants in comparable manners (Calabr-
ese et al., 1992). For example, chronic ingestion of PCB-
contaminated Great Lakes fish caused neurological im-
pairments in captive mink (Aulerich et al., 1971) that
resembled effects observed in human children (Jacobson et
al., 1984; Lonky et al., 1996), and rats (Daly, 1993). In
addition, Hg-induced effects on neurochemistry (Basu et
al., 2005c), histology, and behavior (Watanabe and Satoh,
1996; Wobeser et al., 1976a) are remarkably homologous
among humans, rodents, and mink. Thus, mink represent
N. Basu et al. / Environmental Research 103 (2007) 130–144138
Page 10
Author's personal copy
an alternate model to study the etiology of environmental
diseases in mammals.
6. Recommendations
In summary, mink have innate characteristics that
support their utilit y as a key indicator species of ecosystem
health (Table 1). They inhabit a wide geographical range
(Fig. 1) and are found in sufficient numbers to permit
epidemiological investigations as samples can often be
obtained efficiently from trappers. As a high-trop hic-level,
piscivorous mammal with a limited home range, they can
bioaccumulate environm ental pollutants to relatively high
concentrations and can provide an indication regarding
local contamination. Furthermore, they can be housed in
the laboratory for use in routine toxicity screening
To improve their utility as a key sentinel species, we
conclude by discussing future activities that would benefit
and increase the role of mink in environmental health
1. Wild mink are widely distributed across North
America, Europe, and possibly South America, and
carcasses can readily be obtained from trappers. A
regular monitoring program to collect animals from
key regions would provide a strong platform to
characterize pollution trends (i.e., exposure and effects)
on spatial and temporal scales. Evers et al. (1998), for
example, sampled feathers and blood from nearly 500
common loons (Gavia immer) in five regions across
North America between 1991 and 1996 to address Hg
exposure on a continental scale. Their findings showed
that Hg levels increase from the west to the east, and
resemble the USEPA’s modeled prediction of atmo-
spheric Hg deposition. Mink may also be used to infer
historical trends in Hg pollution because they have
been trapped for nearly 300 yr, and the levels of Hg in
the fur (i.e., pelts) are correlated with those in tissues
(Evans et al., 2000). Such an approach using archival
feather samples from two piscivo rous seabirds showed
that Hg concentrations in our environment have
increased approximately five times over the past
150 yr (Monteiro and Furness, 1997).
2. Inconsistencies exist in the manner in which tissue
residues are reported, as some publications express
concentrations on a wet-, dry-, or lipid-weight basis.
This makes it difficult to compare findings, and is
apparent by examining the available literature from the
Hg (Fig. 2) and PCB (Fig. 3) studies. To minimize the
potential effects of varying moisture content resulting
from freezer storage, tissues should be freeze-dried
prior to the analysis of metals and trace elements, and
the moisture content reported. Freeze-dried samples
are easier to store , as they require less space and can be
kept at room temperature. Organochlorine data should
be expressed on both a lipid- and tissue-wei ght basis.
All analytical studies need to include certified standard
reference materials and present the recovery values.
The establishment of universal protocols for sample
collection, storage, and analysis would be beneficial to
the entire wildlife toxicology communi ty and facilitate
better inter-study comparisons. Along these lines,
laboratory feedi ng trials should provide complete
exposure information, including nominal and actual
concentrations of toxicant in feed, daily intakes per
kilogram of body weight, and total daily intakes.
3. Relevant statistical information should be documented
in all publications (e.g. sample size, mean, range, and a
measure of variance), as well as possible transforma-
tions of the data. Studies should include pertinent
demographic information, such as sex and age, and
determine if these cofactors are affecting the data using
proper statistical techniques. The report of Yates et al.
(2005) is exemplary in these regards. There are
inconsistencies in the literature with respect to the
influence of such factors on tissue residue levels as
some studies have reported effects (Kucera, 1983;
Yates et al., 2005) while other have not (Fortin et al.,
2001; Marti n et al., 2006a; Wren et al., 1986). Future
studies should aim to resolve some of these uncertain-
4. Relationships between contaminant exposure and
effects at the physiological or whole-animal level have
seldom been explored. As many studies have employed
commercial trappers, who utilize lethal approaches to
capture animals, it is difficult to obtain blood samples
or perform body condition analysis. Where budgets
permit, live trapping of animals should be considered,
which may also allow repeated-measures studies on
individuals over time. Furthermore, radiotelemetry
provides information on locomotion and habitat
preference, and newly emerging technologies can even
record simple measures of health, such as heart rate
and temperature (Kramer and Kinter, 2003). These
data can later be correlated with contaminant burdens
measured in the same animals. A strong database exists
on the normal values of many physiological parameters
from mink (Aulerich et al., 1999; Brandt, 1989) and
this provides a foundation for assessing animal health
in natural populations.
5. The development of surrogate and nonlethal ap-
proaches to diagnosing pollutant exposure and effects
are gaining interest in the field of environmental health
(Stamler et al., 2005 ). Fur offers a noninvasive method
for assessing Hg bioac cumulation, as correlations have
been calculated between levels of Hg in the fur and
brain (r
¼ 0:46–0.55, Po0:05; Evans et al., 2000;
Yates et al., 2005). Blood can also be obtained from
animals in a humane manner and levels of pollutants in
blood generally reflect recent exposures. Health status
may be inferred by studying certain components within
blood. For example, measurements of platelet mono-
amine oxidase, serum alkaline phosphatase, and
N. Basu et al. / Environmental Research 103 (2007) 130–144 139
Page 11
Author's personal copy
plasma cortisol can afford clues regarding the function
of the nervous, hepatic, and endocrine systems,
respectively. The analysis of fecal DNA has recently
been used to assess population size and structure of
feral river otters (Hung et al., 2004). Mink are a good
model to validate the use of peripheral biomarkers
because blood, fur, feces, and internal organs can be
obtained from animals in the laboratory and field.
6. Molecular and cellular approaches have increased our
knowledge of the mechanisms underlying toxic out-
comes. However, the application of such techniques in
wildlife toxicology is progressing more slowly than in
the biomedical sciences. Historical studies on mink
have provided important toxicity data (i.e., LC
reproductive impairment) on notable pollutants (Au-
lerich and Bursian, 1996), but future research needs to
place greater emphasis on addressing stressor-induced
changes in biochemistry and gene expression that
precede overt toxicity. This is now possible in both
the laboratory and field. Recent studi es have shown
that DNA can be collected from wildlife samples (e.g.,
blood, blood clots, tissues, saliva) using commercially
available FTA databasing paper, and these samples can
be stored at room temperature for years (Smith and
Burgoyne, 2004). Useable quantities of mRNA have
been obtained from road kill and hunter- killed animals
up to 24 h after death (Doyon et al., 2003), and
neurochemical receptors and enzymes have been
assayed from mink carcasses (Stamler et al., 2005).
Genetic analysis of fecal scats was used to estimate the
size and structure of river otter populations from
Kinmen Island, China (Hung et al., 2004). By studying
events at the genetic and cellular level, perhaps
sensitive and specific biomarkers may be discovered
in the laboratory and validated in the field to better
establish cause-and-effect linkages.
7. For the most part, laboratory studies on mink have
focused on singl e-chemical effects, and few have
considered multiple factors. Now that basic quantita-
tive exposure–response data have been colle cted for
most pollutants of concern in mink, future research
endeavors should incorporate ecological fact ors into
their study design by characterizing the effects of, for
example, sex, age, multiple contaminants, temperature,
disease, and predation on health outcomes. In the
laboratory there are reports that toxic responses are
affected by cofactors, and in the field there is some
uncertainty as to the effects of age or gender on Hg
bioaccumulation. Future experiments need to address
these issues.
8. Behavioral tests are a vital component of toxicity
screening programs. There are many reports that
pesticides, organochlorines, and MeHg can impair
animal behavior, but clinical outcomes (e.g., lethargy,
anorexia, and tremors) have largely been reported as
anecdotal observations. Sophisticated tests to monitor
subtle changes in behavior have been developed for
rodents, and some of these have recently been explore d
in mink, including righting ability, open field test, and
tail-pinch response (Bush et al., 2002). Furthermore,
the use of a swimming maze to assess spatial memory
(K.J. Bec kett, Michigan State University, data in
preparation) and a stick test for aggression (K.
Rouvinen-Watt, Nova Scotia Agricultural College,
data in preparation) have also been explored in mink.
The development and standardization of a battery of
behavioral and neurofunctional tests for mink may
allow researchers to better explore the ethology of wild
mink that inhabit ecosystems stressed by environmen-
tal contaminants.
9. The creation of viable cell lines (e.g., hepatic, neuronal,
lymphocyte) from mink is desirable because of the high
costs and ethical constraints of whole animal studies.
Protocols for the creation of cell lines are now well
established in the biomedical community (Davis, 2002)
and can easily be adapted to mink. For example,
Trowbridge et al. (1982) developed an adherent
neuronal cell line from 6-week-old ferret (Mustela
putoris furo) with a cloning effici ency greater than 45%
and a doubling time of approximately 12 h. This
product is commercially available through American
Type Culture Collection (ATCC Catalog ] CRL-1656).
By studying exposure–response relationships in vitro,
specific hypotheses regarding cause-and-effect relation-
ships can be formulated and then validated in whole
animal bioassays and field studies.
10. There are many similarities among the biological
responses of mink, humans, and rodents following
exposures to PCBs and Hg (Aulerich et al., 1971; Basu
et al., 2005c; Daly, 1993; Jacobson et al., 1984; Lonky
et al., 1996), and such observations support the use of
mink as an alternate model in the field of environ-
mental health. To better improve their utility, research
to compare and contrast interspecies differences in
toxicokinetics and toxicodynamics is necessary at both
the whole animal and cellular levels. A recent study, for
example, studied the effects of Hg on muscarinic
receptor binding among humans, rodents, and mink
(Basu et al., 2005c). The authors found that species
sensitivities, irrespective of brain region studied (cortex
or cerebellum) or type of Hg (MeHg or HgCl
), could
be ranked from most to least sensitive as river
otter4rat4mink4mouse 4 humans. Data such as
these provide important information on interspecies
differences and a platform to interpret results between
wildlife and human studies.
No conflict of interest is declared. Funding for this
review was provided by a grant from the Collaborative
Mercury Research Network (COMERN) to AS and HMC.
NB was a recipient of a NSERC Postgraduate Fellowship.
N. Basu et al. / Environmental Research 103 (2007) 130–144140
Page 12
Author's personal copy
The authors thank Kerrie J. Beckett, Christopher J.
Stamler, and Birgit Braune for valuable discussion and
Arnold, T.W., Fritzell, E.K., 1990. Habitat use by male mink in relation to
wetland characteristics and avian prey abundances. Can. J. Zool. 68,
Atchison, W.D., Hare, M.F., 1994. Mechanisms of methylmercury-
induced neurotoxicity. FASEB J. 8, 622–629.
ATSDR, 1999. Toxicological