Animal Conservation (2003) 6, 11–18 © 2003 The Zoological Society of London
DOI:10.1017/S1367943003003032 Printed in the United Kingdom
Impacts of varying habitat quality on the physiological stress of
spotted salamanders (Ambystoma maculatum)
Rebecca Newcomb Homan1, Jonathan V. Regosin1, Daniel M. Rodrigues1, J. Michael Reed1, Bryan S. Windmiller2
and L. Michael Romero1
1
2
Department of Biology, Tufts University, Medford, MA 02155, USA
Hyla Ecological Services, 1150 Main Street, #7, Concord, MA 01742, USA
(Received 10 February 2002; accepted 14 July 2002)
Abstract
We sampled blood from free-living spotted salamanders (Ambystoma maculatum) to test whether
differences in the concentrations of a stress hormone (corticosterone) were associated with different
qualities of breeding and migration habitat. Spotted salamanders are forest specialists that migrate to
vernal pools to breed, and upland habitat degradation may have sub-lethal effects on animals that lead
to population declines. An individual’s level of physiological stress may function as a biomonitor for
sub-lethal effects, and thus as a biomonitor for ecosystem quality. We compared unstressed (baseline)
and stress-induced corticosterone concentrations in spotted salamanders: (1) at sites that differed in
amount of forest loss; (2) during breeding migration across forest habitat versus pavement; (3) in
microhabitats that varied in soil drainage and canopy cover. Removal of large amounts of terrestrial
habitat surrounding a breeding pond was correlated with lower baseline (in males) and stress-induced
corticosterone concentrations, which may indicate healthy individuals with a reduced ability to respond
to additional stress or individuals experiencing chronic stress. Male salamanders migrating across
pavement had elevated baseline corticosterone concentrations compared to animals migrating through a
forest, consistent with an acute stress response. However, concentrations of corticosterone did not differ
between individuals in microhabitats with canopy cover and well-drained soil versus those in
microhabitats with no canopy cover and/or swampy soil. This endocrinological technique may be one
useful measure of a population’s health, helping to identify populations where further ecological study
is recommended to evaluate conservation concerns.
INTRODUCTION
For more than a decade, conservation biologists have
recorded apparent declines of amphibian populations
world-wide (Bury, 1999). Research has identified many
factors, such as pollution, ultraviolet radiation, introduced
species and fungal outbreaks contributing to declines of
amphibian populations (Blaustein et al., 1994; Gamradt
& Kats, 1996; Hopkins, Mendonça & Congdon, 1997;
Berger et al., 1998; Kiesecker, Blaustein & Belden, 2001).
As with other taxa, however, habitat loss and degradation
is considered the major cause of amphibian decline
(Wyman, 1990; Blaustein et al., 1994; Pounds, Fogden &
Campbell, 1999). The most threatened habitat world-wide
may be wetlands (e.g. Dahl, 1990), and most amphibians,
even if they are primarily terrestrial, require wetlands for
some part of their life-cycle (Duellman & Trueb, 1986).
All correspondence to: Rebecca Newcomb Homan.
Tel.: 617–627–3195; fax: 617–627–3805 (c/o J.M. Read);
[email protected].
However, many pond-breeding amphibian species spend
the majority of each year in adjacent terrestrial habitat
(e.g. Husting, 1965; Windmiller, 1996), and the potential
impacts to amphibians of terrestrial habitat degradation
immediately surrounding wetlands are largely unstudied
(but see Windmiller, 1996; deMaynadier & Hunter, 1998,
1999; Larson et al., 1998). Beyond reducing population
size, it is not clear if terrestrial habitat degradation has
other impacts on bond-breeding amphibians. In particular,
the sub-lethal impacts of environmental degradation on
individuals and populations can be difficult to evaluate
(Belden, Wildy & Blaustein, 2000; Rose, 2000; Reed,
2002).
Baseline and stress-induced patterns of physiological
stress may be useful biomonitors of sub-lethal impacts.
The stress response is being used increasingly as
a biomonitor for potentially threatened populations
(e.g. Creel, Creel & Monfort, 1997; Wasser et al., 1997),
and elevated glucocorticoid concentrations may be
useful in predicting survival of individuals in stressed
populations (Romero & Wikelski, 2001). Glucocorticoid
12
R. N. HOMAN ET AL.
hormones are released in response to stressful
environmental stimuli (such as storms and predators) and
can have beneficial short-term effects, such as temporary
suppression of reproduction, increased foraging, immune
system regulation and increased gluconeogenesis (e.g.
Wingfield et al., 1997; Wingfield & Romero, 2001). If the
stressor is long lasting (i.e. chronic), however,
glucocorticoids can have negative consequences for an
animal, such as shutdown of reproduction, chronic
suppression of both immune function and growth, and
neuronal cell death (Sapolsky, 1992; Wingfield &
Romero, 2001).
The primary glucocorticoid released by amphibians is
corticosterone (Jungreis, Huibregtse & Ungar, 1970), and
the roles of corticosterone in amphibians are similar to
those summarized by Wingfield & Romero (2001) for
other vertebrate taxa. Furthermore, elevated corticosterone in amphibians has been reported to suppress
courtship activity (Moore & Miller, 1984; Lowry &
Moore, 1991) and to increase the rate of metamorphosis
(Denver, 1997, 1998; Larson et al., 1998). Corticosterone
concentrations can fluctuate in amphibians in response to
an environmental stressor, as Hopkins et al. (1997) found
for southern toads (Bufo terrestris) exposed to coal
combustion waste, and Larson et al. (1998) found for tiger
salamander (Ambystoma tigrinum) larvae exposed to the
pesticide atrazine. We propose that habitat degradation
also could act as a stressor, potentially resulting in
negative long-term consequences for individuals and,
thus, the population. To evaluate the hypothesis that
degraded habitat can act as a stressor, we conducted three
studies of spotted salamanders (Ambystoma maculatum)
to determine how their baseline and stress-induced
corticosterone concentrations responded to different types
of apparent habitat degradation or quality.
Spotted salamanders are pond-breeding amphibians
that spend more than 11 months of the year in the
terrestrial habitat surrounding their breeding pond
(Downs, 1989; Windmiller, 1996) and populations appear
to be sensitive to forest loss (Windmiller, 1996;
deMaynadier & Hunter, 1998). Our first study compared
corticosterone concentrations from individuals from two
study sites, one in which approximately 36% of the
terrestrial habitat within 200 m of the breeding pond edge
had been replaced with a housing development within the
last 5 years and one in which the terrestrial habitat
surrounding the breeding pond was largely intact. Our
second study compared corticosterone from individuals
migrating through forested habitat with individuals
migrating across a paved parking lot. Finally, we
compared corticosterone levels of individuals in a single,
undisturbed population, that differed by migrating
through ostensibly good and poor microhabitats, where
habitat quality was attributed by tree canopy and
soil characteristics (Downs, 1989; Windmiller, 1996).
For each study, we predicted that living in or
moving across apparently degraded or poorer-quality
habitat would be associated with elevated baseline
corticosterone concentrations (Wingfield & Romero,
2001).
METHODS
Study 1: Breeding ponds with different amounts of
intact forest habitat
Spotted salamanders were captured during the springs of
1999–2001 in the terrestrial habitat surrounding two
breeding ponds in Sudbury, Massachusetts (N42° 22′,
W71° 25′). These breeding ponds are approximately 56 m
apart and are separated by a narrow two-lane paved road,
which is rarely crossed by amphibians (pers. obs.).
Approximately 36% of the forested habitat within 200 m
surrounding one of the breeding ponds, but not including
habitat on the opposite side of the road, was removed in
1997 and replaced by a housing subdivision (referred to
hereafter as the Disturbed site). The terrestrial habitat
within 200 m of the other breeding pond (Undisturbed
site), again not including habitat on the opposite side of
the road, has been left largely unaltered during the last
6 years, with the exception of a single house lot that was
developed in 1998, converting approximately 9% of the
existing wooded upland. Salamanders were captured as
they emerged from their overwintering sites in the spring
and headed towards the breeding pond, and as they left
the breeding pond a few days to a few weeks later, postbreeding. Drift fencing with pitfall traps (paired 5 gallon
buckets spaced every 15 m) surrounded both ponds, and
animals were captured on rainy spring nights when
they were walking along the fence, or once they fell
into pitfall traps (Dodd & Scott, 1994). Animals captured
in traps were in the buckets for a few minutes to
4 hours.
Study 2: Migration across forest habitat and across
pavement
Spotted salamanders were captured in the spring of 2000
and 2001 in the terrestrial habitat adjacent to two small
breeding ponds in Concord, Massachusetts (N42° 27′,
W71° 21′]. The breeding ponds were located at the edge
of a golf course. Animals migrating to these ponds in the
spring emerge from their overwintering sites in forest,
travel through forested habitat, then cross a well-lit, paved
parking lot that is approximately 17 m across, followed
by a driveway road (approximately 5 m wide) paved with
macadam, before reaching the edge of the golf course and
approaching the ponds. This parking lot is a popular
gathering place for people hoping to spot migrating
salamanders, and on several of the nights of our study
other people were also walking around the paved area,
introducing additional anthropogenic influences. Animals
migrating towards the breeding pond were captured as
they emerged from the woodland before encountering
the parking lot pavements, or after they crossed the parking lot. To ensure that no animal was sampled twice,
animals were transported directly to the pond after being
bled. We located animals by walking along the edges of
the parking lot after nightfall and virtually detecting them.
We also compared these samples to inbound animals at
the Undisturbed pond in Study 1, where migration was
through forested habitat.
Habitat quality and physiological stress
Study 3: Microhabitat variation
For this study, spotted salamanders were captured during
the falls of 1999 and 2000 in the terrestrial habitat
surrounding the Undisturbed pond described in Study 1.
Fall migrations may be a means by which an animal leaves
one burrow to search for a better location before the onset
of cold weather and snow (Duellman, 1954). Animals
were captured on rainy fall nights, using the same drift
fences and pitfall traps detailed previously. We used drift
fences arrayed in five concentric arcs ranging from 3 m
from the pond edge to approximately 200 m from the pond
edge for our sampling. These fence arrays intersected
several habitat types, including well-drained upland forest,
a red maple (Acer rubrum) swamp, and a small clearing
with scattered patches of white pine (Pinus strobus). Drift
fences were constructed both along the edges of these
microhabitat types and within them, so that each enclosure
created by the fence system contained one primary
microhabitat. We categorized a salamander as moving
away from a good microhabitat if it was migrating out of
upland forest, whereas a salamander migrating out of the
old field (poor canopy cover) or out of the red maple
swamp (poorly drained soil) was considered to be
moving from a poor microhabitat (Downs, 1989;
Windmiller, 1996). These are coarse classifications of
habitat quality, but they were used to maximize
differences among sites.
Breeding technique
We obtained the first blood sample within 3 min
of observing a salamander. Changes in corticosterone
concentrations can be detected in the blood 3 to 5 min
after a stressful stimulus, so blood samples taken within
3 min are assumed to represent baseline corticosterone
concentrations (Wingfield et al., 1997). For Studies 1
and 3, animals were then held for 30 min in small, plastic
containers, approximately 1 litre in volume, with a
few moist leaves (to prevent desiccation). At 30 min, a
second blood sample, hereafter referred to as a stressinduced blood sample, was taken to determine the
corticosterone response to an acute stress, which was, in
this case, the combination of capture, the initial blood
draw and subsequent handling (Wingfield et al., 1997;
Homan, Reed & Romero, in press). When bleeding was
completed, the gender of each animal was determined by
external examination of the cloaca (Downs, 1989), and
animals were released at their capture locations. In Study
2, our 30 min bleed (stress-induced) sample size for males
after crossing the parking lot was too small for statistical
analysis, so we report only the results of the baseline
(3 min) values for all samples.
To extract blood from salamanders, we used a 21 gauge
needle to puncture the caudal vein. For each bleeding
event, we collected less than 60 µl of upwelling blood in
heparinized microhematocrit capillary tubes (Fisher
Scientific, St. Louis, MO), which were then stored on ice
for 2 to 18 hours. On the rare occasion that blood flow
exceeded our requirements, pressure applied to the wound
13
for less than 1 min stanched bleeding. In the lab, we
centrifuged samples at approximately 400 g for 10 min,
removed the plasma, which generally composed ≥ 50% of
the whole blood, and stored the plasma at –20°C.
Although we attempted to measure steroid concentrations
in plasma samples that ranged in volume from 2 µl to 20
µl, we found that at least 5 µl of plasma was required to
generate detectable concentrations (Wingfield et al.,
1997). All protocols were approved by the Tufts
University Animal Care and Use Committee.
Radioimmunoassay and statistical analyses
Radioimmunoassay, as described by Wingfield, Vleck
& Moore (1992), was used to measure plasma
corticosterone concentrations. Briefly, we used redistilled dichloromethane to extract the corticosterone,
which was then resuspended in phosphate buffer. A
small amount of titriated corticosterone (NEN, nuclear,
Boston, MA) was used to determine the recovery of
corticosterone through the assay. Sample data were
compared to a standard curve, created for each assay
with known corticosterone concentrations, and adjusted
to account for percent recovery to determine actual
concentrations of corticosterone. Owing to changes in
antibody availability, we used two corticosterone
antibodies with different sensitivities in radioimmunoassays conducted on samples taken from
Studies 1 and 3 (B21–42 and B3–163 from Endocrine
Sciences, Austin, TX). The different sensitivities
resulted in an inter-assay variation of 30% between the
two antibodies. Consequently, for Studies 1 and 3, all
assays were standardized to one another using a pool
of starling (Sturnus vulgaris) plasma. A sample of
the starling plasma pool was included with every assay
as an internal control. The intra-assay variations for
the assays of Studies 1 and 3 ranged from 2 to 16%.
We used one antibody for the analyses of Study 2
(B21–42), so we did not need to standardize among
assays for this experiment. The parking lot interassay variation was 19% and the intra-assay variation
was less than 3%. If corticosterone concentrations
were below the detection limit for the assay, they were
assigned the value of the detection limit for statistical
analysis.
For Studies 1 and 3, in which multiple samples were
taken from the same individual, we performed a repeated
measure analysis of variance looking for effects of gender,
capture location (along fence versus in the trap), direction
of migration (towards or away from the breeding pond),
and habitat condition (disturbed versus undisturbed) for
Study 1 or microhabitat quality (good versus poor) for
Study 3 on baseline and stress-induced corticosterone
concentrations. These analyses were followed by post-hoc
tests to examine differences specific to baseline or stressinduced corticosterone concentrations. For sample sizes
for each study, please see the figures. We used a student’s
t-test for the analysis of data from Study 2. Analyses were
conducted using SAS (1999).
14
R. N. HOMAN ET AL.
RESULTS
Capture protocol
Neither capture location (in Study 1 and Study 3, where
animals were captured either along fences or in the trap)
nor direction of migration (in Study 1) significantly
affected corticosterone concentrations, and they were
removed from the model in a stepwise fashion.
Study 1: Breeding ponds with different amounts of
intact forest habitat
Corticosterone concentrations were significantly higher in
females than in males (baseline F = 5.51, d.f. = 1,54, P =
0.007, stress-induced F = 3.42, d.f. = 1,54, P = 0.04).
Comparisons between ponds, therefore, were done
separately for each gender. For females, baseline levels (≤
3 min) were not significantly different between sites (F =
2.21, d.f. = 1,20, P = 0.15; Fig. 1(a)). Stress-induced
(30 min) corticosterone levels, however, were significantly
higher in females at the Undisturbed site than at the
Disturbed site (F = 6.25, d.f. = 1,20, P = 0.02; Fig. 1(a)).
Males at the Undisturbed site had significantly higher
baseline (F = 5.03, d.f. = 1,33, P = 0.03; Fig. 1(b)) and
stress-induced (F = 11.20, d.f. = 1,33, P = 0.002; Fig. 1(b))
corticosterone concentrations compared to those at the
Disturbed site. Although there was a significant increase
in response to capture and handling for both females and
males at the Undisturbed site (females F = 6.40, d.f. = 1,11,
P = 0.03; males F = 13.32, d.f. = 1,21, P < 0.01), there was
no significant increase in response to capture and handling
for animals at the Disturbed site (females F = 1.51, d.f. =
1,9, P = 0.25; males F = 2.10, d.f. = 1,12, P = 0.17).
Study 2: Migration across forest habitat and across
pavement
Female salamanders had higher corticosterone
concentrations after crossing the forested habitat than did
males (F = 4.42, d.f. = 1,11, P = 0.059; Fig. 2), although
this gender difference was absent after crossing the
parking lot (F = 0.61, d.f. = 1,14, P = 0.45; Fig. 2). Given
the gender difference present before crossing the parking
lot, we analyzed corticosterone concentrations separately
for males and females. Male spotted salamanders had
significantly higher baseline corticosterone concentrations
after they crossed the parking lot than before they crossed
(because Levine’s test for homogeneity of variances was
nearly significant, t = 2.22, d.f. = 8, P = 0.057, we used an
Fig. 1. Baseline (≤ 3 min after capture) and stress-induced
(30 min after capture) corticosterone concentrations of spotted
salamanders in Undisturbed (largely intact forest) (Homan et al.,
in press) and Disturbed (36% of forest converted to subdivision 5
years before present) terrestrial habitat. Data are presented as
means ± 1 standard error. (a) Female spotted salamanders captured
during spring migrations towards and away from the breeding
pond in Undisturbed habitat and in Disturbed habitat. (b) Male
spotted salamanders captured during spring migrations towards
and away from the breeding pond in Undisturbed habitat and in
Disturbed habitat. Sample sizes are given as numbers in the bars.
Fig. 2. Baseline (≤ 3 min after capture) corticosterone
concentrations of spotted salamanders before and after crossing
a parking lot (~17 m across) during spring migration towards
the breeding pond. Individuals in the ‘before’ category had
migrated through forested habitat. Different animals were used
in the ‘before’ and ‘after’ samples. Data are presented as means
± 1 standard error. The asterisk represents a P-value of 0.035.
Sample sizes are given as numbers in the bars.
Habitat quality and physiological stress
unequal variances t-test, t = 2.54, d.f. = 7.9, P = 0.035;
Fig. 2). Females, however, showed no significant
difference in their baseline corticosterone concentrations
after having crossed the parking lot than before crossing
(homogeneous variances, t = 1.36, d.f. = 17, P = 0.19; ttest, t = 1.00, d.f. = 17, P = 0.33; Fig. 2). We also
compared the corticosterone concentrations of animals
migrating through the Undisturbed site from Study 1,
composed primarily of upland forest habitat, to those of
animals in the forested habitat of this study as a
comparison for site differences. There were no significant
differences in baseline corticosterone concentration in
males (t-test, t = 0.02, d.f. = 18, P = 0.98) or females (ttest, t = 0.44, d.f. = 15, P = 0.67) between the Undisturbed
site and the forested (before crossing) treatment at the
parking lot.
Study 3: Microhabitat variation
No significant differences in corticosterone were found
relative to gender (P > 0.1), so it was removed from the
model. Gender differences in corticosterone
concentrations are often related to breeding (Wingfield,
1994; Wingfield, O’Reilly & Astheimer, 1995; Jessop,
2000), so it is not surprising that they exist during the
breeding season, as seen in Studies 1 and 2, and not during
the fall, a non-breeding season. Although mean
corticosterone concentrations were higher in the poor
microhabitat than in the good microhabitat, the overall
differences (baseline and stress-induced combined) were
not statistically significant (F = 2.46, d.f. = 1,43, P = 0.12;
Fig. 3). The higher mean corticosterone concentration in
the poor microhabitat was due largely to one outlier (> 3
SD from the mean) in the baseline measurements and two
outliers (> 2 SD from the mean) in the stress-induced
measurements. When we compared the two types of poor
Fig. 3. Baseline (≤ 3 min after capture) and stress-induced
(30 min after capture) corticosterone concentrations of spotted
salamanders in microhabitats having canopy cover and welldrained soil, labelled ‘good’, and microhabitats having poor
canopy cover and/or poorly drained soil, labelled ‘poor’, during
the fall (Homan et al., in press). Data are presented as means ±
1 standard error. Sample sizes are given as numbers in the bars.
15
microhabitat classification (poor soil drainage and poor
canopy cover) to the good microhabitat separately, we
found that overall corticosterone concentrations of
animals in poorly drained soil did not differ from animals
in good habitat (F = 0.37, d.f. = 1,38, P = 0.55). While our
first test of overall corticosterone concentrations from
animals in microhabitat with poor canopy cover was
significantly higher than for those of animals in good
microhabitat (F = 5.9, d.f. = 1,38, P = 0.02), when we
removed a single outlier from the analysis, the difference
between microhabitats was not there (F = 0.01, d.f. = 1,23,
P = 0.91). There was no significant response to capture
and handling in either microhabitat (F = 2.04, d.f. = 1,43,
P = 0.16).
DISCUSSION
The relationships between habitat and baseline and stressinduced corticosterone concentrations differed among our
three studies. In Study 1, both males and females showed
different corticosterone profiles between Disturbed and
Undisturbed sites, whereas in Study 2, only males showed
increases in corticosterone after crossing a paved parking
lot. In Study 3, different microhabitat characteristics had
no significant impact on corticosterone concentrations.
These varying responses to different types of habitat
suggest that corticosterone concentrations may be useful
for quantifying potential sub-lethal effects on individuals,
which could, in turn, negatively impact the population.
Both baseline and stress-induced corticosterone
concentrations reflected the physiological status of an
individual, and thus may offer clues to the quality of their
environment.
Elevated
baseline
corticosterone
concentrations suggest that focal animals are in a more
stressful environment than are control animals (Wingfield
& Romero, 2001). For example, habitat contamination
from coal combustion waste was linked to increased
baseline corticosterone concentrations in southern toads
(Hopkins et al., 1997). Conversely, lower baseline
corticosterone concentrations relative to control suggest
that focal animals are in a less stressful environment
(Wingfield & Romero, 2001). However, interpretations of
corticosterone concentrations are more complex.
Evidence from many species indicates that robust
corticosterone increases in response to an acute stressor
(e.g. capture) are indicative of a healthy individual free of
long-term chronic stress (Wingfield & Romero, 2001). A
damped or absent corticosterone increase, on the other
hand, is usually interpreted as a disruption of normal
corticosterone negative feedback, and a direct consequence of chronic stress prior to capture (Hontela et al.,
1992; Norris et al., 1999).
Applying these interpretations to the data from two sites
that had different levels of terrestrial habitat lost to
suburban development (Study 1) leads to conflicting
conclusions. Male and female baseline corticosterone
concentrations were significantly lower for salamanders
living in a terrestrial habitat that was recently disturbed,
compared to those living in a relatively undisturbed
terrestrial habitat. The recently disturbed site, therefore,
16
R. N. HOMAN ET AL.
may be a less stressful habitat for these salamanders. The
apparent superiority of the Disturbed habitat could be a
consequence of only catching the healthy survivors of the
habitat degradation. In fact, the density of spotted
salamanders was reduced following the destructive events
associated with the degradation (B. S. Windmiller, R. N.
Homan, J. M. Reed & J. V. Regosin, unpublished data).
The lower density may mean that the surviving
salamanders are in a better habitat, possibly leading to
lower corticosterone levels as an artefact of biased
survival (as is seen with birds, e.g. Silverin, 1998). Data
currently being analyzed on growth rates and breeding
frequency will test this hypothesis.
However, the lack of a corticosterone response to
capture and handling in animals living at the Disturbed
site suggests the opposite conclusion. We assume that
capture and handling are novel and potent stressors to
free-living wild animals (e.g. Wingfield & Romero, 2001),
and, therefore, that habituation is not a factor in the lack
of a corticosterone response. Thus, the inability to respond
to acute stress is consistent with a prior chronic stress,
suggesting that animals were suffering from sub-lethal
decreases in health owing to the degraded habitat. Two
processes could explain these apparently disparate results.
First, lower baseline corticosterone concentrations
indicate that animals are healthy at the Disturbed site.
However, the lack of response to capture and handling
indicates that these same animals are vulnerable because
they are less capable of responding to further stressors,
and the acute stress response is important for individual
survival (Sapolsky, Romero & Munck, 2000).
A second explanation that accounts for both the
baseline and stress-induced corticosterone concentrations
is that animals at the Disturbed site may be experiencing
chronic stress. Some studies report that chronic stress can
lead to the depression of both the baseline levels of
circulating corticosterone and the corticosterone response
to an additional stressful stimulus (Selye, 1971; Hontela
et al., 1992; Norris et al., 1999). It is currently unknown
whether this occurs in salamanders. Distinguishing
between these two potential explanations would require
more study. It is clear that individuals differ between sites,
however, and a possible explanation is chronic stress at
the Disturbed site. This is cause for concern because
vertebrate populations under chronic stress may be
adversely affected by suppression of reproductive and
immune function, and slower growth (Sapolsky, 1992;
Wingfield & Romero, 2001).
Results from Study 2 (crossing a parking lot) suggest
that some forms of terrestrial habitat degradation also
could act as acute stressors. Consistent with an acute stress
response (Wingfield & Romero, 2001), male spotted
salamanders significantly increased baseline corticosterone concentrations after crossing a well-lit parking lot
on their way to the breeding pond. Acute stress responses
are adaptive, temporarily suppressing non-essential
functions like reproduction and immune function and
increasing mobilization of stored energy (e.g. Wingfield
et al., 1997; Wingfield & Romero, 2001). For individuals
in Study 2, stress levels of those before crossing the
parking lot were similar to baseline values of animals in
Study 1, and the values for individuals that crossed the
parking lot were similar to stress-induced values in Study
1 that were a result of capture and handling. Amphibians
frequently cross or attempt to cross roads during migration
and dispersal (e.g. Fahrig et al., 1995; deMaynadier &
Hunter, 2000; Hels & Buchwald, 2001), and crossing a
road might invoke a stress response similar to what we
saw for salamanders crossing a paved parking lot. The
physical structure of these two potential barriers is similar,
although we cannot separate the effects of crossing the lit
pavement from the effects of the pedestrian traffic at our
study site. However, automobile traffic is by far the more
potent danger associated with road crossing by
amphibians (e.g. Fahrig et al., 1995; deMaynadier &
Hunter, 2000; Hels & Buchwald, 2001), so the
physiological effects demonstrated here are unlikely to be
the most important consequence.
Baseline corticosterone concentrations in females,
however, were not affected by crossing the parking lot.
Females in breeding condition down-regulate the acute
stress response in some species of birds (Wingfield, 1994;
Wingfield et al., 1995) and reptiles (Jessop, 2000), which
is thought to be an attempt to prevent relatively minor
stressors from interrupting breeding behaviours
(Wingfield & Romero, 2001). Although this phenomenon
has not been documented in spotted salamanders, it could
explain the lack of response.
In contrast to human-caused degradation of habitat
quality, more natural variation in microhabitat characteristics had no apparent effect on salamander corticosterone concentrations. Corticosterone concentrations did
not differ significantly in animals moving through poorly
drained or exposed microhabitats from those in animals
moving through well-drained, canopy-covered microhabitats. These results suggest that some natural variations
in habitat quality may not act as significant stressors for
spotted salamanders. However, there are several caveats
to this interpretation. First, we do not know whether
animals captured in each microhabitat were residing there,
or were migrating through. Duration of exposure may be
important. Second, the variations at this study site may not
have been extreme enough to elicit significant differences
in the physiological stress response. Finally, our definition
of habitat quality may have been insufficient, so that other
habitat conditions drove corticosterone levels. The lack of
response to capture and handling may be due to seasonal
changes in corticosterone release as seen in amphibians
(Dupont et al., 1979), and reptiles (Wilson & Wingfield,
1994; Lance, Grumbles & Rostal, 2001), or salamanders
moving in the fall may be choosing to migrate because
they are already stressed, thereby altering the pattern of
corticosterone release we are able to measure (Homan et
al., in press). A biomonitor, such as the measure of
physiological stress used here, may improve our ability to
predict which properties of habitat quality have the
greatest physiological impact on amphibian populations.
Other studies also have used physiological stress
as a biomonitor to evaluate habitat quality. Pollution
can elevate glucocorticoid levels in southern toads
Habitat quality and physiological stress
(Hopkins et al., 1997), and Galápagos marine iguanas
(Amblyrhynchus cristatus) (Wikelski, Romero & Snell,
2001). Northern spotted owls (Strix occidentalis caurina)
with territories exposed to different forestry practices also
have elevated glucocorticoid levels (Wassar et al., 1997),
but Creel et al. (1997) found that radio-collaring African
wild dogs (Lycaon pictus) had no effect on their stress
levels. These studies have shown that glucocorticoids can
be a valuable biomonitor to determine which populations
and which general situations require more detailed
ecological study, especially since these techniques do not
require sacrificing animals. Amphibians can be long-lived
(Flageole & LeClair, 1991) and have highly fluctuating
yearly reproductive success (Pechmann et al., 1991; Reed
& Blaustein, 1995), so determining whether an amphibian
population is declining can take many years of intensive
ecological monitoring. Conservation biologists concerned
about amphibians, therefore, might benefit from using the
technique outlined here to obtain relatively quickly
physiological clues about the impact of various practices
on the health of individuals that could lead to long-term
problems.
Acknowledgements
We thank the Sudbury Conservation Commission and
Concord Country Club for allowing us access to their
properties. We also thank E. Cavellarano, J. Fahey, D.
Homan, A. Kennedy, J. O’Brien, M. Pokras, L. RemageHealey, A. Robbins, J. Sarnat and E. Schlüter for help in
the field. Funding for this project came from National
Science Foundation grant IBN-9975502 to LMR, from a
Massachusetts Environmental Trust grant, from migration
monies made available to the Massachusetts Natural
Heritage and Endangered Species Program, and from
contributions from Hyla Ecological Services.
REFERENCES
Belden, L. K., Wildy, E. L. & Blaustein, A. R. (2000). Growth,
survival and behaviour of long-toed salamanders (Ambystoma
macrodactylum) exposed to ambient levels of UV-B radiation. J.
Zool. 251: 473–479.
Berger, L., Speare, R., Daszak, P., Green, D. E., Cunningham,
A. A., Goggin, C. L., Slocombe, R., Ragan, M. A., Hyatt, A. D.,
McDonald, K. R., Hines, H. B., Lips, K. R., Marantelli, G. &
Parkes, H. (1998). Chytridomycosis causes amphibian mortality
associated with population declines in the rain forests of Australia
and Central America. Proc. Natl. Acad. Sci. 95: 9031–9036.
Blaustein, A. R., Hoffman, P. D., Hokit, D. G., Kiesecker, J. M.,
Walls, S. C. & Hays, J. B. (1994). UV repair and resistance to
solar UV-B in amphibian eggs: a link to population declines?
Proc. Nat. Acad. Sci. 91: 1791–1795.
Bury, R. B. (1999). A historical perspective and critique of the
declining amphibian crisis. Wildl. Soc. Bull. 27: 1064–1068.
Creel, S., Creel, N. M. & Monfort, S. L. (1997). Radiocollaring and
stress hormones in African wild dogs. Conserv. Biol. 11: 544–548.
Dahl, T. E. (1990). Wetlands: losses in the United States 1780’s to
1980’s. Washington, DC: US Department of Interior, Fish and
Wildlife Service.
deMaynadier, P. G. & Hunter, M. J. (1998). Effects of silvicultural
edges on the distribution and abundance of amphibians in Maine.
Conserv. Biol. 12: 340–352.
17
deMaynadier, P. G. & Hunter, M. J. Jr (1999). Forest canopy closure
and juvenile emigration by pool-breeding amphibians in Maine.
J. Wildl. Mgmt. 63: 441–450.
deMaynadier, P. G. & Hunter, M. J. (2000). Road effects on
amphibian movements in a forested landscape. Nat. Areas J. 20:
56–65.
Denver, R. J. (1997). Environmental stress as a developmental cue:
corticotropin-releasing hormone is a proximate mediator of
adaptive phenotypic plasticity in amphibian metamorphosis.
Horm. Behav. 31: 169–179.
Denver, R. J. (1998). Hormonal correlates of environmentally
induced metamorphosis in the Western spadefoot toad,
Scaphiopus hammondii. Gen. Comp. Endocrin. 110: 326–336.
Dodd, C. K. & Scott, D. E. (1994). Drift fences encircling breeding
sites. In Measuring and monitoring biological diversity: standard
methods for amphibians: 125–130. Heyer, W. R., Donnelly, M.
A., McDiarmid, R. W., Hayek, L. A. C. & Foster, M. S. (Eds).
Washington, DC: Smithsonian Institution Press.
Downs, F. L. (1989). Ambystoma maculatum (Shaw), spotted
salamander. In Salamanders of Ohio: 108–124. Pfingsten, R. A. &
Downs, F. L. (Eds). Bulletin of the Ohio Biological Survey.
Duellman, W. E. (1954). Observations on autumn movements of
the salamander Ambystoma tigrinum tigrinum in southeastern
Michigan. Copeia 1954: 156–157.
Duellman, W. E. & Trueb, L. (1986). Biology of amphibians.
Baltimore: The John Hopkins University Press.
Dupont, W., Bourgeois, P., Reinberg, A. & Vaillant, R. (1979).
Circannual and circadian rhythms in the concentration of
corticosterone on the plasma of the Edible frog (Rana esculenta
L.). J. Endocrin. 80: 117–125.
Fahrig, L., Pedlar, J. H., Pope, S. E., Taylor, P. D. & Wegner, J. F.
(1995). Effect of road traffic on amphibian density. Biol. Conserv.
71: 217–222.
Flageole, S. & LeClair, R. Jr (1991). Etude démographique d’une
population des salamanders (Ambystoma maculatum) à l’aide de
la méthode squeletto-chronologique. Can. J. Zool. 70: 740–749.
Gamradt, S. C. & Kats, L. B. (1996). Effect of introduced crayfish
and mosquitofish on Clifornia newts. Conserv. Biol. 10:
1155–1162.
Hels, T. & Buchwald, E. (2001). The effect of road kills on
amphibian populations. Biol. Conserv. 99: 331–340.
Homan, R. N., Reed, J. M. & Romero, L. M. (In press). Gender and
seasonal differences in corticosterone concentrations of free-living
spotted salamanders (Ambystoma maculatum). Gen. Comp.
Endocrin.
Hontela, A., Rasmussen, J. B., Audet, C. & Chevalier, G. (1992).
Impaired cortisol stress response in fish from environments
polluted by PAHs, PCBs, and mercury. Arch. Environ. Contam.
Toxicol. 22: 278–283.
Hopkins, W. A., Mendonça, M. T. & Congdon, J. D. (1997).
Increased circulating levels of testosterone and corticosterone in
southern toads, Bufo terrestris, exposed to coal combustion waste.
Gen. Comp. Endocrin. 108: 237–246.
Husting, E. L. (1965). Survival and breeding structure in a
population of Ambystoma maculatum. Copeia 1965: 352–362.
Jessop, T. S. (2000). Modulation of the andrenocortical stress
response in marine turtles (Cheloniidae): evidence for a hormonal
tactic maximizing maternal reproductive investment. J. Zool. 254:
57–65.
Jungreis, A. M., Huibregtse, W. H. & Ungar, F. (1970).
Corticosteroid identification and corticosterone concentration in
serum of Rana pipiens during dehydration in winter and summer.
Comp. Biochem. Physiol. C 34: 683–689.
Kiesecker, J. M., Blaustein, A. R. & Belden, L. K. (2001). Complex
causes of amphibian population declines. Nature 410: 681–684.
Lance, V. A., Grumbles, J. S. & Rostal, D. C. (2001). Sex
differences in plasma corticosterone in desert tortoises, Gopherus
agassizii, during the reproductive cycle. J. Expt. Zool. 289:
285–289.
18
R. N. HOMAN ET AL.
Larson, D. L., McDonald, S., Fivizzani, A., Newton, W. &
Hamilton, S. (1998). Effects of herbicide atrazine on Ambystoma
tigrinum metamorphosis: duration, larval growth, and hormonal
response. Physiol. Zool. 71: 671–679.
Lowry, C. A. & Moore, F. L. (1991). Corticotropin-releasing factor
(CRF) antagonist suppresses stress-induced locomotor activity in
an amphibian. Horm. Behav. 25: 84–96.
Moore, F. L. & Miller, L. J. (1984). Stress-induced inhibition of
sexual behavior: corticosterone inhibits courtship behaviors of a
male amphibian (Taricha granulosa). Horm. Behav. 18: 400–410.
Norris, D. O., Donahue, S., Dores, R. M., Lee, J. K., Maladonado,
T. A., Ruth, T. & Woodling, J. D. (1999). Impaired adenocortical
response to stress by brown trout, Salmo trutta, living in metalcontaminated waters of the Eagle River, Colorado. Gen. Comp.
Endocrin. 113: 1–8.
Pechmann, J. H. K., Scott, D. E., Semlitsch, R. D., Caldwell, J. P.,
Vitt, L. J. & Gibbons, J. W. (1991). Declining amphibian
populations: the problem of separating human impacts from
natural fluctuations. Science 253: 892–895.
Pounds, J. A., Fogden, M. P. L. & Campbell, J. H. (1999).
Biological response to climate change on a tropical mountain.
Nature 398: 611–615.
Reed, J. M. (2002). Animal behavior as a tool in conservation
biology. In Conservation medicine: ecological health in practice:
145–163. Aguirre, A. A., Ostfeld, R. S., House, C. A., Tabor,
G. M. & Pearl, M. C. (Eds). Oxford: Oxford University Press.
Reed, J. M. & Blaustein, A. R. (1995). Assessment of ‘nondeclining’ amphibian populations using power analysis. Conserv.
Biol. 9: 1299–1300.
Romero, L. M. & Wikelski, M. (2001). Corticosterone levels predict
survival probabilities of Galapagos marine iguanas during El Niño
events. Proc. Nat. Acad. Sci. 98: 7366–7370.
Rose, K. A. (2000). Why are quantitative relationships between
environmental quality and fish populations so elusive? Ecol.
Applic. 10: 367–385.
Sapolsky, R. M. (1992). Stress, the aging brain and the mechanisms
of neuron death. Cambridge, MA: MIT Press.
Sapolsky, R. M., Romero, L. M. & Munck, A. J. (2000). How do
glucocorticoids influence stress responses? Integrating permissive,
suppressive, stimulatory, and preparative actions. Endo. Rev. 21:
55–89.
SAS (1999). SAS Statistics, Version 8. Cary, NC: SAS Institute,
Inc.
Selye, H. (1971). Hormones and resistance. Berlin: Springer.
Silverin, B. (1998). Territorial behaviour and hormones of pied
flycatchers in optimal and suboptimal habitats. Anim. Behav. 56:
811–818.
Wasser, S. K., Bervis, K., King, G. & Hanson, E. (1997). Noninvasive physiological measures of disturbance in the northern
spotted owl. Conserv. Biol. 11: 1019–1022.
Wikelski, M., Romero, L. M. & Snell, H. L. (2001). Marine iguanas
oiled in the Galapagos. Science 292: 437–438.
Wilson, B. S. & Wingfield, J. C. (1994). Seasonal and
interpopulation variation in plasma levels of corticosterone in the
side-blotched lizard (Uta stansburiana). Physiol. Zool. 67:
1025–1049.
Windmiller, B. S. (1996). The pond, the forest, and the city: spotted
salamander ecology and conservation in a human-dominated
landscape. Doctoral thesis, Tufts University.
Wingfield, J. C. (1994). Modulation of the adenocortical response
to stress in birds. In Perspectives in comparative endocrinology:
520–528. Davey, K. G., Peter, R. E. & Tobe, S. S. (Eds). Ottawa:
National Research Council of Canada.
Wingfield, J. C., Hunt, K., Breuner, C., Dunlop, K., Fowler, G. S.,
Freed, L. & Lepson, J. (1997). Environmental stress, field
endocrinology, and conservation biology. In Behavioral
approaches to conservation in the wild: 95–131. Clemmons, J. R.
& Buchholz, R. (Eds). Cambridge: Cambridge University Press.
Wingfield, J. C., O’Reilly, K. M. & Astheimer, L. B. (1995).
Modulation of the adenocortical responses to acute stress in Arctic
birds: a possible ecological basis. Am. Zool. 35: 285–294.
Wingfield, J. C. & Romero, L. M. (2001). Adrenocortical responses
to stress and their modulation in free-living vertebrates. In
Handbook of physiology, section 7: The endocrine system, Vol.
IV: Coping with the environment: neural and endocrine
mechanisms: 211–234. McEwen, B. S. & Goodman, H. M. (Eds).
New York: Oxford University Press.
Wingfield, J. C., Vleck, C. M. & Moore, M. C. (1992). Seasonal
changes of the adrenocortical response to stress in birds of the
Sonoran Desert. J. Expt. Zool. 264: 419–428.
Wyman, R. L. (1990). What’s happening to the amphibians?
Conserv. Biol. 4: 350–352.