A Sensory System at the Interface
between Urban Stormwater Runoff
and Salmon Survival
J A S O N F . S A N D A H L ,†
D A V I D H . B A L D W I N ,‡
J E F F R E Y J . J E N K I N S ,†A N D
N A T H A N I E L L . S C H O L Z *, ‡
Oregon State University, Department of Molecular and
Environmental Toxicology, 333 Weniger Hall,
Corvallis, Oregon 97331, and NOAA Fisheries, Northwest
Fisheries Science Center, Ecotoxicology and Environmental
Fish Health Program, 2725 Montlake Boulevard East,
Seattle, Washington 98112
Motor vehicles are a major source of toxic contaminants
such as copper, a metal that originates from vehicle exhaust
and brake pad wear. Copper and other pollutants are
deposited on roads and other impervious surfaces and
then transported to aquatic habitats via stormwater runoff.
In the western United States, exposure to non-point
source pollutants such as copper is an emerging concern
for many populations of threatened and endangered
Pacific salmon (Oncorhynchus spp.) that spawn and rear
we used conventional neurophysiological recordings to
investigate the impact of ecologically relevant copper
exposures (0-20 µg/L for 3 h) on the olfactory system of
juvenile coho salmon (O. kisutch). These recordings were
combined with computer-assisted video analyses of
behavior to evaluate the sensitivity and responsiveness of
copper-exposed coho to a chemical predation cue
(conspecific alarm pheromone). The sensory physiology
and predator avoidance behaviors of juvenile coho were
both significantly impaired by copper at concentrations as
low as 2 µg/L. Therefore, copper-containing stormwater
runoff from urban landscapes has the potential to cause
in exposed salmon.
the coastal margins of countries such as the United States
(1, 2). Urbanization and other forms of coastal development
increase the runoff of pollutants from terrestrial landscapes
to the aquatic environment. Upon completing the most
coasts, and the Great Lakes in more than three decades, the
threats to aquatic species (3). A similar review by the Pew
Ocean Commission found that non-point sources represent
the greatest pollution threat to oceans and coasts (4). For
at-risk aquatic species, the current conservation challenges
associated with toxic runoff are global in scope, complex,
expanding and poorly understood.
Pavement is a universal feature of urbanized landscapes,
and impervious surfaces accumulate chemical pollutants
from automobile traffic as well as from other sources (5).
During rainfall events, these contaminants are mobilized by
and residential uses of copper, including the incorporation
and various pesticide formulations. In addition, vehicle
emissions via exhaust and brake pad wear represent major
watershed, the loading of copper to surface waters will
traffic, and rainfall patterns. As an example of measured
concentrations in aquatic habitats, recent monitoring of
dissolved copper at levels that varied from 3.4 to 64.5 µg/L,
with a mean of 15.8 µg/L (9).
stocks of coho and other species of anadromous Pacific
natural range in the western U.S. (10). Currently, 26 distinct
population segments (evolutionary significant units; ref 11)
of coho, chinook (O. tshawytscha), sockeye (O. nerka), and
chum (O. keta) salmon as well as steelhead (O. mykiss) are
listed as either threatened or endangered under the U.S.
Endangered Species Act (ESA). In the case of coho, several
historical runs have been extirpated throughout California,
Oregon, Washington, and Idaho (12). To reverse salmon
and restore the quality of freshwater and estuarine habitats
(e.g., ref 13). Freshwater habitat quality is particularly
important for coho salmon that rear for more than a year in
lowland streams and ponds before beginning their seaward
Copper is a neurobehavioral toxicant in fish, and it has
been known for more than three decades that the metal
disrupts the normal function of the fish olfactory system
(15). Ultrastructural analyses have shown that dissolved
copper damages the olfactory sensory epithelium (16-19),
and previous studies using direct neurophysiological record-
ings from the fish nose (15, 17, 20, 21) or observations of
interferes with the ability of fish to detect and respond to
chemical signals in aquatic environments. Chemosensory
deprivation has important implications for salmon, as these
migratory animals rely on their sense of smell to find food,
from the ocean to freshwater spawning habitats, and assess
the reproductive status of prospective mates.
To determine whether short term (3 h) exposures to
dissolved copper at concentrations typical of urban storm-
water runoff (0-20 µg/L) interfere with olfaction and
olfactory-mediated behaviors in juvenile coho salmon, we
imaging to quantify predator avoidance behaviors that are
normally triggered in juvenile salmon by a conspecific
chemical alarm pheromone (25). For each copper exposure
* Corresponding author phone: (206) 860-3454; fax: (206) 860-
3335; e-mail: Nathaniel.Scholz@noaa.gov.
†Oregon State University.
Environ. Sci. Technol. 2007, 41, 2998-3004
29989ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 8, 200710.1021/es062287r CCC: $37.00
2007 American Chemical Society
Published on Web 03/14/2007
concentration, recordings from olfactory sensory neurons
were matched to behavioral observations from the same
animal. This allowed us to evaluate the sublethal neurobe-
havioral effects of copper at two different biological scales
toxicity (i.e., electrophysiology) are predictive of behavioral
Animals. Coho salmon eggs were obtained from the Uni-
versity of Washington hatchery (Seattle, WA) at the eyed egg
stage and raised at the Northwest Fisheries Science Center’s
parr were maintained in tanks supplied with filtered,
dechlorinated municipal water (hereafter referred to as
hatchery water; 120 mg/L total hardness as CaCO3, pH 6.6,
dissolved oxygen 8.1 mg/L, temperature 11-13 °C) on a
single-pass flow system. Fish were raised on standard
commercial salmon pellets (Bio-Oregon, Warrenton, OR).
Fish were 4-5 months of age with an average ((SD) length
of 4.6 ( 0.4 cm and a weight of 0.9 ( 0.2 g.
The homogenate was then filtered through polyester floss,
diluted to a final concentration of 100 cm2skin/L in distilled
water, mixed, aliquoted into 10 mL glass vials, and stored at
-20 °C. Immediately before use, aliquots were thawed,
filtered, and diluted 1:100 in hatchery water to a final
concentration of 1 cm2skin/L. Control blank solutions
consisted of hatchery water only. Although the as-yet
the pheromone is contained within specialized club cells
in ref 27). Thus, the concentration of pheromone is likely to
vary in proportion to the protein content of the skin extract.
Moreover, protein assays are more precise and more
reproducible than estimates of epidermal surface area.
Accordingly, we measured the total protein content of the
(Coomassie Plus-2000 Protein Assay Reagent, Pierce, Rock-
ford, IL). Odor stimulus concentrations are reported as mg
(or µg) of protein/L. As a point of reference, 1 cm2skin was
empirically determined to be equivalent to 5 mg of protein.
Moreover, a mechanical disruption of the skin as small as 1
mm2(50 µg of protein) would be sufficient to fill 100 L to a
concentration of 0.5 µg/L protein, a concentration within
the experimental range examined here.
Copper Exposures and Chemical Analysis. Copper-
copper chloride (Sigma Chemical Co., St. Louis, MO; 99%
purity cupric chloride, dihydrate) in distilled water. A total
of five stock solutions was prepared, such that adding 100
dissolved copper concentrations of 0, 2, 5, 10, and 20 µg/L
were visually isolated from each other. Prior to the introduc-
tion of fish, 100 mL water samples for dissolved copper
analysis were collected in acid-washed, Teflon bottles and
refrigerated at 4 °C. Fish were then exposed to copper for 3
per exposure concentration) in separate tanks using freshly
stock. Individual exposures were staggered to maintain a
and the onset of either behavioral or electrophysiological
trials. Different combinations of copper-exposed fish were
group was tested on each day. Water temperature, pH, and
dissolved oxygen (dO) remained relatively constant over the
course of the exposure period, with a mean and range (in
parentheses) of 10.8°C (10-12 °C), pH 6.7 (6.5-7.1), and 8.2
mg/L dO (6.5-9.6 mg/L).
Nominal exposure solutions were analyzed for total
dissolved copper by an outside laboratory using inductively
coupled plasma mass spectrometry (Frontier Geosciences,
from exposure tanks ranged from 84 to 102% of nominal
expressed in terms of nominal concentrations.
Following a 3 h exposure, the behavioral response of each
juvenile coho to a chemical predation cue was monitored
using a computer-assisted, three-dimensional data acquisi-
tion system (29). The experimental design was modified
slightly from Scholz et al. (25). For the behavioral trials,
filled with 25 L of hatchery water. Continuous, closed
circulation mixing in the aquarium was provided by a small
aquarium pump. Conspecific skin extract was injected into
the behavioral observation tank via a 50 cm length of Tygon
tubing. Initial tests with dye indicated an even distribution
of odor stimulus throughout the tank within approximately
The three-dimensional position of fish was monitored
using two orthogonally placed Firewire digital cameras
computer (iBook, Apple Computer, Cupertino, CA), as
simultaneous images of the fish from the front and side of
the tank every 2 s. Each pair of images was then analyzed to
determine the position of the fish via triangulation, with a
correction for refraction. The three-dimensional distance
between subsequent pairs of images (divided by 2 s) was
used to calculate the swimming speed at each time point.
to acclimate for 30 min. A baseline, pre-stimulus swimming
speed for each animal was subsequently recorded for a 3
min interval (t ) -180-0 s). Following this, a small volume
of the chemical alarm substance (0.5 mL; 5 mg of protein/L)
was injected into the circulation system (t ) 0 s) to achieve
a final diluted concentration of 0.1 µg of protein/L in the
observation chamber. The post-stimulus swimming speed
of the fish was then monitored for an additional 4 min. To
in the initiation of the avoidance response among animals,
we selected a fixed 30 s interval (t ) 45-75 s) to measure the
post-stimulus swimming speed. On the basis of initial trials,
TABLE 1. Effects of Dissolved Copper on the Swimming
Behavior of Coho Salmona
0.3 ( 0.2
1.9 ( 0.4
4.7 ( 0.6
10.2 ( 1.6
16.8 ( 1.7
5.6 ( 0.4
6.0 ( 0.3
5.6 ( 0.3
5.2 ( 0.5
2.3 ( 0.4*
1.4 ( 0.3
3.7 ( 0.7
4.8 ( 0.7
4.1 ( 0.5
2.4 ( 0.5
aMeasured copper values are from three composite samples for
response was a 50% or greater reduction in locomotory activity (see
Experimental Procedures). Data are presented as mean ( SE or as
asterisks represent a statistical difference from controls (p < 0.05, one-
way ANOVA with a Dunnett’s post hoc and Fisher’s exact test,
VOL. 41, NO. 8, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY92999
this interval included the behavioral responses of almost all
the fish. The magnitude of the response was quantified by
comparing the change (reduction) in swimming speed over
Figure S1A). Additionally, the reaction to alarm pheromone
was scored as a predator avoidance response if the animal
exhibited motionlessness, as indicated by a reduction in
swimming speed of 50% or more. To reduce inter-animal
variability arising from risk-taking behavior (i.e., motivation
to forage in the face of a predation threat), we did not feed
juvenile coho during behavioral trials and thus did not
monitor food strikes (25).
Odor-Evoked Neurophysiological Recordings from the
Coho Olfactory Epithelium. Once the behavioral observa-
the peripheral olfactory epithelium of each juvenile coho
using established procedures (30). Fish were anaesthetized
in tricaine methanesulfonate (MS-222; 50 mg/L) and trans-
ferred to a vibration isolation table for electrophysiological
recordings. For each animal, the EOG evoked by an odorant
was measured in triplicate and then averaged to produce a
single response value. The size or amplitude of the EOG was
of the evoked peak relative to the pre-stimulus electrical
baseline (refs 20 and 21 and Supporting Information Figure
in hatchery water. The olfactory chamber of each animal
(with the nare intact) was perfused with a sequence of the
three different odorants: skin extract (10 µg of protein/L),
l-serine (10-5M), and TCA (10-6M). L-Serine and TCA are
well-studied odorants in salmon and were included for the
purposes of comparing the results of this study to previous
elicit similar, robust EOGs from the olfactory epithelium of
unexposed animals. Fish were euthanized by decapitation
after recording EOGs.
Initial Alarm Substance Range-Finding Experiments.
Several range-finding experiments were performed to mea-
sure the behavioral and physiological responses of fish to a
range of skin extract dilutions. For the alarm behavior,
unexposed juvenile coho salmon were presented with the
skin extract at nominal concentrations of 0 (hatchery water
blank), 0.04, 0.1, 0.4, and 1.0 µg of protein/L (n ) seven to
one skin extract dilution. For the physiological response,
in response to five dilutions of skin extract (0.1-10 µg of
to evaluate the effects of copper on the stimulus-response
relationship for the alarm substance, a third group of fish
was exposed to 2 µg/L of copper for 3 h, and EOG responses
to skin extract were then recorded at dilutions ranging from
0.4-40 µg of protein/L (n ) three to six fish per dilution).
Statistical Analysis. The electrophysiological and behav-
groups (followed by a Dunnett’s test for comparisons with
controls), Fisher’s exact test (for freeze responses), or
regression analysis to test for concentration-dependent
relationships. Paired t-tests were used to determine differ-
ences in pre-stimulus baseline activity and post-stimulus
activity for antipredator responses. Correlations were de-
termined by using the Pearson correlation procedure.
Statistical analyses and graphing were performed with
GraphPad Prism 4.0 (San Diego, CA) and SAS Institute JMP
5.1 (Cary, NC).
Neurobehavioral Responses to a Chemical Predation Cue
over a Range of Stimulus Concentrations. The onset of
s after the introduction of skin extract to the observation
tank. This brief delay presumably reflected variability in the
introduction, and variation in inter-animal behavior. In a
typical response, juvenile coho oriented to the direction of
in a relatively fixed position. Responsive fish also tended to
slowly settle toward the bottom of the tank (Supporting
Information, Movie S1). Although the stereotypical anti-
predator response was a rapid onset of motionlessness, the
duration of the response varied, with some animals freezing
for tens of seconds and others freezing for several minutes
The degree of acclimation to the observation tank was
consistent across the groups of fish, as indicated by a
comparable amount of baseline (pre-stimulus) swimming
activity among groups (mean ( SE; 5.2 ( 0.2 cm/s; one-way
ANOVA, p > 0.5). Fish presented with hatchery water only
(blank) showed no change in swimming speed over the pre-
relative to the pre-stimulus interval (paired t-test, p ) 0.12).
However, three of eight animals exhibited motionlessness
or freezing. A more pronounced antipredator response
occurred when the alarm stimulus concentration was
increased to 0.1 µg of protein/L. This included a 74 ( 6%
reduction in speed relative to controls (paired t-test, p <
changes were similarly pronounced at higher stimulus
concentrations (0.4 and 1.0 µg of protein/L; paired t-test, p
FIGURE 1. Odorant stimulus-response curves were determined
to skin in control (unexposed fish) and for EOG responses to skin
in fish exposed to 2 µg/L of copper (open circles). Fractions within
parentheses correspond to the proportion of fish tested in each
group that showed a >50% reduction in activity (number of fish
to hatchery water alone (blank stimulus). Unlike the behavioral
of skin extract. In both graphs, error bars represent one standard
error. Asterisks denote the skin extract concentration used in
subsequent copper exposure experiments.
30009ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 8, 2007
epithelium of coho indicated a concentration-dependent
increase in EOG amplitude in response to skin extract
(0.1-10 µg of protein/L; n ) eight to nine fish per stimulus
ranging from 0.2-1.7 mV, after subtraction of the response
indistinguishable from responses to the blank solution (p >
threshold for conspecific skin extract under these experi-
mental conditions was between 0.1 and 0.4 µg of protein/L.
Exposure to 2 µg/L of copper for 3 h reduced the EOG
the concentration relationship to the right (Figure 1). For
copper-exposed fish, responses to concentrations of skin
extract of 1 µg of protein/L or less were indistinguishable
from blank responses (p > 0.05, one-way ANOVA followed
by Dunnett’s post hoc), indicating an increase in response
threshold by about 1 log unit.
In summary, the conspecific skin extract elicited measur-
able electrophysiological and behavioral responses from
juvenile coho salmon at concentrations either above or at
0.1 µg of protein/L. The behavioral stimulus-response curve
was steep, with the skin extract evoking maximal predator
avoidance behaviors at a concentration that was below the
lowest concentration detectable via olfactory neurophysi-
ology. Also, since motionlessness was observed in response
to alarm substance at a concentration subthreshold for
evoked EOGs (0.04 µg of protein/L), the behavioral measure-
assays. On the basis of these initial observations, a stimulus
behavioral trials involving copper-exposed fish. To elicit a
robust EOG response, the olfactory chamber was perfused
with 10 µg of protein/L during neurophysiological experi-
Relative Thresholds for Neurophysiological and Behav-
ioral Impairment in Juvenile Coho Exposed to Dissolved
Copper. To determine the relative impacts of short-term
copper exposures (3 h; 2-20 µg/L) on olfactory sensitivity
to copper, monitored a behavioral response to 0.1 µg of
protein/L of skin extract, and then recorded odor-evoked
EOGs from each animal’s olfactory epithelium using con-
specific skin extract and two other natural odorants (the
amino acid L-serine and the bile salt TCA) as stimuli. The
Supporting Information includes examples of paired etho-
the behavioral responses of a control fish and a fish exposed
to 10 µg/L of copper (Movie S1).
dependent manner (ANOVA, p < 0.001, Figure 2A). In
unexposed animals, the mean EOG responses to 10 µg of
protein/L of skin extract, 10-5M L-serine, and 10-6M TCA
were 1.2, 2.8, and 4.0 mV, respectively. At the lowest copper
exposure concentration (2 µg/L), the mean skin extract-
evoked EOG amplitude was 0.6 mV, a significant reduction
relative to controls (ANOVA, Dunnett’s test, p < 0.01). At 20
µg/L of copper, EOG responses to all three odorants were
EOG ) max/(1 + (copper/EC50)slope), which was applied
shown in Figure 2A, the mean olfactory response of the
control group was used to define the value of max in the
Dissolved copper also disrupted odor-evoked predator
avoidance behaviors (Figure 2B and Table 1). For juvenile
coho exposed to copper at concentrations up to 10 µg/L,
pre-stimulus baseline swimming activity was indistinguish-
FIGURE 2. Exposure to copper diminished olfactory sensitivity and
alarm behavior in juvenile coho. (A) Electro-olfactogram (EOG)
responses shown here were blank-subtracted. The results of
nonlinear regressions are shown with solid lines (see Results for
details). (B) Copper exposure also reduced the alarm response
elicited by 0.1 µg of protein/L of skin extract in a dose-dependent
manner. The result of a nonlinear regression is shown with a solid
response means were highly correlated (i.e., fish with reduced
olfactory sensitivity showed reduced alarm behavior). Error bars
in all graphs represent one standard error.
VOL. 41, NO. 8, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY93001
as indicated by a reduction in the mean baseline activity in
these fish relative to controls (2.3 ( 0.4 cm/s; ANOVA, p <
0.05). The alarm pheromone triggered an average reduction
in swimming speed of 74 ( 6% (mean ( SE) among
unexposed animals. The behavioral change was highly
significant (paired t-test, p < 0.001), with all but one of the
unexposed animals (n ) 11 of 12 fish) becoming motionless
during the post-stimulus interval (Table 1). While the
reduction in swimming speed among fish exposed to 2 µg/L
fewer animals (n ) six of 12) became motionless. At higher
copper concentrations (5, 10, and 20 µg/L), there were no
significant reductions in swimming speed (paired t-tests, p
> 0.1), and the majority of fish did not become motionless.
The effect of copper on the alarm reaction also showed a
reasonable fit to the same sigmoidal function as the EOG
responses (r2) 0.80, Figure 2B). The duration of the alarm
reaction for the few fish that did respond to the pheromone
at these higher copper exposures was generally shorter than
114 ( 27 s for the 11 unexposed fish). Overall, however, too
few of the copper-exposed fish responded to allow for a
comparison of duration (not shown).
A direct comparison of the inhibitory effects of dissolved
is shown in Figure 2C. The relationship between olfactory
inhibition and diminished alarm response was significantly
correlated (Pearson r ) -0.97, r2) 0.94, p < 0.01). From the
slope of the correlated measures (linear regression, slope )
to a ∼29% decrease in the magnitude of the pheromone-
mediated predator avoidance behavior. Consequently, the
relative impacts of dissolved copper exposure are similar at
these two different scales of biological organization.
Our current findings provide an important link between
habitat degradation (i.e., dissolved copper exposure) and
that short-term exposures to dissolved copper diminish the
olfactory sensitivity of juvenile coho salmon and that this
loss of sensory function leads, in turn, to a failure to initiate
predator avoidance behaviors in response to a conspecific
(reviewed by ref 31) as well as for surviving encounters with
predators (32). Notably, these neuroethological effects of
range of measured copper levels in surface waters of urban
and urbanizing watersheds (e.g., 3-64 µg/L; ref 9).
The effective range of chemical alarm pheromone-
mediated signaling in aquatic systems is likely to vary with
the strength of the signal at the source (i.e., the degree of
damage to the skin of another fish), the turbulent dispersal
of the chemical cue, and the sensory capabilities of the
which a conspecific alarm signal is effective. Moreover,
cue at concentrations that would normally trigger anti-
predator behaviors in uncontaminated systems. The neu-
robehavioral basis for this shift can be seen in Figure 1 and
is illustrated in Figure 3. In the present study, the EOG
response of juvenile salmon following a 3 h exposure to
copper at 2 µg/L was reduced by ∼40% over the entire range
of odor concentrations (Figure 1), thereby shifting the
stimulus-response curve to the right nearly a log unit. This
by the continued reduction in EOGs following exposure to
copper content in surface waters increases, the responsive-
ness of the peripheral olfactory system to a predation cue
will diminish until it falls below the threshold required to
initiate an appropriate behavioral response (Figure 3).
exposed fish will make behavioral decisions that are inap-
propriately risky for a particular ecological situation (33).
The consequences of this for actual rates of predation on
an important area for future research.
Salmon will avoid copper originating from point sources
contaminated with diffuse non-point source runoff. For fish
will be reversible, with physiological recovery taking place
(21). At higher concentrations, including those sufficient to
trigger cell death in the sensory epithelium (i.e., g25 µg/L;
over days or weeks. In either case, intermittent rainfall can
be expected to drive a dynamic process of neurobehavioral
toxicity and recovery among salmon in urban creeks.
Finally, our current results in juvenile coho should be
applicable to other fish species in urbanizing watersheds
worldwide. In addition to coho (this study and refs 20 and
21), dissolved copper has been shown to impair olfaction in
chinook salmon (17, 23), rainbow trout (15, 18, 24), brown
FIGURE 3. Conceptual model to illustrate how shifts in olfactory
behavior. On the basis of the data in Figure 1, sigmoidal and power
functions were used to approximate the behavioral and olfactory
stimulus-response curves, respectively. In this theoretical model,
fish. Following exposure to copper, a shift in olfactory sensitivity
increases the strength of the stimulus needed to reach this
and the previous stimulus now effectively fails to elicit the alarm
30029ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 8, 2007
trout (Salmo trutta; ref 19), fathead minnow (Pimephales
ref 16), and tilapia (34). It is also likely that the neurotoxic
effects of copper extend beyond the olfactory networks that
et al. (21) found that copper reduces the sensitivity of coho
salmon to distinct classes of natural odorants in a similar,
dose-dependent manner. This suggests that copper is a
general-purpose inhibitor of fish olfaction and may thus
mechanosensory behaviors such as shoaling, prey capture,
and predator evasion. For these reasons, non-point source
stormwater runoff from roads has the potential to interfere
with a wide variety of behaviors in a diversity of fish species.
This work was supported by the NOAA Coastal Storms
Program, a Public Health Service grant (T32ES07060) from
the National Institute of Environmental Health Sciences to
J.F.S., and an internship from the Oak Ridge Institute for
Science and Education to D.H.B. We thank Jana Labenia,
Kate Macneale, and Jenifer McIntyre for comments on the
Supporting Information Available
Measured behavioral and EOG responses of fish (Figure S1);
examples of the behavioral and EOG responses from four
S1); and behavioral responses of a control fish and a fish
exposed to 10 µg/L of copper (Movie S1). This material is
(1) Crosset, K. M.; Culliton, T. J.; Wiley, P. C.; Goodspeed, T. R.
Population trends along the coastal United States: 1980-2008;
National Oceanic and Atmospheric Administration: Silver
Spring, MD, 2004.
(2) Beach,D.Coastalsprawl: Theeffectsofurbandesignonaquatic
ecosystems in the United States; Pew Ocean Comission: Ar-
lington, VA, 2002.
(3) An ocean blueprint for the 21st century; United States Com-
mission on Ocean Policy: Washington, DC, 2004; final report.
(4) America’s living oceans. Charting a course for sea change: A
report to the nation, recommendations for a new ocean policy;
Pew Ocean Commission: Washington, DC, 2003.
(5) Wheeler, A. P.; Angermeier, P. L.; Rosenberger, A. E. Impacts of
new highways and subsequent landscape urbanization on
stream habitat and biota. Rev. Fish. Sci. 2005, 13, 141-164.
(6) Sansalone, J. J.; Buchberger, S. G. Partitioning and first flush of
metals in urban roadway storm water. J. Environ. Eng. 1997,
(7) Hamilton, P. A.; Miller, T. L.; Myers, D. N. Water quality in the
nation’s streams and aquiferssOverview of selected findings,
1991-2001; U.S. Geological Survey Circular 1265: Reston, VA,
2004; p 62.
(8) Davis, A. P.; Shokouhian, M.; Ni, S. Loading estimates of lead,
copper, cadmium, and zinc in urban runoff from specific
sources. Chemosphere 2001, 44, 997-1009.
(9) Soller, J.; Stephenson, J.; Olivieri, K.; Downing, J.; Olivieri, A. W.
Evaluation of seasonal scale first flush pollutant loading and
2005, 76, 309-318.
(10) NationalResearchCouncil(NRC).Upstream: Salmonandsociety
(11) Waples, R. S. Pacific salmon (Oncorhynchus spp.) and the
Fish. Rev. 1991, 53, 11-22.
(12) Nehlsen, W.; Williams, J. E.; Lichatowich, J. A. Pacific salmon
and Washington. Fisheries 1991, 16, 4-21.
(13) Columbia River Basin salmon and steelhead: Federal agencies’
General Accounting Office: Washington, DC, 2002; GAO-02-
(14) Nickelson, T. E.; Rodgers, J. D.; Johnson, S. L.; Solazzi, M. F.
Seasonal changes in habitat use by juvenile coho salmon
Aquat. Sci. 1992, 49, 783-789.
(15) Hara, T. J.; Law, Y. M. C. MacDonald, S. Effects of mercury and
copper on the olfactory response in rainbow trout. J. Fish. Res.
Board Can. 1976, 33, 1568-1573.
(16) Beyers, D. W.; Farmer, M. S. Effects of copper on olfaction of
Colorado pikeminnow. Environ. Toxicol. Chem. 2001, 20, 907-
(17) Hansen, J. A.; Rose, J. D.; Jenkins, R. A.; Gerow, K. G.; Bergman,
trout (Oncorhynchus mykiss) exposed to copper: Neurophysi-
Toxicol. Chem. 1999, 18, 1979-1991.
(18) Julliard, A. K.; Saucier, D.; Astic, L. Time-course of apoptosis in
the olfactory epithelium of rainbow trout exposed to a low
copper level. Tissue Cell 1996, 28, 367-377.
(19) Moran, D. T.; Rowley, J. C.; Aiken, G. R.; Jafek, B. W. Ultra-
structural neurobiology of the olfactory mucosa of the brown
trout, Salmo trutta. Microsc. Res. Tech. 1992, 23, 28-48.
(20) Sandahl, J. F.; Baldwin, D. H.; Jenkins, J. J.; Scholz, N. L. Odor-
evoked field potentials as indicators of sublethal neurotoxicity
in juvenile coho salmon (Oncorynchus kisutch) exposed to
copper, chlorpyrifos, or esfenvalerate. Can. J. Fish. Aquat. Sci.
2004, 61, 404-413.
receptor pathways in the peripheral olfactory nervous system.
Environ. Toxicol. Chem. 2003, 22, 2266-2274.
(22) Carreau, N. D.; Pyle, G. G. Effect of copper exposure during
fathead minnows (Pimephales promelas). Ecotoxicol. Environ.
Saf. 2005, 61, 1-6.
(23) Hansen, J. A.; Marr, J. C. A.; Lipton, J.; Cacela, D.; Bergman, H.
mykiss) exposed to copper and cobalt: Behavioral avoidance.
Environ. Toxicol. Chem. 1999, 18, 1972-1978.
(24) Saucier, D.; Astic, L.; Rioux, P. The effects of early chronic
exposure to sublethal copper on the olfactory discrimination
(25) Scholz, N. L.; Truelove, N. K.; French, B. L.; Berejikian, B. A.;
Quinn, T. P.; Casillas, E.; Collier, T. K. Diazinon disrupts anti-
predator and homing behaviors in chinook salmon (Onco-
(26) Brown, G. E.; Adrian, J. C., Jr.; Patton, T.; Chivers, D. P. Fathead
minnows learn to recognize predator odor when exposed to
ioral-response threshold. Can. J. Zool. 2001, 79, 2239-2245.
(27) Smith, R. J. F. Alarm signals in fish. Rev. Fish Biol. Fish. 1992,
(28) Bradford, M. M. A rapid and sensitive method for the quan-
tification of microgram quantities of protein utilizing the
(29) Sandahl, J. F.; Baldwin, D. H.; Jenkins, J. J.; Scholz, N. L.
Environ. Toxicol. Chem. 2005, 24, 136-145.
(30) Baldwin, D. H.; Scholz, N. L. The electro-olfactogram: An in
vivo measure of peripheral olfactory function and sublethal
2; Ostrander, G. K., Ed.; CRC Press, Inc: Boca Raton, FL, 2005;
Vol. 2, pp 257-276.
(31) Brown, G. E. Learning about danger: Chemical alarm cues and
local risk assessment in pre fishes. Fish Fish. 2003, 4, 227-234.
(32) Mirza, R. S.; Chivers, D. P. Chemical alarm signals enhance
with predatory chain pickerel (Esox niger). Ethology 2001, 107,
(33) Kats, L. B.; Dill, L. M. The scent of death: Chemosensory
assessment of predation risk by prey animals. Ecoscience 1998,
(34) Bettini, S.; Ciani, F.; Franceschini, V. Recovery of the olfac-
tory receptor neurons in the African Tilapia mariae fol-
VOL. 41, NO. 8, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY93003
lowing exposure to low copper level. Aquat. Toxicol. 2006, 76,
copper triggers cell death in the peripheral mechanosensory
system of larval fish. Environ. Toxicol. Chem. 2006, 25, 597-
(36) Herna ´ndez, P. P.; Moreno, V.; Olivari, F. A.; Allende, M. L. Sub-
lethal concentrations of waterborne copper are toxic to lateral
Received for review September 25, 2006. Revised manuscript
received January 27, 2007. Accepted February 6, 2007.
30049ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 8, 2007