-
extracted text
-
Behavioral response of juvenile Chinook salmon (Oncorhynchus
tshawytscha) to combinations of dissolved copper and submerged structure
in freshwater and seawater.
by
Frank Sommers
A Thesis
Submitted in partial fulfillment
of the requirements for the degree
Masters of Environmental Studies
The Evergreen State College
March 2012
© 2012 Frank Sommers. All rights reserved.
This Thesis for the Master of Environmental Study Degree
by
Frank Sommers
has been approved for
The Evergreen State College
by
________________________
Amy Cook
Member of the Faculty
________________________
Maria Bastaki
Member of the Faculty
________________________
David Baldwin
Research Biologist NOAA Fisheries/NWFSC
________________________
Date
ABSTRACT
Behavioral response of juvenile Chinook salmon (Oncorhynchus tshawytscha) to
combinations of dissolved copper and submerged structure in freshwater and seawater.
Frank Sommers
Dissolved copper is one of the more pervasive and toxic constituents of stormwater
runoff and is commonly found in stream, estuary, and coastal marine habitats of juvenile
salmon. While stormwater runoff does not usually carry copper concentrations high
enough to result in acute fish kills, exposure to sublethal concentrations of copper is
known to both impair salmon health (e.g. olfactory function) and alter behavior. To
evaluate behavior of juvenile Chinook to the presence of sublethal concentrations of
dissolved copper, I conducted laboratory studies using a multi-chambered experimental
tank. The circular tank was divided into six segments so that water flowed outward from
the center of the tank through each of the segments separately, yet fish could move freely
between them. The presence of individual fish in each of the segments was recorded at 3second intervals for one hour. The number of occurrences in each segment was counted
before and after introduction of a sublethal and environmentally realistic concentration of
dissolved copper (< 20 µg/L) to one of the segments (exposure segment). Use of
exposure segment by each fish was compared before and after introduction of dissolved
copper in both freshwater and seawater. To address whether use of preferred habitat is
altered by the presence of copper, experiments were also conducted with a submerged
structural element. Juvenile Chinook salmon avoided segments with dissolved copper in
both freshwater and seawater with no structure present. Fish preferred physical structure
in freshwater without copper but that preference disappeared when copper was present.
There was no preference for structure in seawater. The presence of sub-lethal levels of
dissolved copper altered the behavior of juvenile Chinook salmon in both freshwater and
seawater, this could potentially affect behaviors beneficial to growth, survival and
reproductive success.
TABLE OF CONTENTS
LIST OF FIGURES........................................................................................................v
LIST OF TABLES..........................................................................................................v
ACKNOWLEDGEMENTS ..........................................................................................vi
INTRODUCTION .........................................................................................................1
METHODS.....................................................................................................................6
Animals ....................................................................................................................6
Experimental Chamber ............................................................................................7
Copper Concentration ............................................................................................10
Evaluation of Acclimation .....................................................................................13
Data Collection and Analysis .................................................................................14
RESULTS.....................................................................................................................16
Structure .................................................................................................................16
Copper exposure without structure ........................................................................17
Copper exposure with structure .............................................................................18
DISCUSSION ..............................................................................................................19
BIBLIOGRAPHY ........................................................................................................25
APPENDIX 1 ...............................................................................................................29
iv
LIST OF FIGURES
Figure 1.
Overhead view of experimental chamber showing
segment number (1-6) ................................................................................8
Figure 2.
Overhead view of dye test showing containment of
treatment water within the exposure segment (1) ......................................9
Figure 3.
Overhead view of the exposure segment with added
physical structure .....................................................................................12
Figure 4.
Mean percent time spent (± se) by individual fish in
the exposure segment before exposure ....................................................17
Figure 5.
Mean percent time (± se) spent in exposure segments
before and during exposure ......................................................................18
LIST OF TABLES
Table 1.
Results of pre-experiment water samples ................................................11
v
ACKNOWLEDGEMENTS
I owe a great deal of thanks to many people who helped with this project, particularly
my readers David Baldwin (NOAA Fisheries/NWFSC/Ecotox) who provided a great deal
of technical support as well as patience, encouragement and advice, Maria Bastaki (The
Evergreen State College) for her excellent reviews and attention to detail and Amy Cook
(The Evergreen State College) who was willing to be my primary reader in spite of a very
busy schedule. The 2006 NWFSC Internal Grants program provided the funding to make
this study possible and the 2004 NOAA Advanced Studies Program provided the support
to begin the process. Special thanks The folks at NOAA Fisheries/NWFSC especially
Linda Rhodes for the encouragement to pursue the Advanced Studies Program and Tracy
Collier who was willing to give his support. Correigh Greene suggested the idea of
ecological traps that eventually led to the idea for this study. Vera Trainer gave great
encouragement and assistance writing the grant proposal. Thanks to Casey Rice, James
Meador, Julann Spromberg, Kathleen Neely and many others for help and feedback on
the transcript. Finally I am incredibly grateful for the love and support of my family,
Charlie and Arlene Sommers, Shizuku Fujita and Stephen Claussen.
vi
INTRODUCTION
Juvenile Chinook salmon depend heavily on estuarine habitats that provide food
resources for rapid growth and shelter from predators (Aitkin et al. 1998, Healy 1982,
Thorpe 1994). Loss and degradation of estuarine habitat is considered to be one factor
contributing to the decline of wild populations of Chinook salmon and other Pacific
salmonids throughout their range. Estuarine habitat is disproportionately affected by
urban and rural development which tends to be more concentrated in coastal areas (POC
2003). Non-point source pollution such as contaminated stormwater runoff is recognized
as being a significant and growing threat to aquatic species (Hart-Crowser 2007).
Dissolved copper is one of the more pervasive and toxic constituents of stormwater
runoff causing behavioral alterations at low levels to death at higher concentrations.
Copper is commonly found in streams, estuaries, and nearshore waters receiving
stormwater runoff.
Copper enters the aquatic environment from a variety of sources including treated
wood products, pesticides and anti-fouling paint from recreational vessels though metalcontaining dust shed from trucks and automobiles is the greatest contributor (Davis et al.
2001). Concentrations in stormwater runoff vary greatly depending on time of the year
and surrounding land use. A study on the Governor Albert D. Rosellini Bridge-Evergreen
point on SR-520 in Seattle, WA, which isolated stormwater runoff primarily to road
deposition on a busy highway found copper levels between 36 and 76.5 µg/L (King
1
County 2005). Other measurements of stormwater runoff from outputs in and around
Sinclaire/Dyes Inlets watersheds’ ranged from a median of 1-75 µg/L (Hart-Crowser
2007). A total of more than 100 metric tons of copper are estimated to enter Puget Sound
each year from stormwater runoff (Hart-Crowser 2007, King County 2005). With
increasing urbanization, it seems likely that animals that inhabit the estuarine and
nearshore waters, including listed species of Pacific salmon, will be exposed to copper.
Salmon encounter dissolved copper across a continuum of water conditions between
rearing in fresh water until entering the seawater environment. The toxicity of copper
varies depending on fish species, life stage and water chemistry, and the mechanism of
toxicity appears to be different in freshwater than in seawater (Blanchard and Grosell
2006, Grosell et al. 2007, Hansen et al. 1999). In freshwater, fish must take up sodium to
maintain a higher internal ionic concentration than the surrounding environment. Sodium
ions are primarily taken up by active transport across the gill epithelium and in the gut.
Acute copper exposure causes osmoregulatory dysfunction by interfering with ion
transport primarily at the gill epithelium leading to death at relatively low concentrations;
exposure to 26 µg/L was found to be the lethal concentration to 50% of test animals
(LC-50) over a 96 hour time period for Chinook salmon smolts (Chapman 1978).
Dissolved organic matter (DOM), pH, alkalinity, and hardness can reduce toxicity by
complexing with free ionic copper to form a variety of copper species, reducing levels of
Cu2+ and CuOH+, the forms of copper most toxic in fresh water or by competing with
ionic copper for sodium binding and uptake pathways on the gill. This relationship is
modeled with the Biotic Ligand Model (BLM), though the mechanism is complicated and
2
not fully understood (Di Toro et al. 2005, Niyogi and Wood 2004, Santore et al. 2001).
Relatively few studies have examined the effects of copper on fish in seawater and the
mechanism of toxicity and chemical interactions are less well understood than in
freshwater. In seawater as in freshwater, copper toxicity appears to involve ionoregulation
but the buildup of nitrogenous products in plasma may also contribute to toxicity along
with the disruption of Na+ and Cl- regulation (Blanchard and Grosell 2006, Grosell et al.
2007, Grosell et al. 2004a, Wilson and Taylor 1993). The abundance of inorganic ligands
in seawater is likely to make copper less bioavailable and be protective against acute
toxicity (Neff 2002). More importantly, in seawater the osmoregulatory balance is
reversed and fish are excreting ions that they would be taking up in freshwater. Fish are
more likely to encounter sublethal concentrations of copper in both freshwater and
seawater and the sublethal effects of copper are likely to be a more ecologically relevant
measure of the effects of copper on fish (Kleerekoper 1976).
Studies of the sub-lethal effects of contaminants on fish physiology in freshwater
have demonstrated copper-linked neurophysiological effects on the olfactory system in
salmonids resulting in loss of smell and olfactory-triggered predator avoidance behaviors
(Baldwin et al. 2003, City of San Jose 2005, McIntyre et al. 2008, Sandahl et al. 2007,
Sandahl et al. 2006) and neurotoxicity in zebrafish (Danio rerio) that targets lateral line
neurons (Linbo et al. 2009). These studies effectively link copper exposure to
physiological damage and related behavioral effects. Behavioral avoidance is a sensitive
indicator of exposure to copper and other heavy metals (Scott and Sloman 2004,
Woodward et al. 1995) and may manifest before any physiological damage occurs. Many
3
studies have demonstrated avoidance in freshwater (Atchison et al. 1987, Scott and
Sloman 2004), but only two have done so in seawater (Koltes 1985, Labenia et al. 2006).
Though copper may be acutely toxic to fish in seawater only at high concentrations,
studies in seawater have demonstrated avoidance of copper by fish at environmentally
relevant levels. The results of copper avoidance studies in seawater are consistent with
behavioral responses in freshwater studies; in a two-choice test chamber, cutthroat trout
avoided copper at the single test concentration of 40 µg/L in seawater (Labenia et al.
2006); hyperactivity and altered schooling behavior were observed in Atlantic silverside
at copper concentrations less than 100 µg/L in seawater, however the test chamber was
completely flooded with the exposure solution and did not physically allow avoidance so
the altered behavior may have been an avoidance response (Koltes 1985). Because of the
lack of studies examining the affects of copper on fish in seawater there exists a gap in
information about how copper may affect Chinook salmon in the estuarine and nearshore
environment.
Fish function in a complex environment and make choices about how to use that
environment such as tradeoffs between avoidance of predators and access to food
resources. Avoidance is likely to be the initial response to the presence of copper, and
may reduce or prevent further damage that could result from prolonged exposure. Few
studies have addressed avoidance in the field in freshwater and none have done so in
seawater. Results of field studies in freshwater are consistent with avoidance studies
conducted in the lab. In one study, Chinook salmon avoided a metal contaminated
tributary compared to one relatively free of metal contamination during spawning
4
migration (Goldstein et al. 1999), potentially reducing available spawning habitat for a
portion of the fish population. A related study (Woodward et al. 1997) found cutthroat
trout avoided a metals mixture simulating that found in the same system and that
avoidance was driven by levels of zinc and copper. Another study illustrates how the
avoidance of one stressor may override avoidance of metal contamination. Trout in a
Montana stream suffered damage from chronic metal exposure when favoring areas of
cooler water contaminated with sub-lethal levels of metals over warmer areas relatively
free of metal contamination (Harper et al. 2009). Metals in the cooler areas were above
concentrations that have been shown to elicit avoidance behavior but temperatures in the
cleaner areas may have reached nearly lethal levels. Behavior affects how fish use their
environment and has important ecological implications and avoidance of copper may
limit the use of otherwise suitable habitat. Alternately, preference for habitat may increase
the risk of exposure to contaminants. Estuaries and nearshore habitats are critical in the
life-cycle of many salmon, especially Chinook. Disruption of juvenile Chinook salmon
behavior in these environments from dissolved copper may have important implications
for health and survival.
To evaluate whether juvenile Chinook salmon may be affected by dissolved copper
concentrations typical of urban estuaries, I conducted laboratory experiments using a
multi-chambered experimental tank, similar to that used by Scarfe (1985). Distribution of
individual fish within the tank was compared in multiple trials before and after
introduction of dissolved copper in both freshwater and seawater. To evaluate interactive
5
effects of copper and physical habitat complexity on fish behavior, additional
experiments were conducted with a submerged structural element.
METHODS
Animals
Four thousand Fall Chinook salmon were acquired as eyed eggs from the University
of Washington’s School of Fisheries hatchery, (Seattle, WA, USA) in December 2006 and
transported to the NOAA Fisheries Mukilteo Field Facility in Mukilteo, WA for rearing
and experiments. Fish were fed standard commercial salmon pellets (Bio-Oregon,
Warrenton, OR, USA) and held in a 1.8 m diameter circular fiberglass tank on a
recirculating freshwater system supplied by carbon filtered city water (100-300 mg/L
total hardness as CaCO3, pH 7.1-7.3, temperature 8-12 °C, oxygen 9-12 mg/L) until
smolted over the week of 11 June 2007 (Meador et al. 2006). Forty fish to be used in the
freshwater portion of the experiment were transferred to an indoor circular 0.9 m
diameter tank on a separate recirculating freshwater system supplied by the same water
source on 8 May 2007, prior to smolting. The remaining fish were held on flow-through,
sand-filtered seawater, (salinity 30‰, pH 7.8, oxygen 9.5 mg/L, temperature 10-12 °C).
The recirculating system for the freshwater trials of the experiment was replenished from
city water passed through a 5µm filter and an activated carbon filter and chilled to
between 12-14 °C. Seawater for the experiment was sand filtered and passed through a 20
6
µm cartridge filter and 10 µm bag filter to remove particulates that would otherwise coat
the floor of the tank. Temperature was maintained at 10-12 °C.
Experimental Chamber
The experimental chamber consisted of a 1.8 m diameter 74 cm tall circular fiberglass
tank, divided into six segments with a central distribution manifold (Figure 1). Each
segment contained an inner distribution segment, and an outer experimental segment. The
distribution segments and dividers were made out of 3.2 mm Plexiglas and attached to the
circular tank using silicone glue. The distribution manifold was made from 200 mm
diameter PVC pipe to allow water to enter from the top and flow out through 1.3 mm,
CPVC pipe fittings into the distribution chamber at the head of each of six segments.
Water flow was directed into each experimental segment, to either a center section along
the bottom of the tank, or to the sides where sheets of twinwall polycarbonate sheet
created a laminar flow effect to help isolate water between segments while allowing fish
to swim freely around the tank (Figure 1). Copper solution was pumped into the
distribution chamber of the exposure segment (segment 1) so that it mixed with water
flowing into the segment, entering at the bottom center at the head of the experimental
portion of the exposure segment. Drains were placed at the center of the external wall of
each segment, approximately 31 cm above the bottom of the tank providing a water depth
of 35 cm. Adjustable 90° elbows leaving the drains on the outside of the tank, allowed
outflow of each segment to be regulated separately. Flows were adjusted so that the
7
exposure segment (segment 1) had negative pressure relative to its neighbors, allowing
incoming water to be contained within that segment (Figure 2).
Drain
2
Water Input
Drain
Flow
3
1
4
6
Exposure
substance
5
From Metering Pump
30 cm
Figure 1. Overhead view of experimental chamber showing segment numbers (1-6).
Double-dashed arrows indicate flow from dividers; single-dashed arrow indicates
direction of flow of copper solution. Flow and drains are radially symmetrical. Copper
solution was introduced in segment 1 only. Water input lines run over the top of the tank
on the borders of segment 3.
Isolation of flow within the exposure segment and between it and neighboring
segments was verified using a fluorescent Rhodamine water tracing dye (Bright Dyes
FTW Red 25liquid, Bright Dyes Miamisburg, OH). Dye was pumped into the exposure
tank through the same delivery system used to supply the copper solution and samples
collected over time to determine onset, saturation and dispersion. Dye traveled from the
8
input to the far wall of the segment in about 3 minutes and the segment was saturated by
7 minutes. Dye was well contained in the exposure segment.
2
3
1
4
6
5
Figure 2. Overhead view of dye test showing containment of treatment water within the
exposure segment (1). Distribution chambers are covered with fabric to eliminate
reflections. Tank segments are numbered.
A solution of 0.01 g/L, dissolved copper was added to the exposure segment at a rate
of 12.6 mL/min via a metering pump (Liquid Metronics Incorporated, model LMIA11).
Delivery of the copper solution at this rate provided a nominal concentration of 17 µg/L
for one-sixth of the total flow of the tank. Total flow into the experimental tank was 45
L/min and flow out of the exposure segment was a minimum of 8 L/min. The dissolved
copper solution was prepared before each experimental trial by adding 10 mL/L to dH2O
of a stock solution of 1 g/L Cu (2.68 g/L CuCl2.2H2O), which was made up weekly. The
9
delivery line from the copper solution source was purged daily and fresh solution run
through the line.
Copper Concentration
Samples of exposure solutions were drawn from the exposure tank at 0, 10, and 30
minutes. Samples were taken from the center of each segment and the border between the
exposure segment (segment 1) and adjacent segments 2 and 6. Single samples were taken
at each time point. Samples were submitted to an outside laboratory (Frontier
Geosciences, Seattle, WA, USA) for analysis of total dissolved copper by inductively
coupled plasma mass spectrometry (detection limit 0.1µg/L in freshwater and 0.02 µg/L
in seawater). In freshwater, total dissolved copper ranged from 2.6-3.6 µg/L at the various
sampling locations at time zero. At 30 minutes, the exposure segment (segment 1)
reached a concentration of 7.9 µg/L while surrounding segments remained low. In
seawater, segment one had 0.4 µg/L dissolved copper at time zero and 16.9 µg/L at 30
minutes (Table 1). The higher time zero levels of copper in the freshwater portion of the
experiment may be due to the recirculating system that provided water for the experiment
as well as fish held in a separate tank for use in the experiment. Trace amounts of copper
from food as well as chillers and pumps may have caused the elevated time zero values.
Levels of copper at the 10 and 30 minute sampling points in freshwater were roughly half
of those in seawater. This may be due to insufficient priming of the pump delivering the
copper solution at the time of the sampling.
10
Table 1. Results of Pre-experiment water samples
Copper Concentration (µg/L) by Segment
Medium
Time
1
1-2
1-6
2
3
4
5
6
Freshwater
0
2.6
NT
NT
2.6
3.0
2.8
2.7
2.7
10
3.5
5.9
5.2
2.5
2.6
2.7
2.6
2.7
30
7.9
7.9
4.1
3.9
2.7
2.7
2.7
3.2
0
0.41
NT
NT
NT
NT
NT
NT
NT
10
7.32
0.88
2.91
0.42
0.39
0.38
0.38
0.95
30
16.9
5.75
3.74
1.15
0.43
0.42
0.49
1.68
Seawater
NT indicates samples not taken. 1-2 and 1-6 indicates samples taken at the borders of
adjacent segments.
The freshwater portion of the experiment was conducted between 14 August 2007, and
6 September 2007. Each replicate trial was run using a single fish placed into the
experimental chamber. The fish was given an hour to acclimate before beginning the trial.
Fish were evaluated after acclimation (see Evaluation of Acclimation below) before
beginning the experiment. If fish behavior was acceptable, frame capture began. At 20
minutes, the metering pump was started and dissolved copper delivered to the exposure
chamber. The run was concluded after 60 minutes. The fish was euthanized and the tank
was flushed twice before the next run.
For trials where fish were kept on a freshwater recirculating system, a separate drain
system was constructed so that the drain was switched immediately before adding copper
to the experimental chamber, so that copper contaminated water would not return to the
11
recirculating system. After each run, the system was replenished with filtered, dechlorinated city water and the drain was switched back to return to the recirculating
system. Seawater trials were conducted using a flow-through system.
For trials with added physical structure, a bundle of 12.7 mm PVC pipes, forming a
loose teepee shape was placed in the exposure segment. The structure took up most of the
segment but did not extend beyond the segment (Figure 3). Otherwise, trials were run
identically to those without structure. Seven trials were completed with and without
structure.
2
1
6
Figure 3. Overhead view of the exposure segment with added physical structure.
Methods were essentially the same in seawater as in freshwater except that
acclimation time was increased to 2 hours, the number of trials was increased to 15 for
trials with structure and 16 without structure, and the experimental tank was drained and
refilled once between trials.
12
Evaluation of Acclimation
When placed into the experimental chamber, fish adopted one of three behaviors:
hiding, rapid swimming, or exploration. When hiding, fish dropped to the bottom of the
tank usually near a wall and stopped moving, sometimes for hours. Some fish would
begin rapidly swimming around the tank, usually around the perimeter and at the surface.
This appeared to be a panic swim as the fish paid no attention to structure or the presence
of copper and seemed to ignore their surroundings. When exploring, fish would move
around the tank, often in mid-water but varying between the bottom and the surface
between sections occasionally reversing direction. With structure present, some fish
would immediately dart into structure while others would approach slowly before
entering structure. Often fish would initially display hiding or rapid swimming behavior
before exploring. Hiding and rapid swimming responses were deemed unsuitable for the
experiment because while displaying these behaviors, fish did not appear to be interacting
with their environment and no useful information could be gained from their inclusion.
Fish which were not displaying exploratory behavior by the end of the acclimation period
were rejected.
Acceptance criteria in trials without structure:
1. Fish must complete at least one full circuit of the chamber.
2. Fish must be moving enough to cross border between segments at least once every five
minutes (except when structure was present).
13
3. Fish must spend at least half of the time in mid water column.
4. Fish must have been displaying this behavior during the last ten minutes of the
acclimation period.
In trials with structure, the criteria needed to be different because of the attraction of
structure and criterion 2 was not used. Not all fish exhibited behavior that was considered
acceptable for use in the experiment. In fresh water, 11 fish were rejected out of 25
attempts. In seawater 55 fish were rejected out of 86 total attempts. In the seawater
portion, acclimation time was increased to 2 hours. This allowed more time for fish to
settle into exploratory behavior and many of the fish that were accepted would have been
rejected if the acclimation time were 1 hour. Published avoidance studies with fish are
inconsistent with their reporting of rejection criteria and rates but this study is
comparable to those published in Labenia, 2007. Additionally, it is difficult to compare
the rejection rates in this study with other published avoidance studies because of the
variation in equipment, species and rejection criteria.
Data Collection and Analysis
Fish movement in the experimental chamber was recorded using a digital camera with
a vari-focal lens (DFK 21F04, Unibrain Inc., San Ramon, CA, USA). Images of the fish
were captured using time-lapse software on a laptop computer. Overhead lighting was
provided by seven 60-watt floodlights. Images were analyzed for presence or absence of
fish in each segment at each frame capture, using image analysis software (Image-J, NIH)
14
and manually verified. Individual trials had a one-hour duration and images were
captured at 3-second intervals producing 1200 images for each trial. These intervals were
frequent enough to capture the movement of fish in the experimental chamber and
provided sufficient duration to capture overall fish behavior based on initial observations
of fish performance. Each trial was analyzed in two periods: the 20 min before copper
was introduced into the exposure segment, and the final 30 minutes, after introduction of
the copper solution, which was counted as the exposure portion. The first ten minutes
after exposure began were not analyzed while copper permeated the exposure segment.
Frames were captured at 400 time points in the pre-copper portion and 600 in the
exposure portion.
The percentage of time spent in the exposure segment was calculated for both precopper and post-copper portions of individual trials by dividing the number of frames
where a fish appeared in the exposure segment by the total number of frames for that
portion. For all analyses, percentages were arcsin transformed before statistical analysis
(Zar 1984). Baseline use of all segments in the experimental tank was analyzed by
comparing all non-structure, pre-copper trials to a theoretical even distribution of 16% in
each segment by one-sample t-test. The attraction to structure was analyzed by two
methods: First, by comparing pre-copper use of the structure segment (segment 1)
between trials using structure and those not using structure by unpaired t-test. Second,
percentage of time spent in the test segment was compared to a hypothesized even
distribution of 16% with a one-sample t-test. Response to copper was similarly analyzed
by two methods: First, using paired t-tests for percentage of time spent in the exposure
15
segment before and during copper exposure. Second, percentage of time spent in the test
segment was compared to a hypothesized even distribution of 16% with a one-sample ttest. Data were analyzed using Statview statistical software (Abacus Concepts, Inc.
Berkeley, CA).
RESULTS
On average, fish did not display bias in the mean percent time spent in the exposure
segment without structure and without copper when compared to a hypothesized even
distribution between all segments of 16% (one sample t-test, n = 22, p = 0.14).
Structure
In freshwater without copper, fish were present in the exposure segment (segment 1)
over four times as often (76% versus 14%, p = 0.016) when structure was present than
when structure was not present (Figure 4a). Without structure, presence in the exposure
segment was not different (p = 0.34) from a hypothesized mean of 16% but was greater
(p = 0.02) when structure was present (Figure 4a). In seawater, the presence of structure
did not affect the likelihood that fish would be found in the exposure segment (12% with
structure, 15% without, p = 0.37, Figure 4b). Presence in the exposure segment was not
different from a hypothesized mean of 16% whether structure was present or not (no
structure p = 0.30, with structure p = 0.15).
16
a
b
Freshwater
100
100
80
80
60
60
40
40
20
20
0
0
No Structure
Structure
Seawater
No Structure
Structure
Figure 4. Mean percent time spent (± se) by individual fish in the exposure segment
before exposure. Individual treatments were compared to a theoretical even distribution
between all segments of 16% (dashed line).
Copper exposure without structure
In freshwater without structure, fish spent less time in the exposure segment (12%
versus 14%) when copper was present than before copper was present (Figure 5a, paired
t-test, n = 6, p = 0.05). Time spent in the exposure segment was similar to a hypothesized
mean of 16% before copper was introduced (p = 0.34) but less than 16% (p = 0.05) while
copper was present.
In seawater without structure, fish were present in the exposure segment 15% of the
time before the introduction of copper and 11% when copper was present (Figure 5b,
paired t-test, n = 16, p = 0.02). Time spent before copper was introduced was not different
from a hypothesized mean of 16% (p = 0.30) but declined to less than 16% when copper
was present (p = 0.0004).
17
Figure 5. Mean percent time (± se) spent in exposure segments before and during
exposure. Individual treatments were compared to a theoretical even distribution between
all segments of 16% (dashed line). Note different scale on y axis of panel b.
Copper exposure with structure
While fish were more likely to be in the exposure segment when structure was present
in freshwater (76%), the preference for structure was not apparent when copper was
present (23%, n = 7, p = 0.06, figure 5c). Presence in the exposure segment differed from
18
the hypothesized mean of 16% (One sample t-test n = 7, p = 0.02) before copper, and did
not differ (One sample t-test n = 7, p = 0.44) when copper was present.
In seawater with structure present, fish were no more likely to be found in the segment
with structure than in segments without and were no less likely to be in the exposure
segment with copper present than without (12% with copper, 12% without, paired t-test,
n = 15, p = 0.86, Figure 5d). Presence in the exposure segment was not different from a
hypothesized mean of 16% whether copper was present or not (p = 0.15 with copper, p =
0.23 without).
DISCUSSION
The findings of this study indicate that sublethal concentrations of copper altered the
behavior of juvenile Chinook salmon in seawater and freshwater at environmentally
realistic levels at a size and age when fish enter the estuarine environment and transition
from freshwater to seawater. Also, the use of physical structure in freshwater was altered
by the presence of copper. The use of submerged structure by pre-smolt salmon has been
well documented primarily as the use of large woody debris (Quinn and Peterson 1996,
Roni et al. 2002). In this study, pre-smolt juvenile fish also showed a preference for
submerged structure in freshwater in the laboratory but that preference disappeared when
copper was elevated. Fish were not attracted to submerged structure in seawater which
may indicate behavioral difference between pre- and post-smolt life stages.
19
Water chemistry can be protective against acute toxicity but results of this study
suggest that avoidance behavior is less affected by water chemistry than acute toxicity.
There is a great deal of experimental evidence supporting the general effectiveness of the
biotic ligand model which predicts the effect of water chemistry on acute toxicity in
freshwater (Niyogi and Wood 2004, Paquin et al. 2002). The model predicts the
chemistry of seawater to be highly protective against lethal copper toxicity and
experimental evidence supports that prediction. The total copper LC50 for coho smolts is
63 µg/L in freshwater and 601 µg/L in seawater (EPA 2007). Whether the mechanism for
this difference lies in the competition for binding sites on the gill by other ligands as in
the biotic ligand model, the difference in osmotic physiology between freshwater and
seawater, or some other mechanism is not clear. In both freshwater and seawater
however, dissolved copper will elicit an avoidance response at concentrations below
those reported for acute toxicity in a variety of fish species (Atchison et al. 1987,
Blanchard and Grosell 2006, Grosell et al. 2007, Grosell et al. 2004b). In freshwater, the
Lowest Observed Effect Concentration (LOEC) for copper avoidance was reported for
rainbow trout at 0.1 µg/L (Folmar 1976), though most studies report LOECs for
salmonids in the 2-7 µg/L range (Atchison et al. 1987). Threshold avoidance levels have
not been explicitly studied in seawater and at 18 µg/L this study finds avoidance at a level
lower than found in previous studies (Koltes 1985, Labenia et al. 2006). While water
chemistry is protective against acute copper toxicity, it may have less of an effect on
sensory function and avoidance.
20
The two endpoints are dissimilar in terms of the physical mechanism of copper
toxicity; acute toxicity, binding of copper to sites on the gill and attenuation of toxicity by
ligands described by water chemistry and dissolved organic matter (DOM); and sensory
toxicity as seen in studies of the effects of copper on olfactory cells and sensory cells of
the lateral line. Although avoidance behavior has not been mechanistically linked to
physiological toxicity, it seems likely to be linked to sensory input like olfaction or the
lateral line. The specific sensory input that elicits avoidance has not yet been identified
but the loss of alarm response behavior has been shown to correspond to the loss of
olfactory response to alarm chemicals (Sandahl et al. 2006). Additionally, the effect of
water chemistry on avoidance behavior appears to be mechanistically more similar to
sensory toxicity than it is to acute toxicity in that water chemistry is only minimally
protective against sensory toxicity (Linbo et al. 2009, McIntyre et al. 2008), though the
effects of water chemistry on avoidance behavior have not been studied specifically.
While the consequences of acute toxicity are obvious, those of avoidance behavior are
less so. The alteration in the use of submerged structure in this study may have
implications for the effect of copper avoidance on the utilization of estuarine critical
habitat.
The relationship between submerged structure primarily as large woody debris
(LWD) and juvenile salmon has been extensively studied in freshwater and the preference
for structure in this study corroborates this work. Few studies have focused on the use of
submerged structure by juvenile salmon in the estuary and nearshore after the transition
to seawater but the importance of large woody debris appears to weaken as fish enter
brackish water and ocean-type Chinook in particular make extensive use of shallow
21
shoreline, mudflats, and seagrass beds (Simenstad 1982, Wick 2002). In this study, the
loss of attraction to submerged structure between freshwater and seawater seems to
corroborate this. While this study was unable to determine whether the presence of
dissolved copper would cause fish to abandon attractive habitat in seawater using
submerged structure, it is still possible that the presence of copper would cause fish to
avoid estuarine and nearshore habitats that they would otherwise use. The modification
of behavior shown by individual fish in this study could impair their ability to use their
environment, effectively resulting in a loss of habitat. Additional research is needed to
understand how the presence of copper affects fish in the field as well as at what levels
the effects become apparent.
Increasing urbanization in the Puget Sound region makes it likely that copper will
continue to be a concern. The evidence from the few published studies that exist indicate
that the effects of copper in the estuary and nearshore may cause sublethal effects in fish
and deserves further study. Future work should include the determination of copper
avoidance LOECs for salmon at a range of salinities including full strength seawater. The
results of this study show effects at the lowest level of copper in seawater yet published
and would likely find effects at lower concentrations. Damage to the olfactory sensory
system is another sublethal effect of copper demonstrated in freshwater that may also
occur in seawater. The ability to smell is disrupted after exposure to dissolved copper in
freshwater (Baldwin et al. 2003, McIntyre et al. 2008, Sandahl et al. 2007, Sandahl et al.
2006, Scott and Sloman 2004). Loss of smell can affect imprinting on home streams
22
(Wisby WJ and AD 1954) as well as a variety of olfactory-mediated behavior such as
predator avoidance (Brown 2003) and cohort identification (Quinn and Busack 1985).
The methods used in this study, though based on techniques used in other behavioral
experiments, are unique to this experiment. These methods proved effective by providing
results in freshwater that are comparable to other studies of copper avoidance in
freshwater. There are however areas where the method used could be improved upon. At
16% the theoretical likelihood of presence in a given segment provides a low baseline for
detecting avoidance and is better suited to detecting preference for a given segment.
Flooding two or three segments with the test substance would provide a greater baseline
usage (i.e. 50% rather than 16%) and increase the ability to detect a reduction in usage.
Additionally, quantifying additional behavioral changes might increase the ability to
detect avoidance. Fish entering the exposure segment before sensing the copper and
either reversing direction to avoid the copper or accelerating through the segment, are
counted as being in the segment while actively leaving it. Measuring reversal of
direction or change in speed would make the experiment more sensitive to behavioral
changes from the presence of copper though would make the process of collecting data
considerably more complicated and laborious. Also, a relatively high number of fish
were rejected from the experiment because of behavior that may have been caused by the
stress of the experimental chamber. Experimental evidence exists that suggests that the
presence of multiple fish may reduce some of this stress (Grand and Dill 1999). The use
of multiple test subjects may reduce the rejection rate and behavior may become more
natural. Finally, field studies of the effects of copper on fish are almost nonexistent due
23
most likely to the amount of effort and the logistical difficulties of performing field
studies with toxic substances but efforts should be made to verify laboratory findings in
the field.
In conclusion, the results of the present study contribute to a sparse body of work
focusing on the behavioral effects of copper on fish in seawater. Dissolved copper can
elicit avoidance behavior in salmonids in seawater at levels below those commonly found
in urban streams feeding into Puget Sound. Further, the use of preferred habitat can be
disrupted at a life stage where estuarine habitat is likely critical to healthy salmon
populations. All Puget Sound estuaries and coastal marshes have been classified as
critical habitat to threatened Puget Sound salmon species. With growing urbanization in
the Puget Sound region, toxics like copper will continue to enter streams flowing into
estuarine habitat at a time when coastal development has reduced the area of estuaries
and salt marshes to 40% of historic levels (Collins BD 2005, Hart-Crowser 2007).
24
BIBLIOGRAPHY
Aitkin, J.K., U.S. Fish and Wildlife Service. Western Washington Office. Aquatic
Resources Division., and Puget Sound Estuary Program. 1998. The importance of
estuarine habitats to anadromous salmonids of pacific northwest a literature
review. U.S. Fish & Wildlife Service, Western Washington Office, Aquatic
Resources Division,, Lacey, Wash. (510 Desmond Dr., SE, Suite 102, Lacey
98503-1273).
Atchison, G.J., Henry, M.G., and Sandheinrich, M.B. 1987. Effects of Metals on Fish
Behavior - a Review. Environmental Biology of Fishes 18(1): 11-25.
Baldwin, D.H., Sandahl, J.F., Labenia, J.S., and Scholz, N.L. 2003. Sublethal effects of
copper on coho salmon: Impacts on nonoverlapping receptor pathways in the
peripheral olfactory nervous system. Environmental Toxicology and Chemistry 22
(10): 2266-2274.
Blanchard, J., and Grosell, M. 2006. Copper toxicity across salinities from freshwater to
seawater in the euryhaline fish Fundulus heteroclitus: Is copper an ionoregulatory
toxicant in high salinities? Aquat. Toxicol. 80(2): 131-139.
Brown, G.E. 2003. Learning about danger: chemical alarm cues and local risk assessment
in pre fishes. Fish and Fisheries 4(3): 227-234.
Chapman, G.A. 1978. Toxicieties of Cadmium, Copper, and Zinc to Four Juvenile Stages
of Chinook Salmon and Steelhead. Trans. Am. Fish. Soc. 107(6): 841-847.
City of San Jose, W.I.S., Watershed Protection, Environmental Services Department.
2005. Review fo the Efffects of Copper on Salmonid Olfaction, San Francisco
Bay Copper Site-Specific Objective Workgroup, San Jose.
Collins BD, S.A. 2005. Historical reconstruction, classification and change analysis of
Puget Sound tidal marshes., Report to Washington Department of Natural
Resources, Aquatic Resources Division, Olympia, WA.
Davis, A.P., Shokouhian, M., and Ni, S.B. 2001. Loading estimates of lead, copper,
cadmium, and zinc in urban runoff from specific sources. Chemosphere 44(5):
997-1009.
Di Toro, D.M., McGrath, J.A., Hansen, D.J., Berry, W.J., Paquin, P.R., Mathew, R., Wu,
K.B., and Santore, R.C. 2005. Predicting sediment metal toxicity using a sediment
biotic ligand model: Methodology and initial application. Environmental
Toxicology and Chemistry 24(10): 2410-2427.
EPA, U.S. 2007. Aquatic Life Ambient Freshwater uality Criteria - Copper. Edited by
O.o.W. U.S. Environmental Protection Agency, Office of Sceince and
Technology, Washington, D.C.
25
Folmar, L.C. 1976. Overt avoidance reaction of rainbow trout fry to nine herbicides. Bull.
Environ. Contam. Toxicol. 15(5): 509-514.
Goldstein, J.N., Woodward, D.F., and Farag, A.M. 1999. Movements of adult Chinook
salmon during spawning migration in a metals-contaminated system, Coeur
d'Alene River, Idaho. Trans. Am. Fish. Soc. 128(1): 121-129.
Grand, T.C., and Dill, L.M. 1999. The effect of group size on the foraging behaviour of
juvenile coho salmon: reduction of predation risk or increased competition?
Anim. Behav. 58(2): 443-451.
Grosell, M., Blanchard, J., Brix, K.V., and Gerdes, R. 2007. Physiology is pivotal for
interactions between salinity and acute copper toxicity to fish and invertebrates.
Aquat. Toxicol. 84(2): 162-172.
Grosell, M., McDonald, M.D., Walsh, P.J., and Wood, C.M. 2004a. Effects of prolonged
copper exposure in the marine gulf toadfish (Opsanus beta) II: copper
accumulation, drinking rate and Na super(+)/K super(+)-ATPase activity in
osmoregulatory tissues. Aquat. Toxicol. 68(3): 263-275.
Grosell, M., McDonald, M.D., Wood, C.M., and Walsh, P.J. 2004b. Effects of prolonged
copper exposure in the marine gulf toadfish (Opsanus beta). I. Hydromineral
balance and plasma nitrogenous waste products. Aquat. Toxicol. 68(3): 249-262.
Hansen, J.A., Marr, J.C.A., Lipton, J., Cacela, D., and Bergman, H.L. 1999. Differences
in neurobehavioral responses of chinook salmon (Oncorhynchus tshawytscha) and
rainbow trout (Oncorhynchus mykiss) exposed to copper and cobalt: Behavioral
avoidance. Environmental Toxicology and Chemistry 18(9): 1972-1978.
Harper, D.D., Farag, A.M., Hogstrand, C., and MacConnell, E. 2009. Trout Density and
Health in a Stream with Variable Water Temperatures and Trace Element
Concentrations: Does a Cold-Water Source Attract Trout to Increased Metal
Exposure? Environmental Toxicology and Chemistry 28(4): 800-808.
Hart-Crowser, I., Washington Department of Ecology;, U.S. Environmental Protection
Agency;, and Puget Sound Partnership.,. 2007. Phase 1: Initial Estimate of Toxic
Chemical Loadings to Puget Sound, Washington State Department of Ecology,
Olympia, Washington.
Healy, M.C. 1982. Estuarine comparisons. Academic Press, New York, Ny (USA).
King County. 2005. Highway 520 Bridge Storm Water Runoff Study. Edited by W.a.L.R.
Division. Dean Wilson, , Seattle.
Kleerekoper, H. 1976. Effects of Sublethal Concentrations of Pollutants on Behavior of
Fish. Journal of the Fisheries Research Board of Canada 33(9): 2036-2039.
Koltes, K.H. 1985. Effects of Sublethal Copper Concentrations on the Structure and
Activity of Atlantic Silverside Schools. Trans. Am. Fish. Soc. 114(3): 413-422.
Labenia, J.S., Baldwin, D.H., French, B.L., Davis, J.W., and Scholz, N.L. 2006.
Behavioral impairment and increased predation mortality in cutthroat trout
exposed to carbaryl. Marine Ecology-Progress Series 329: 1-11.
Linbo, T.L., Baldwin, D.H., McIntyre, J.K., and Scholz, N.L. 2009. Effects of water
hardness, alkalinity, and dissolved organic carbon on the toxicity of copper to the
26
lateral line of developing fish. Environmental Toxicology and Chemistry preprint
(2009): 0000-0000.
McIntyre, J.K., Baldwin, D.H., Meador, J.P., and Scholz, N.L. 2008. Chemosensory
deprivation in juvenile coho salmon exposed to dissolved copper under varying
water chemistry conditions (vol 42, 1352, 2008). Environmental Science &
Technology 42(17): 6774-6775.
Meador, J.P., Sommers, F.C., Ylitalo, G.M., and Sloan, C.A. 2006. Altered growth and
related physiological responses in juvenile Chinook salmon (Oncorhynchus
tshawytscha) from dietary exposure to polycyclic aromatic hydrocarbons (PAHs).
Canadian Journal of Fisheries and Aquatic Sciences 63(10): 2364-2376.
Neff. 2002. Bioaccumulation in Marine Organisms-Effect of Contaminants from oil Well
Produced Water. Battelle, Coastal Resources and Environmental Management,
Duxbury, Massachusetts 02332, USA.
Niyogi, S., and Wood, C.M. 2004. Biotic Ligand Model, a Flexible Tool for Developing
Site-Specific Water Quality Guidelines for Metals. Environ. Sci. Technol. 38(23):
6177-6192.
Paquin, P.R., Gorsuch, J.W., Apte, S., Batley, G.E., Bowles, K.C., Campbell, P.G.C.,
Delos, C.G., Di Toro, D.M., Dwyer, R.L., Galvez, F., Gensemer, R.W., Goss,
G.G., Hogstrand, C., Janssen, C.R., and et al. 2002. The biotic ligand model: a
historical overview. Comp. Biochem. Physiol., C 133(1-2): 3-35.
POC, P.O.C. 2003. 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.
Quinn, T.P., and Busack, C.A. 1985. Chemosensory Recognition of Siblings in Juvenile
Coho Salmon (Oncorhynchus-Kisutch). Anim Behav 33(Feb): 51-56.
Quinn, T.P., and Peterson, N.P. 1996. The influence of habitat complexity and fish size on
over-winter survival and growth of individually marked juvenile coho salmon
(Oncorhynchus kisutch) in Big Beef Creek, Washington. Canadian Journal of
Fisheries and Aquatic Sciences/Journal Canadien des Sciences Halieutiques et
Aquatiques. Ottawa. 53(07): 1555-1564.
Roni, P., Beechie, T.J., Bilby, R.E., Leonetti, F.E., Pollock, M.M., and Pess, G.R. 2002. A
Review of Stream Restoration Techniques and a Hierarchical Strategy for
Prioritizing Restoration in Pacific Northwest Watersheds. N. Am. J. Fish. Manage.
22(1): 1-20.
Sandahl, J.F., Baldwin, D.H., Jenkins, J.J., and Scholz, N.L. 2007. A sensory system at
the interface between urban stormwater runoff and salmon survival.
Environmental Science & Technology 41(8): 2998-3004.
Sandahl, J.F., Miyasaka, G., Koide, N., and Ueda, H. 2006. Olfactory inhibition and
recovery in chum salmon (Oncorhynchus keta) following copper exposure.
Canadian Journal of Fisheries and Aquatic Sciences 63(8): 1840-1847.
Santore, R.C., Di Toro, D.M., Paquin, P.R., Allen, H.E., and Meyer, J.S. 2001. Biotic
ligand model of the acute toxicity of metals. 2. Application to acute copper
27
toxicity in freshwater fish and Daphnia. Environmental Toxicology and Chemistry
20(10): 2397-2402.
Scarfe, A.D., Steele, C.W., and Rieke, G.K. 1985. Quantitative Chemobehavior of Fish an Improved Methodology. Environmental Biology of Fishes 13(3): 183-194.
Scott, G.R., and Sloman, K.A. 2004. The effects of environmental pollutants on complex
fish behaviour: integrating behavioural and physiological indicators of toxicity.
Aquat. Toxicol. 68(4): 369-392.
Simenstad, C.A., K.L. Fresh, and E.O. Salo. 1982. The role of Puget Sound and
Washington coastal estuaries in the life history of Pacific salmon: An
unappreciated function. . In The Estuary As A Filter. Edited by V. Kennedy.
Academic Press, Orlando, Florida. pp. 343-364
Thorpe, J.E. 1994. Salmonid fishes and the estuarine environment. Estuaries 17(1A):
76-93.
Wick, A.J. 2002. Ecological function and spatial dynimcs of large woody debris in
oligohaline-brackish estuarine sloughs for juvenile Pacific salmon. M.S., School
of Aquatic and Fishery Sciences, University of Washington, Seattle.
Wilson, R.W., and Taylor, E.W. 1993. Differential responses to copper in rainbow trout
(Oncorhynchus mykiss) acclimated to sea water and brackish water. J. Comp.
Physiol., B. 163(3): 239-246.
Wisby WJ, and AD, H. 1954. Effect of occlusion on migrating silver salmon
(Onchorhynchus kisutch). J. Fish REs. Board Can. 11: 472-478.
Woodward, D.F., Goldstein, J.N., Farag, A.M., and Brumbaugh, W.G. 1997. Cutthroat
trout avoidance of metals and conditions characteristic of a mining waste site:
Coeur d'Alene River, Idaho. Trans. Am. Fish. Soc. 126(4): 699-706.
Woodward, D.F., Hansen, J.A., Bergman, H.L., Little, E.E., and Delonay, A.J. 1995.
BROWN TROUT AVOIDANCE OF METALS IN WATER CHARACTERISTIC
OF THE CLARK-FORK RIVER, MONTANA. Canadian Journal of Fisheries and
Aquatic Sciences 52(9): 2031-2037.
Zar, J.H. 1984. Biostatistical Analysis. Prentice-Hall, Englewood Cliffs, New Jersey
07632.
28
APPENDIX 1
Percentage time spent by individual fish in exposure segment 1.
Date
Freshwater
8/28/07
8/30/07
9/1/07
9/4/07
9/5/07
9/6/07
Seawater
10/1/07
10/3/07
10/11/07
10/12/07
10/17/07a
10/17/07b
10/30/07
10/31/07
11/1/07a
11/1/07b
11/2/07
11/29/07
11/30/07a
11/30/07b
12/4/07
12/5/07
No Structure
No Copper Copper
Date
Structure Present
No Copper Copper
10.0
16.0
18.8
17.5
9.0
14.0
7.5
14.0
18.2
12.8
9.8
9.8
8/14/07
8/16/07
8/17/07
8/20/07
8/24/07
8/25/07
8/26/07
49.0
32.8
12.5
91.8
88.5
47.3
100.0
17.7
6.0
16.8
29.8
21.8
62.2
0.0
12.0
12.3
19.0
5.5
11.5
8.5
15.0
22.5
17.3
16.3
14.8
18.3
9.8
18.8
8.3
23.0
12.7
6.5
7.2
1.2
17.2
15.0
10.0
14.0
14.5
15.5
10.2
9.0
5.8
16.5
7.8
10.0
10/18/07
10/19/07a
10/19/07b
10/22/070
10/24/07
10/25/07
10/26/07
11/5/07
11/8/07
11/9/07a
11/9/07b
11/14/07
11/16/07
11/20/07
11/26/07
13.3
6.3
16.0
1.3
41.5
12.5
4.8
9.0
15.5
6.5
19.8
3.8
15.0
4.3
8.8
0.2
13.2
18.2
5.2
51.7
15.7
0.8
11.2
13.5
6.7
14.8
2.8
10.8
2.3
5.7
29
30
