Assessment of chronic toxicity of upper Columbia River sediments to early life stage
white sturgeon (Acipenser transmontanus)
David W. Vardy1, Jon A. Doering1, Shawn C. Beitel1, Brett T. Tendler1, Adam Ryan2 , Robert Santore2, John P. Giesy1,3,4,5, Markus Hecker1,6,7
1Toxicology
Sampling Sites and Sediments
Sediments were collected from seven locations along the upper Columbia River (Map 1).
In addition, a water only (H2O) and an artificial sediment (CTRL) were included as controls.
Two reference sites, Lower Arrow Lake (LALL) and Genelle (GE), located upstream of a
metallurgical facility in Trail, B.C., Canada (Map 1B), and five locations between the transboundary reach of Canada and the USA, and Kettle falls, USA, representing a sediment
exposure gradient for contaminants of potential concern, were sampled (Map 1C).
Specifically, sediments were collected at Deadman’s Eddy (DE), Northport (NP), Little
Dalles, (LD), Upper Marcus Flats (UMF), and Lower Marcus Flats (LMF). Note: DE was
beach sediment that was collected above the water line due to logistical restraints.
[B]
[A]
[C]
[A]
Cumulative Survival (%)
White Sturgeon Early Life Stage
[B]
Passive Diffusive Gradient Thin Films (DGTs)
Fig 1. [A] Flow-through experimental exposure system, [B]
Experimental exposure arrangement, [C] Exposure
chamber with sediment and sturgeon.
Days of Exposure
Passive Diffusive Sampling Devices (Peepers)
Suction Sampling Devices (Airstones)
Water chemistry analyses
• Basic water quality parameters (temp., pH, dissolved oxygen, conductivity, and hardness) were
measured daily. Alkalinity, ammonia, nitrate, and nitrite were measured once a week, at minimum.
• Metal, cation/anion, and DOC analyses were performed weekly by inductively coupled plasma
mass spectrometry (ICP-MS), ion chromatography, and infrared detection, respectively, by
Columbia Analytical Services, Keslo, WA, USA.
Sediments
Data analyses
• Biological endpoints included mortality and body condition, and were assessed by Kaplan-Meier
survival analysis and Fulton's condition factor (K).
• ANOVA and Dunnett’s Tests were used to assess statistical significance between treatment and
reference groups.
Biotic Ligand Model
The Biotic Ligand Model (BLM), a metal speciation and predictive toxicity model, can be used to
predict metal speciation for various metals as well as toxicity to a number of aquatic organisms
depending on ambient water quality parameters (BLM, 2007). The most sensitive BLM parameter files
that were developed for previous work with white sturgeon (Vardy et al. 2011) were used to make
predictions for 4 metals (Cu, Cd, Zn, and Pb) based on water quality values obtained from the
different matrices in the present study. Toxic units (TUs) were then calculated by dividing the
measured metal concentrations in the matrices by the associated BLM prediction.
Metal Analyses
• Concentrations of metals fluctuated throughout the experiment within the matrices of the various
exposure sediments.
• There was a trend towards greater concentrations of metals in PW versus SWI and OW. Greatest
Cu concentrations were measured in systems containing DE, NP, LD, and LMF sediments while no
such trends were observed for the other metals (Fig 3).
Cd Concentrations in Matrices
Cu Concentrations in Matrices
[C]
LALL
OW
SWI
1
PW (1 cm)
Cadmium (µg/L)
Copper (µg/L)
DE
10
0.1
SWI
0.01
H2O CTRL LALL GE
DE
NP
LMF
Experimental Design
• WS embryos were obtained from the Kootenay Trout Hatchery, B.C., Canada.
• Newly hatched larvae were exposed to sediments and controls for 60d in flow-through
experimental chambers (Fig 1).
• Three different waters were sampled weekly for assessment of exposure: overlying
water (OW), sediment-water interface water (SWI), and porewater (PW), by use of
syringes, suction devices, and peepers and diffusive gradient thin films (DGTs),
respectively (Fig 2).
NP
10
1000
1
100
OW
SWI
0.1
PW (1 cm)
OW
SWI
10
PW (1 cm)
PW (2.5 cm)
1
H2O CTRL LALL GE
DE
NP
Sediments
LD UMF LMF
1.12
1.1
1.08
1.06
1.04
1.02
CTRL
LALL
GE
DE
NP
LD
UMF
LMF
Biotic Ligand Model Toxic Units
Comparison for Copper
10
1
0.1
< <
<
<<
<
H20 CTRL LALL GE
DE
NP
Fig 5. Mean body condition index of white sturgeon. Fish
in treatment sediment exposures were compared to fish
exposed to reference sediments. From left to right, the
dashed lines separate the control, reference, and treatment
groups, respectively.
Error bars represent standard
deviation.
<
<
OW
<
<
<
0.01
SWI
PW (1 cm)
PW (2.5 cm)
0.001
H20 CTRL LALL GE
DE
NP
LD UMF LMF
Sediments
Fig 6. Toxic unit (TU) analyses for copper based on Biotic
Ligand Model predictions. Bars represent average TUs for
overlying water (OW), sediment-water interface (SWI), and
porewater (PW) for the various sediment exposures. Error
bars represent standard deviation. TUs were considered
“<“ when more than 80% of measurements were below DL
or associated with a contaminated blank.
Conclusions
PW (2.5 cm)
0.01
1.14
LD UMF LMF
Zn Concentrations in Matrices
Zinc (µg/L)
Map 1. [A] Columbia River basin. [B] Reference sediment locations on the Columbia River circled in red [Lower
Arrow Lake (LALL), and Genelle (GE)]. [C] Treatment sediment locations on the Columbia River circled in red
[Deadman’s Eddy (DE), Northport (NP), Little Dalles, (LD), Upper Marcus Flats (UMF), and Lower Marcus Flats
(LMF)]. Map adapted from McLelllan and Howell.
DE
Sediments
Pb Concentrations in Matrices
Lead (µg/L)
LD
PW (1 cm)
H2O CTRL LALL GE
LD UMF LMF
Sediments
NP
OW
0.001
UMF
Body Condition Index of Sturgeon at
Termination
Sediments
PW (2.5 cm)
0.1
Body Condition Index (K)
• There were no significant differences (p > 0.05) in K of fish exposed to treatment sediments
versus reference sediments at termination (Fig 5).
BLM Toxic Units
• TUs were calculated for all matrices (OW, SWI, and PW at 1 and 2.5 cm) for every exposure
chamber (only Cu data shown; Fig 6).
• Mean TU values approached or exceeded the threshold of 1.0 TU for only Cu in deep
porewater (Fig 6) for DE and LD sediment exposures.
• TUs for Cd, Zn, and Pb were less than TU threshold values of 1.0 (data not shown).
H20
PW (2.5 cm)
GE
Fig 4. [A] Survival analyses over the duration of the experiment and [B] overall survival
in the control, reference, and treatment exposures at the termination of the experiment.
From left to right, the dashed lines separate the control, reference, and treatment
groups, respectively.
1
1
100
CTRL
White Sturgeon Survival at Termination
[B]
Fig 2. Schematic of exposure chamber with DGTs and peepers for
passive sampling (PW 1 cm) and airstones (PW 2.5 cm) for active
sampling of porewater at 1 and 2.5 cm depths, respectively.
Results
[C]
White Sturgeon Survival Analyses
[A]
Survival at Termination (%)
Methods
Survival Analyses
• Average survival in the reference groups was greater than 75%, and average survival in all
treatment groups was greater than 70% (Fig 4).
• There were no significant differences in survival between treatment and reference groups at
the termination of the study (p > 0.05).
• NP treatment group had the greatest average survival (90%) while the UMF treatment group
had the least average survival (70%).
Condition Index
Sturgeon (Acipenseridae) populations are threatened throughout the world and have
been decreasing over the past century in North America, Asia, and Northern Europe. In
North America, populations of white sturgeon (WS; Acipenser transmontanus) are
declining in the north-western USA and British Columbia, Canada, primarily due to poor
annual recruitment. Pollution has been hypothesized as one potential cause for poor
recruitment in the Columbia River (CR) between Grand Coulee Dam in the USA, and the
Hugh L. Keenleyside Dam in southern British Columbia, Canada. Potential past and/or
present sources of pollution include metallurgical facilities, pulp and paper mills, and
other industrial and municipal sources. There are concerns that sediments in the CR
may be contaminated and are potentially toxic to WS given their epibenthic nature.
Specifically, there are concerns about the potential toxicity to WS early life stages,
including the early hiding stage where fry are in proximity to sediments, to contaminants
such as metals that are associated with sediments. The present study evaluated effects
of metals in sediments collected from the CR on early life stage WS. WS were exposed
to field-collected sediments from areas considered to be suitable WS habitat, and
containing a range of metal concentrations. Reference sediments were collected from
sites in Canada, upriver from the study area. Studies were conducted at the Aquatic
Toxicology Research Facility, University of Saskatchewan, Saskatoon, SK, under
simulated fluvial, flow-through, conditions from hatch to 60 days post hatch (dph). To
aid in the risk assessment of WS exposed to sediments of concern, the Biotic Ligand
Model (BLM), a metal speciation and predictive toxicity model, was employed to
characterize potential toxicity of CR sediments to early life stage WS. Results
presented here are preliminary and are the opinion of the authors. The study will be
submitted for consideration as part of a baseline ecological risk assessment being
conducted at the upper Columbia River in eastern Washington State, USA.
Centre, University of Saskatchewan, Saskatoon, SK, Canada.
2HDR | HydroQual, East Syracuse, NY, USA.
3Dept. Veterinary Biomedical Sciences, University of Saskatchewan, Saskatoon, SK, Canada.
4City University of Hong Kong, Hong Kong, China.
5King Saud University, Riyadh, Saudi Arabia.
6Cardno ENTRIX Inc., Saskatoon, SK, Canada.
7School of the Environment & Sustainability, University of Saskatchewan, Saskatoon, SK, Canada.
Geometric mean of Cu Toxic Units
Introduction
LD UMF LMF
Sediments
Fig 3. Average measured metal concentrations in control, reference, and treatment exposures in overlying water (OW), sediment water
interface water (SWI), and porewater at 1 cm (PW 1 cm) and 2.5 cm (PW 2.5 cm) depth.
• There were no statistically significant effects on any of the measurement endpoints.
• Concentrations of metals in sediment exposure matrices were less than the most sensitive
LC values determined in previous studies (Vardy et al. 2011) for Cd, Zn, and Pb, and only
approached LC values in PW for Cu in DE and LD sediment exposures. This is consistent
with the BLM toxic units comparison for copper (Fig 6.)
• Results of this study indicate that exposure to contaminants through sediments in the
upper Columbia River is unlikely to significantly contribute to the poor recruitment of WS.
References
BLM, 2007. The Biotic Ligand Model Windows Interface, Version 2.2.3: User’s Guide and Reference
Manual, HydroQual, Inc, Mahwah, NJ, USA.
Vardy, D.W., Tompsett, A.R., Sigurdson, J.L., Doering, J.A., Zhang, X., Giesy, J.P., Hecker, M., 2011.
Effects of sub-chronic exposure of early life stages of white sturgeon (Acipenser
transmontanus) to copper, cadmium and zinc. Environ. Toxicol. Chem. (In press )
Acknowledgments: Funding for this project was provided by an unrestricted grant from Teck American Incorporated to the Univ. Sask. Thanks to the Kootenay Trout Hatchery, the UofS graduate program and team, the UofS ATRF, the US-EPA and the UCR RI/FS technical advisory team for their advise and support during these studies.