Ecological Impacts of Contaminants in an Urban Watershed

Ecological Impacts of Contaminants in an Urban Watershed
DOE FRAP 1998-25
Prepared for:
Environment Canada
Environmental Conservation Branch
Aquatic and Atmospheric Sciences Division
700-1200 West 73rd Avenue
Vancouver, BC V6P 6H9
Prepared by:
John S. Richardson1, Ken J. Hall1, Peter M. Kiffney1, and Jacqueline A. Smith1, and
Patricia Keen2
1
Dept. of Forest Sciences, Institute for Resources and Environment, and Dept. of
Civil Engineering, The University of British Columbia, Vancouver, B.C. ,
2
B.C. Research Incorporated, 3650 Wesbrook Mall, Vancouver, B.C.
June 1998
DISCLAIMER
This report was funded by Environment Canada under the Fraser River Action Plan through the
Environmental Quality Technical Working Group. The views expressed herein are those of the
authors and do not necessarily state or reflect the policies of Environment Canada
Any comments regarding this report should be forwarded to:
Aquatic and Atmospheric Sciences Division
Environmental Conservation Branch
Environment Canada
700-1200 West 73rd Avenue
Vancouver, B.C.
V6P 6H9
ii
Acknowledgements
We are grateful for the financial support of this work by Environment Canada's Fraser River
Action Plan, the Natural Sciences and Engineering Research Council (Canada), and the University
of British Columbia.
iii
Abstract
Small, urban streams are affected by a variety of alterations including contamination, and the
Brunette River drainage (Burnaby, BC) was studied as an example. We studied sediment
chemistry, direct toxicity of sediments, patterns of benthic invertebrate community structure, and
then experimentally tested the direct effects of metals on a “pristine” stream community. The
Brunette watershed has concentrations of heavy metals, particularly copper, lead, and zinc,
associated with sediments and transported during high flow that regularly exceed water quality
guidelines. Sediments were toxic to organisms in toxicity assays but highly variable which may be
attributable to interactions amongst contaminants and the variation in organic content of the
sediments, a known binding site. A survey of benthic invertebrates across the lower mainland
showed that there were large differences in the organisms present in urban versus rural streams.
Species that were largely absent from the Brunette River watershed were also those species most
sensitive to heavy metal exposure in our flow-through experimental stream experiment beside a
clean, mountain stream. These results show that contaminants have a large effect on urban stream
communities.
iv
Résumé
Les petits cours d’eau urbains subissent toute une gamme de modifications incluant la
contamination, et le bassin versant de la rivière Brunette (Burnaby, C.-B.) a été étudié à titre
d’exemple. Nous avons étudié la chimie des sédiments, la toxicité directe des sédiments, la
structure de la communauté des invertébrés benthiques, avant d’éprouver expérimentalement les
effets directs de métaux sur la communauté d’un cours d’eau «vierge». Le bassin versant de la
rivière Brunette présente des concentrations de métaux lourds, en particulier de cuivre, de plomb
et de zinc, associés aux sédiments et transportés en période d’écoulement élevé, qui dépassent
régulièrement les seuils établis dans les lignes directrices en matière de qualité de l’eau. Lors de
tests de toxicité, les sédiments étaient toxiques pour des organismes, mais présentaient une grande
variabilité qui peut être attribuable à des interactions entre contaminants et à la variation de la
teneur en matière organique, un site de liaison connu. Un relevé des invertébrés benthiques
exécuté dans la vallée du bas Fraser a révélé de grandes différences quant aux organismes présents
entre les cours d’eau urbains et les cours d’eau ruraux. Les espèces en grande partie absentes du
bassin de la rivière Brunette étaient également les espèces les plus sensibles à l’exposition aux
métaux lourds dans le cadre de notre expérience en cours d’eau expérimental coulant à côté d’un
cours d’eau de montagne propre. Ces résultats montrent que les contaminants ont un effet
important sur les communautés des cours d’eau urbains.
v
Table of Contents
Acknowledgements .................................................................................................................... iii
Abstract ..................................................................................................................................... iv
Résumé ....................................................................................................................................... v
Table of Contents....................................................................................................................... vi
List of Tables ............................................................................................................................ vii
List of Figures........................................................................................................................... vii
Introduction ................................................................................................................................ 1
Comparison of Water and Sediment Quality to Established Criteria and Guidelines...................... 2
Toxicity Bioassays....................................................................................................................... 2
Benthic Invertebrate Community Structure.................................................................................. 4
Response of a stream macroinvertebrate community from a pristine, southern BC stream to metals
in experimental mesocosms. ........................................................................................................ 5
Summary..................................................................................................................................... 7
References .................................................................................................................................. 7
Figures........................................................................................................................................ 8
Tables ....................................................................................................................................... 11
vi
List of Tables
Table 1. Water quality in the Brunette River Watershed during storm events. ............................ 12
Table 2. Trace metals in surface sediments in the Brunette River Watershed.1 ............................ 13
Table 3. Sediment toxicity in the Still Creek area of the Brunette Watershed.............................. 14
Table 4. Chironomid and Hyalella sediment toxicity in the Brunette Watershed......................... 15
List of Figures
Figure 1. Ordination (principal components analysis) of benthic communities from small, urban,
agricultural and forestland streams.. ...................................................................................... 9
Figure 2. Relationship between final benthos density and metal dose rate from experimental
stream channels at Mayfly Creek, 1995. .............................................................................. 10
vii
Introduction
Water and sediment quality monitoring have demonstrated that a variety of trace metal and
organic contaminants are entering the aquatic environment of the Brunette River watershed as a
result of non-point pollution from urban stormwater runoff. These contaminants vary considerably
over a rainfall event, seasonally, and spatially throughout the watershed.
To determine the ecological impacts of these contaminants a variety of techniques can be utilised.
The water and sediment quality values can be compared to quality guidelines and criteria to
determine when the values, considered safe for the protection of aquatic life, are exceeded. There
are a variety of criteria and objectives in the literature that can be selected for comparison. The
problems with this kind of comparison is that the values can vary considerably over space and
time compared to the movements and life history characteristics of the organisms that live in these
environments. Also, the established values only consider a single substance at a time and in reality
an aquatic organism has to respond to the cumulative effects of these contaminants. Therefore
other techniques in addition to the chemical and physical conditions of the environment are often
used to determine these overall integrative impacts. In the Triad approach, a combination of
sediment chemistry, toxicity bioassays, and benthic invertebrate community structure studies are
utilised to evaluate the ecological impacts of contaminants discharged to the aquatic environment
(Chapman,1990; Chapman et al. 1996). To determine cause and effect relationships it is often
necessary to set up controlled ecosystem experiment studies using containment systems such as
mesocosms or microcosm studies so that one can compare the variables that impact the
components of the ecosystem being investigated. For all of these studies, it is always important to
establish appropriate controls for comparison purposes.
This section of the report uses these techniques to determine the health of the aquatic ecosystem
in this urban watershed.
1
Comparison of Water and Sediment Quality to Established Criteria and
Guidelines
The data presented in Table 1 summarises the water quality conditions at the three stormwater
stations monitored over the period of a year in the Brunette watershed. The data are presented as
the percent of the time that the storm event mean concentration of contaminant exceeds the
water quality objectives. This comparison indicated that copper is a problem throughout the
watershed during all storm events, while other metals such as zinc and lead are more predominant
in the reaches of Still Creek that are more heavily impacted by traffic activity and have high levels
of impervious surface.
Since many of the contaminants are associated with sediments, it is relevant to compare the
sediment contamination to sediment criteria (Table 2). These data are also presented as a
percentage of the number of sites monitored (36) to provide some indication of the ecological
risk involved in living in different areas of the watershed if you are a sediment dwelling organism.
Lead appears to still be the most critical sediment contaminant when the data are compared to the
federal PEL values, even though the removal of lead from gasoline has reduced levels when
compared to previous studies (see Section 3.5). A comparison to the watershed specific criteria
established by the provincial government indicate that Cu, Pb, and Zn all present major ecological
problems in the sediments of this urban watershed. This creates a problem in determining which
criteria should be used in trying to manage the contamination problem. Again, these criteria only
consider the contaminants one at a time and do not consider the potential availability of the
sediment bound contaminants to the organisms.
Toxicity Bioassays
Previous studies demonstrated that the stormwater runoff was periodically toxic to Daphnia (Hall
and Anderson, 1985). Further laboratory investigations in the same study demonstrated that other
water quality characteristics such as pH and the suspended solids level were also important in
2
regulating trace metal toxicity to daphnia. The Microtox® photoluminescence bioassay
demonstrated that it was the suspended solids component of the urban runoff that was the most
toxic to bacteria (see section 3.3, Figure 8). Therefore toxicity studies during the current
investigation placed emphasis on monitoring the sediment toxicity.
Chironomid (Chironomus tentans) and amphipod (Hyalella azteca) bioassays were conducted on
stream sediment samples using the Environment Canada draft biological test methods
(Environment Canada, 1997,1996). The chironomid bioassays included both survival and weight
change compared to controls. Early chironomid bioassays compared changes to a control
sediment from Musqueam Creek (Pacific Spirit Park) (Smith, 1994). while later bioassays used
silica sand for controls and did standard toxicant tests. The Microtox solid phase
photoluminesence microbial bioassay was also conducted on stream bed sediments.
The sediment toxicity at three of the Still Creek stations was higher for all three toxicity indicators
(chironomid survival, weight and the Microtox bioassay) when compared to the control
Musqueam station (Table 3). During the month of August the toxicity of two Still Creek stations
and the Musqueam station for the chironomid bioassay were similar with both areas showing poor
survival. (22-36%) This is attributable to the low flow at the Musqueam station in August when
there was some increase in the trace metals associated with the < 0.63 µ fraction which could
come from the Marine Drive road crossing above the site. There were no good correlations
between the bioassay results and the contaminant (trace metal and PAH) concentrations which
indicates that the toxicity is probably related to a variety of contaminants some of which maybe be
interacting . The three Still Creek stations which were the most toxic also had the highest
sediment organic matter content which could relate to the biological availability of some of the
contaminants.
Another set of sediment bioassays were conducted in 1996 with chironomids and Hyalella (Table
4). The controls in these bioassays were conducted on silica sand and standard toxicant tests were
performed as per the Environment Canada protocol to compare the sensitivity of different broods
3
of invertebrates. Overall these bioassays did not show the same level of sediment toxicity as the
previous bioassays and there was no consistent relationship between the toxicity response of the
two different organisms. Only one sample (Oct.1, 1996- Still Cr. at Renfrew) was statistically
more toxic than the controls to Hyalella while two samples(April 24, 1997, at Still Cr. stations
Renfrew and Douglas ) demonstrated chironomid toxicity significantly different from the controls.
One concern with the controls in this established protocol is that there is considerable build-up of
ammonia in the samples (2-10 mg/L NH3-N) which could affect their growth and survival
(sometimes in the 80-85% region) and their statistical comparison to the field sediments.
Benthic Invertebrate Community Structure
The benthic invertebrates were sampled in the shallow water areas with a Surber sampler.
Triplicate samples were taken at each stations. Several control streams with few anthropogenic
impacts were also sampled as controls for comparison. The invertebrates were classified to the
lowest practical taxon and a principal component analysis was conducted to segregate the stations
that were impacted the most (Figure 1). The more impacted stations show up in the lower left
hand corner (negative axes area) of this principal component analysis diagram. This area of the
diagram is weighted heavily by high numbers of oligochaetes and this area of the diagram contains
many of the Brunette watershed stations (BR- 08,11,16,28,30,and 32). The three Brunette
stations at the top of this diagram (BR-06,19,& 20) are just downstream of the lakes in the
watershed and are obviously affected by invertebrate drift from the lakes. The control streams,
namely Pepin Cr. (PEPN), Bertrand Cr. (BTRD), Anderson Cr. (ANDN), the Little Campbell
River (CAMP) and Salmon River (SA#), located in less developed areas of the Lower Fraser
River valley, are on the far right side of the diagram and their positions on the diagram reflects
larger numbers of the more pollution sensitive species. These community bioassays should be
useful indicators of the cumulative effects of chemical and physical stress in these aquatic
environments
4
Response of a stream macroinvertebrate community from a pristine, southern
BC stream to metals in experimental mesocosms.
Small, urban streams are potentially impacted by an assortment of effects from changes in
hydrology, modifications of the channel form, to alteration of water quality. In the Brunette River
in the lower mainland, surface water concentrations of trace metals regularly exceed drinking
water standards and are likely to contribute to impairment of the biological communities of the
river and its tributaries. Survey work on a series of streams within the Brunette River drainage
and into agricultural and forested land showed a clear gradient in the community structure of
benthic invertebrates associated with the degree of urban development.
With increasing
development there was a decline in densities of several species of mayflies that are otherwise
common in less disturbed streams, and increases in oligochaete worms and chironomids. This
indicates modification of the stream network has detrimental effects on the ecosystem but does
not identify the actual cause.
One hypothesis for the apparent impairment of the biological community in urban streams is the
toxic effect of elevated concentrations of various trace metals. To test this hypothesis we chose
to experimentally elevate metal concentrations to streamside experimental stream troughs
receiving water and colonists from an otherwise “pristine”, forested stream. The site for the
experiment was Mayfly Creek in the Malcolm Knapp Research Forest of the University of British
Columbia. The experiment was performed twice, first in September 1995 and then in July 1996,
each time using mixtures of metals in fixed ratios but at a range of dose levels. Copper and zinc
were added in both years, and in 1995 lead and manganese were also included. After a period to
allow colonisation of the experimental channels the metal doses were added for a one-week
period during which measures of emigration rate, dissolved metal concentrations, and final benthic
densities were taken.
The elevated metal concentrations had a significant negative impact on the final densities of the
overall benthic community (p<0.025, Figure 2). These impacts were not uniformly distributed
amongst taxa. Some taxa showed strong declines in abundance as metal doses increased, such as
5
the mayflies Paraleptophlebia spp. and Baetis spp. and the oligochaete worms. Other taxa
showed no significant change in densities associated with trace metal additions, such as
chironomid midges or the caddisflies Lepidostoma spp. It is possible that there may have been
shifts in the composition of chironomid species since we did not identify that taxon beyond the
family level.
Rates of emigration (“drift”) turned out to be a poor measure of biological response to the
experiment. There was a weak trend to increasing rates of emigration as dose increased, but there
was too much variation in the measure to detect a significant relationship. The latter was true
regardless of whether species or total numbers were considered.
Other measures, such as
bacterial respiration rates, chlorophyll a concentrations, and algal composition showed no
detectable impact of the experimental additions.
The results of these experiments indicate that taxa that were sensitive to the toxic effects of trace
metals in the stream channels were many of the same taxa that occurred at lower densities in
urban streams. These results don’t necessarily confirm that the trace metals are the sole causal
agent for impairment of urban streams, but are consistent with the hypothesis that metal pollution
has impacts on the stream ecosystem. Whether other taxa are reduced in abundance in urban
streams for other reasons and whether there are interactive effects of other modifications to the
streams in reducing benthic densities and diversity cannot be addressed at the experimental scale.
Trace metals are a contributing factor to modification of urban stream ecosystems.
Summary
From the invertebrate sediment toxicity bioassays and the community structure analysis it is
obvious that the aquatic ecosystem of the Brunette is being impacted. However, there is
considerable variability in the sediment toxicity at the different stations to the different organisms
during the different sampling periods. This is not unexpected considering the highly dynamic
nature of the flow and contaminant transport and deposition in the watershed. The mesocosm
6
flow through studies at the low trace metal concentrations indicated that some groups of
organisms are affected. It would be very difficult to set up the complexity of flow variation,
contaminant ratios, and loading variation that the natural populations experience in some streams
of the highly urbanised Brunette watershed.
References
Chapman, P.M. 1990. The Sediment Quality Triad approach to determining pollution induced
degradation. Sci. Total Envir. 97/98: 815-825.
Chapman, P.M., M.D.Paine, A.D. Arthur and L.A. Taylor. 1996. A Triad Study of Sediment
Quality Associated with a major, Relatively Untreated Marine Sewage Discharge. Mar.
Poll. Bull. 32: 47-64.
Environment Canada. 1997. Test for Growth and Survival in Sediment Using Larvae of
Freshwater Midges Chironomus tentans. Environment Canada Ottawa.
Environment Canada.1996. Test for Growth and Survival in Sediment Using the Freshwater
Amphipod Hyalella azteca. Preview to final manuscript. Environment Canada, Ottawa.
Hall, K.J. and B. Anderson. 1988. The Toxicity and Chemical Composition of Urban Stormwater
Runoff. Can. J. of Civil Engr. 15: 98-106.
Smith, J. A. 1994. Sediment Toxicity Testing: Battery Test Evaluation of Shallow Urban Streams
and the Effect of Sampling Method on Toxicity. MASc. Dept. of Civil Engr., The
University of British Columbia, Vancouver, B.C.
7
Figures
8
5
BR20
4
Principal Component 2 (15.5%)
BR19
3
BR06
Tricladida
Cheumatopsyche
Tanytarsini
2
Amphipod A
1
SU03
SU13
0
-1
-2
BR17A
SU04
CAMP
PEPN
BR16
Tubificidae
BR28
BR17B
Orthocladiinae
Naididae
BR30 BR11
Lymnaeidae
SA19
Hydra
Leptophlebiidae
Nemouridae
Ephemerella
SA04
SA09
BR32
SA20
SA05
ANDN
Baetis
BTRD
BR08
-3
BR33
-3
-2
-1
0
1
2
Principal Component 1 (34.7%)
3
4
Figure 1. Ordination (principal components analysis) of benthic communities from small,
urban, agricultural and forestland streams. Sample sites identified by circles with site code
beside them (BR# - Brunette River watershed, SU# - Sumas River watershed, SA# - Salmon
River watershed, others are additional “control” streams: ANDN - Anderson; BTRD - Bertrand;
PEPN - Pepin; CAMP - Little Campbell River). Lines with taxonomic names indicate the loading
and direction of the numerically predominant taxa with significant loadings.
9
500
Total Benthos (log10x)
Rs= -0.5, p<0.025
slope (log10-log10) = -0.122
100
1
10
100
log10 Mn (ug / L)
1000
Figure 2. Relationship between final benthos density and metal dose rate from
experimental stream channels at Mayfly Creek, 1995. The coefficient from the Spearman’s
Rank Correlation (Rs) is shown along with the slope from a linear regression of log-log data.
10
Tables
11
Table 1.
Water quality in the Brunette River Watershed during storm events.
Parameter1
Specific
Conductivity
Total Suspended
Solids
COD
NO3-N
NH3-N
Total Phosphorus
Total Copper
Total Lead
Total Zinc
Total Manganese
Exceedence of
Water Quality Objectives3
Storm Event SMC2
Renfrew
Gilmore
Eagle
Renfrew
Gilmore
Eagle
118
132
108
---
---
---
55
98
39
80
90
50
40
0.91
0.43
0.22
0.066
0.007
0.096
0.139
46
0.88
0.62
0.39
0.060
0.011
0.122
0.250
23
0.68
0.29
0.15
0.020
0.003
0.043
0.270
----10
--100
0
92
---
----20
--100
10
100
---
----20
--100
0
58
---
All values in mg/L except for Specific Conductivity in µS/cm.
Storm event sample mean concentration is a technique where each sample during the storm
event is normalized for flow and a weighted single value calculated for the entire storm event.
Data is for 10 to 12 storm events over a 16 month period.
3
Water quality objectives are for the Brunette watershed and are presented as the percent of the
SMC values that exceeded the objectives. For trace metals the quality data were compared to the
maximum allowable objective not the average objective since storm events are episodic in nature
and do not persist for long
1
2
12
Table 2.
Trace metals in surface sediments in the Brunette River Watershed.1
Median Concentration2
Trace Metal
Cadmium
Chromium
Copper
Iron
Lead
Manganese
Mercury
Nickel
Zinc
Total
<3
26
71
2.1
50
768
0.102
13.6
154
Extractable
4
31
0.62
46
486
<6
75
1
Exceedence of Sediment Quality Criteria
(%)3
TEL
PEL
Brunette
19.4
0
83.3
5.5
94.4
86.1
41.6
100
14.2
2.8
57.1
25
0
74.2
2.8
97.2
Three surface sediments samples were collected in 1993-94 at each of 36 stations.
All values in mg/kg dry weight except iron as %. Values are for the <180µ fraction of sediment.
Total for hot nitric acid digestion, extractable for cold 0.5 M HCl digestion.
3
Exceedence relates to total metal of the <177µ fraction and is expressed as % of the total
number of stations: n=36 except for mercury, n=33. TEL = threshold effects level and PEL =
probable effects level reported by the Federal government, while Brunette criteria have been
established specifically by the B.C. Provincial government for the Brunette River watershed.
2
13
Table 3.
Sediment toxicity in the Still Creek area of the Brunette Watershed.
Chironomid Survival (%)
Chironomid Weight (mg)
Microtox (EC50)
Date
July 4
Aug. 2
Aug. 24
Oct. 14
Dec. 4
Still
Musqueam
Still
Musqueam
Still
Musqueam
30
36
26
12
13
73
35
22
69
73
1.3±0.6
1.5±0.7
2.0±1.0
1.1±0.6
0.4±0.2
2.1±1.1
1.5±0.5
2.0±0.6
2.7±1.1
1.2±
3.6
11.5
2.96
0.41
4.28
19.0
NT
10.98
NT
NT
Microtox is for 15 min. contact and is for whole sediment solid phase test. Still Creek stations
were all different for each bioassay and sediments were from Willingdon, Douglas, Atlin,
Lougheed, and Westburne stations in the sequential order presented above. The Musqueam
station sediment samples were from the same station on all dates and served as a control
14
Table 4.
Chironomid and Hyalella sediment toxicity in the Brunette Watershed.
Date
Oct.1, 1996
Apr.24, 1997
Station
Chironomid Survival
(%)
90
90
100
97.5
82.5
95
85
75
Still Cr. (Renfrew)
Still Cr. (Gilmore)
Still Cr.(Douglas)
Eagle Cr.(Piper)
Still Cr. (Renfrew)
Still Cr. (Gilmore)
Still Cr. (Douglas)
Eagle Cr. (Piper)
15
Hyalella Survival
(%)
20
97.5
90
100
100
95
100
100
Download PDF